Emerging Infectious Diseases of the 21st Century

I. W. Fong · David Shlaes · Karl Drlica

Editors

Antimicrobial Resistance
in the 21st Century
Second Edition
Chapter 1
Introduction: Coordinated Global Action
Is Needed to Combat Antimicrobial
Resistance

I. W. Fong

Antimicrobial resistance is a global dilemma that threatens the health and safety of
populations in all countries of the world. Urgent actions are needed to be taken
before it reaches a critical stage, when large numbers of people in communities
cannot be treated for life-threatening infections due to lack of effective drugs.
Although the threat is most imminent from antibacterial resistance to commonly
used antibiotics for infections seen regularly in intensive care units and hospitals, it
is more prevalent and widespread and involves a wide spectrum of microbes. This
second edition of “Antimicrobial Resistance and Implications for the 21st Century”
provides not only updates and advances since the original edition but provides a
wider spectrum of topics on the issue. Although most chapters of this new edition
address issues of common bacterial resistance, others provide up-to-date reviews on
resistance trends with viruses, including human immunodeficiency virus [HIV] and
human herpes group of viruses.
To understand the evolution of microbial resistance, it is appropriate to review
historical aspects. Development of antimicrobial chemotherapy is usually attributed
to Paul Ehrlich [“father of chemotherapy”] based on his quest to find a cure for
parasitic infections, toward the latter part of the nineteenth century, with natural
dyes and heavy metals [mercury and arsenicals]. Penicillin was subsequently
discovered in 1928 and administered clinically in the 1940s; sulfonamides were
introduced clinically in 1937 [1]. Thus, the “antibiotic era” was under way by the
early 1940s. Penicillin and other antibiotics were initially derived from environmental
fungi and bacteria, often with improvements made by chemical synthesis. Hence,
the origin of antibiotics is through naturally derived substances produced to
antagonize or inhibit the growth of other microorganisms, probably due to an evo-
lutionary process that protects environmental niches of the producing organisms.

I. W. Fong (*)
Department of Medicine, University of Toronto, Toronto, ON, Canada
e-mail: [email protected]

© Springer International Publishing AG, part of Springer Nature 2018 1

I. W. Fong et al. (eds.), Antimicrobial Resistance in the 21st Century, Emerging Infectious
Diseases of the 21st Century, https://doi.org/10.1007/978-3-319-78538-7_1
2 I. W. Fong

Streptomycin and the precursor of cephalosporins were also obtained from soil
microbes in the 1940s. Penicillin was first used in 1941, and by 1944 penicillinase-
producing Staphylococcus aureus was described. Streptomycin was introduced
clinically in 1944 for treatment of tuberculosis, but resistance soon developed dur-
ing treatment [2]. By the mid-1950s, most of the major antibiotic families, including
aminoglycosides, chloramphenicol, tetracycline, and macrolides, had been devel-
oped [1]. Synthetic chemical agents with antibacterial activity were introduced in
the early 1950s with para-aminosalicylic acid and isoniazid as antituberculosis
agents; nitrofurantoin was found around the same time, followed by trimethoprim
in 1956. Nalidixic acid, discovered in the early 1960s, was the precursor of the fluo-
roquinolones, with norfloxacin and ciprofloxacin developed in the 1980s. Rifampin
was introduced for tuberculosis in 1968.
The spectacular success of antimicrobial therapy led to widespread use and
emergence of resistance. Pharmaceutical companies saw profit in new, more potent
derivatives, which gave rise to broad-spectrum antipseudomonal penicillins and
second-generation cephalosporins in the 1970s, with subsequent introduction in the
1980s of third-generation cephalosporins. Later, beta-lactamase inhibitors com-
bined with broad-spectrum penicillins and carbapenems, new glycopeptides, newer
macrolides, later-generation quinolones, linezolid [a new class of oxazolidinone],
and glycylcyclines [tigecycline] were introduced. Currently, there are more than
100 antimicrobial compounds available.

1.1 Evolution of Resistance

It is important to understand the evolution of antimicrobial resistance in order to

tackle the problem. It was initially thought that antimicrobial resistance was a
modern, man-made phenomenon. Although penicillinase-producing bacteria were
recognized soon after the discovery of penicillin, it was not recognized as a problem
until clinical use of penicillin became widespread. By the end of the 1950s, 80% of
Staphylococcus aureus isolated from hospital patients were penicillin resistant due
to β-lactamase production [1]. This observation led to the development of
penicillinase-stable penicillins [methicillin, cloxacillin, and oxacillin] which were
introduced in the early 1960s to treat S. aureus. But resistance to methicillin was
seen within a year or so of its clinical introduction. Thus, this was a sign that
microbial resistance would be a problem, but in most cases this was overcome with
development of new, more potent compounds.
Many agents were developed that initially failed to compete well in the market-
place. The lack of use kept resistance from emerging, and now they are being used
more frequently to fill a niche. Vancomycin is one of these agents. It was approved
for use in 1958, but it was used sparingly due to concerns of toxicity and efficacy.
When methicillin-resistant S. aureus [MRSA] appeared in hospitals during the
1970s and then spread to other health-care facilities, vancomycin became more fre-
quently used, beginning in the 1980s [3]. Similarly, polymyxins were developed in
1 Introduction: Coordinated Global Action Is Needed to Combat Antimicrobial… 3

the 1950s, but toxicity kept them from widespread use for systemic infections. Now
they are gaining clinical use with infections caused by Pseudomonas aeruginosa,
Acinetobacter baumannii, and carbapenemase-producing Enterobacteriaceae that
are resistant to all other antibiotics [4]. Full vancomycin resistance among MRSA
strains is rare, probably because mutational changes cause impaired fitness for
MRSA as susceptibility decreases. As a result, vancomycin-­intermediate resistance
[VISA] is slowly becoming a problem. Resistance to polymyxins appears to be
emerging readily with increased use.
Over the years, development of new antibiotics to counter resistant bacteria led
to appearance of novel resistant strains to these new drugs, which become wide-
spread with increasing use of the antibiotics. This pattern has been seen with every
new class of antibiotic developed over the years. There is strong correlation
between the frequency and quantity of antibiotics used in humans and animals and
the rate of development of antibacterial drug resistance. The logarithmic growth of
resistant bacteria since the 1970s is reflected by the number of β-lactamase
enzymes identified during the antibiotic era. Before 1970, there were only several
β-lactamase enzymes described, and now about 900 β-lactamase enzymes have
been identified [2].
Mobile genetic elements, extrachromosomal self-replicating structures [plas-
mids] and transposons, found in bacterial cells provide a resistance threat that
maybe unconquerable. Our understanding of horizontal transfer of bacterial
resistance was heralded by the discovery of antibiotic resistance plasmids that could
be disseminated by bacterial conjugation in the mid-1950s [2]. Since 1989 we have
gained much greater knowledge of the genetics of bacterial resistance following the
discovery of integrons. Integrons are versatile genetic elements, commonly found in
bacterial genomes, that allow efficient capture and expression of exogenous genes.
Plasmids and transposons are considered mobile integrons. Integrons play a major
role in the spread of antibiotic resistance, particularly in Gram-negative pathogens;
the majority of these pathogens carry integrons with resistant genes [5]. The process
of microbial resistance is complex, and besides plasmids and DNA mutations
[acquired or heritable and transferable], other mechanisms include biofilms, which
harbor hypermutator bacteria that select for resistance more frequently, and pheno-
typic tolerance, a situation in which bacteria are not killed by antimicrobials [6].
Antibiotic pressure predisposes to resistance and tolerance. Despite the call for
intensive research on mechanisms of microbial resistance and development of novel
compounds to counteract the spread in 2012 [6], no innovative agent is on the
horizon.
The origins, evolution, genetics, and biochemistry of antibiotic resistance have
been studied over the last 60 years. Figure 1.1 outlines the history of antibiotic dis-
covery and subsequent development of antibiotic resistance. The emergence of anti-
biotic resistance of pathogenic bacteria after antibiotic development and clinical
use, plus the absence of resistance in bacteria of the pre-antibiotic era [7], suggested
that resistance is a modern phenomenon. However, metagenomic analyses of
authenticated ancient DNA from 30,000-year-old Beringian permafrost sediments
identified genes encoding resistance to β-lactam, tetracycline, and glycopeptide
4 I. W. Fong

Fig. 1.1 Discovery of antibiotics and evolution of antimicrobial resistance. Abbreviations: MRSA
methicillin-resistant S. aureus, VRE vancomycin-resistant enterococci, ESBL extended
β-lactamase

antibiotics [8]. Thus, antibiotic resistance is a natural phenomenon that predates the
modern selective pressure associated with clinical and animal use. This explains the
rapid emergence of resistance to new antibiotics; resistance will continue to emerge
with drugs now in development. Hence, it is predictable that new antibiotics will
select for preexistent resistance determinants that have been present and circulating
in universal, environmental microbial pangenome for hundreds of years [8]. Thus,
the era of antibiotics has shifted naturally to the era of antibiotic resistance. Our
challenge is to minimize the problem.

1.2 Defining the Problem

It is now evident that antimicrobial resistance is inevitable. Viable solutions are

needed to postpone the inevitable; development of novel antibiotics is only a
temporary remedy, especially since most “new” agents are chemical refinements of
old ones. Of the 8 new antibiotics approved by the US Food and Drug Administration
[FDA] since 2010, only one [bedaquiline for tuberculosis] is from a new drug class
[9]. Thus, resistance to most of these agents will likely develop rapidly. A
comprehensive, multipronged, coordinated approach is needed to combat
1 Introduction: Coordinated Global Action Is Needed to Combat Antimicrobial… 5

antimicrobial resistance. The World Health Organization [WHO] is best suited to

lead the fight and is already involved in the process. What are the necessary steps to
combat microbial resistance?
Curbing the global overuse of antimicrobials is probably the greatest challenge.
Unnecessary use of antibiotics for humans and animals is a major concern, even
after two decades of effort to reduce the flagrant abuse and overuse. A ban on
nontherapeutic use of antibiotics in animals and agriculture has been recommended
since 1969, but it has been very difficult to gain worldwide acceptance [4]. For
example, the European Union banned use of antibiotics for growth promotion in
animals in 2006, but this practice is still widespread in many other countries. Recent
analyses estimate that from 2010 to 2030 global utilization of antibiotics in the
livestock industry will increase by two thirds and that it will double in the growing
economies of Brazil, China, India, Russia, and South Africa [10]. There appears to
be no political will to ban antibiotics as growth stimulants for animal husbandry.
Companies that use deception to provide antibiotics to animals under cover of
different names should be charged hefty fines. As of 2010 only the Netherlands and
Scandinavia had successfully reduced antibiotic resistance levels by enforcing
antibiotic restrictions [4]. The Dutch government instituted a policy of requiring a
70% reduction of antibiotic use in animals between 2009 and 2015 and prohibited
use of new antimicrobials. These initiatives resulted in a 56% reduction in animal
antimicrobial use between 2007 and 2012 [11]. Thus, rollbacks can be achieved, but
how will multidrug-resistant bacteria that developed in other countries be kept out
of countries that restrict use?
The availability of inexpensive antibiotics is still largely uncontrolled in many
developing countries, and these drugs can be obtained from pharmacies without a
prescription. Even in Europe, persons can purchase antibiotics over the counter or
through the Internet in 19 countries. In 12 countries antibiotics can be obtained on
the black market or veterinary clinics [WHO Regional Office for Europe,
antimicrobial resistance [http://www.euro.who.int/en/health-topics/disease-
prevention/antimicrobial-resistance]. These practices continue to play a role in the
abuse and overuse of antimicrobials. One approach may involve education. For
example, enabling pharmacists to deliver accurate information and counseling on
proper antibiotic should be implemented.
Antibiotic-resistant bacteria respect no borders; international travelers can
acquire and spread these microbes. In a prospective longitudinal study of Dutch
international travelers, 34% of 1847 travelers acquired extended β-lactamase-­
producing Enterobacteriaceae [ESBL] that persisted for 12 months in 11% of the
respondents [12]. Among travelers to southern Asia, 75% acquired ESBL; the
frequency was 89% with travelers to India. Antibiotic use is a strong predictor of
carrying resistant genes, and travelers should be discouraged from using antibiotics
for self-limited infection such as traveler’s diarrhea.
6 I. W. Fong

1.3 Moving Forward

It has been argued that the lack of access to life-saving antibiotics is as important an
issue as antibiotic resistance. It has been estimated that universal access to antibiotics
could prevent 445,000 deaths out of 590,000 deaths from pneumonia [75% reduc-
tion] in 101 countries [13]. However, increased use of pneumococcal and
Haemophilus influenza type B vaccines could prevent up to 11.4 million days on
antibiotics – a 47% reduction in 75 countries. Carriage of multiresistant bacteria is
not restricted to travelers or developing countries. In a study from Germany, of 4376
patients admitted to general wards, third-generation cephalosporin-resistant
Enterobacteriaceae were detected from rectal swabs in almost 10% of patients
immediately after admission [14]. Risk factors for presence of these multiresistant
bacteria included prior antimicrobial treatment, travel outside Europe, stay in a
long-term care facility, and use of proton pump inhibitors for gastroesophageal
reflux.
Many high-income countries, including the US and parts of Europe, have created
national plans as well as regulation to address antibiotic resistance issues. However,
the brunt of the problem will be borne by low-income and middle-income countries
that cannot afford the newer, expensive drugs.
Health care-associated infections are a major source of the problem, and inten-
sive care units are “generators” of resistant bacteria. The empiric institution of
broad-spectrum antimicrobials for infections in very ill patients is the force behind
this problem. There is a great need for inexpensive, rapid and reliable microbiologi-
cal/molecular diagnostic tests to alleviate some of the empiric overuse of antibiotics
in health-care settings. Conventional culture methods usually take >2 days for iden-
tification and susceptibility determination, but a rapid multiplex polymerase chain
reaction [rmPCR] can provide identification in 1.3 h [15]. Moreover, rapid point-of-
care tests are needed to distinguish viral and noninfectious inflammatory conditions
from bacterial infections. Antibiotic stewardship in hospitals in North America and
other countries reduces antibiotic use, improves patient outcome, decreases adverse
events such as superinfection with Clostridium difficile and antibiotic resistance [to
a modest degree], and is cost-effective [16]. Wider adoption of stringent steward-
ship programs is needed for all community hospitals globally, but it will be difficult
to implement in resource-poor countries.
Inappropriate antibiotic use is still widespread for acute upper respiratory tract
infections despite attempts to curb the abuse in outpatient primary-care practice by
education. Efforts had been made to improve prescription behavior and provide
guidelines for antibiotic use across the USA, but success has been limited. Overall
antibiotic use for acute respiratory infections has significantly declined in children
[17], but use in adults remains high, especially for broad-spectrum antibiotics and
macrolides [18], as confirmed by recent data from the Veterans Affairs health system
[19]. A review of outpatient antibiotic use in the USA reported that about 13% of all
visits [about 154 million per year] resulted in antibiotic prescriptions of which at
least 30% were considered unnecessary [20]. Thus, unnecessary use of antibiotic
1 Introduction: Coordinated Global Action Is Needed to Combat Antimicrobial… 7

for acute respiratory or minor infections remains high. One method worth exploring
is for public health officials and medical associations to send frequent e-mail
messages to primary-care physicians concerning the dangers of antibiotic
overprescribing [no proven benefit]. Another option is to provide financial incentives
for not prescribing antibiotics by medical insurance companies. Delayed prescribing
[delay between receiving the prescription and collecting the drugs] has shown some
success in reducing antibiotic use [21]. In Thailand, the Antibiotic Smart Use
program has shown that alternative treatment options were important in restricting
antibiotic outpatient use, such as oral rehydration and zinc for diarrheal diseases,
and herbal drugs packaged in antibiotic-like capsules for viral upper respiratory
infections [22].
What is being implemented to combat antimicrobial resistance? Several coun-
tries are now taking steps to improve antibiotic prescribing, and the World Health
Day in 2011 was dedicated to antimicrobial resistance. The Infectious Disease
Society of America in 2011 outlined a road map to counter antimicrobial resistance:
regular surveillance and data collection on resistance patterns and prevalence, uni-
versal antibiotic stewardship for hospitals, and provision of research and develop-
ment [R&D] incentives for drug companies to facilitate licensing of novel
antimicrobials [23]. But as pointed out, new antimicrobials will simply delay the
problem. More recently, the Presidential Advisory Council on Combating Antibiotic
Resistant Bacteria and Innovative Medicines Initiative suggested public-private
partnership to provide financial resources to assist R&D [24]. However, proposed
funding cuts by the Trump administration to the CDC's antimicrobial resistance
[AMR] fund by 14% and the NIAID by 23% threatens the progress in the fight
against antimicrobial resistance [Boucher et al. Proposed US funding cuts threaten
progress on antinicrobial resistance. Ann Int Med 2017; 167:738–9]
The WHO or United Nations could provide leadership to facilitate multinational
global collaboration in this effort. The WHO has just released priority pathogens
list for R&D of new antibiotics: priority 1 [critical] includes carbapenem-
resistant A. baumannii, P. aeruginosa, and multiresistant Enterobacteriaceae;
priority 2 [high] includes vancomycin-resistant Enterococcus faecium [VRE],
vancomycin-intermediate MRSA, clarithromycin-resistant Helicobacter pylori,
fluoroquinolone-resistant Campylobacter and Salmonella spp., and third-
generation-resistant Neisseria gonorrhoeae; and priority 3 [medium] includes
penicillin-non-susceptible Streptococcus pneumonia, ampicillin-resistant H.
influenzae, and fluoroquinolone-­resistant Shigella spp. [25]. R&D for new drugs
for multiresistant tuberculosis and artemisinin-resistant malaria were previously
noted as high priority by the WHO. Development of new antibiotics to combat
resistant bacteria is a short-term solution to meet current needs. Resistance will
eventually develop to these agents as well. Innovative biological substances for
therapeutics where resistance is unlikely to develop are needed; research in this
area should be encouraged. This could include use of probiotics to counter and
prevent enteric colonization of resistant bacteria or bacteriophages to lyse
colonized resistant organisms such as MRSA and VRE.
8 I. W. Fong

On a global scale, there is much that can be done to reduce the risk of infection
and decrease the need for antibiotics, mainly in low-income countries. These
activities include wider and more universal use of vaccines, such as the pneumococcal
conjugate vaccine, H. influenzae type B vaccine, pertussis vaccine in pregnancy,
rotavirus vaccine, and measles vaccine, which could save lives and dramatically
reduce the use of pediatric antibiotics worldwide. Yearly universal influenza vacci-
nation of children and adults could also reduce the outpatient use of antibiotics for
respiratory infections in all countries. A major problem in developing and low-­
income countries is poor sanitation and lack of clean water, which predisposes per-
sons to a variety of infectious diarrheas that leads to antibiotic overuse and increased
antimicrobial resistance. In general, better infection control practices in health-care
institutions could reduce the need for antibiotics and lead to reduced prevalence of
resistance. In the USA alone, it is predicted that within 5 years multiresistant bacte-
ria could cause 340,000 deaths per year, but immediate implementation of a national
intervention strategy involving all elements of the healthcare network [hospitals,
nursing homes, etc.] through infection control and universal antibiotic stewardship
could save 37,000 lives and avert 619,000 infections over the next 5 years [26].
Current estimates are that antibiotic-resistant bacteria cause 2 million illness and
23,000 deaths each year in the USA with a annual cost to the health care system of
over $20 billion [CDC. Antibiotic resistance threats in the United States, 2013.
www.cdc.gov/drugresistance/pdf/ar-trhreats-2013-508.pdf]

1.4 Concluding Thoughts

Should antibiotic regulation and stewardship be instituted at the national societal

level? Public education on antibiotic use and national guidelines have had limited
impact. Policies that involve withdrawal of subsidies for expensive antibiotics
can have an effect as shown in Australia with quinolone prescriptions [27].
Overuse and abuse of antibiotics in humans and animals is causing pollution of
the coastal aquatic ecosystems with antibiotic-resistant pathogens, and concern
for effects on human and animal health should be of similar concern as global
climate change. A recent study from China has documented widespread pollution
of the estuaries along the coastal environment of China with over 200 different
antibiotic-resistant genes that affect almost all major classes of antimicrobials
[28]. The United Nations has now recognized the universal importance of antimi-
crobial resistance on human and animal health. On September 21, 2016, a high-
level meeting was convened with Heads of State for commitment to taking a
broad, coordinated approach in addressing the issue of antimicrobial resistance
across multiple sectors of human and animal health and agriculture [http://via-
jwat.ch/2nb4Dec2]. In summary, antimicrobial resistance is a global threat to
humanity with no immediate end in sight and it is considered an intetrnational
crisis. The cost in lives and to health care systems worldwide are huge and will
continue to rise in the foreseeable future. Evolution science indicates that
1 Introduction: Coordinated Global Action Is Needed to Combat Antimicrobial… 9

microbes will continue to develop resistance to future antimicrobials and cannot

be prevented, but we can limit the speed and magnitude of antimicrobial resis-
tance by a coordinated, multiprong approach.

References

1. Historical introduction. In: Greenwood D, Davey P, Wilcox M, editors. Antimicrobial chemo-

therapy. 5th ed. Oxford, UK: Oxford University Press; 2007. p. 1–10.
2. Davies J, Davies D. Origins of antibioptic resistance. Microbiol Mol Biol Rev. 2010;74:417–33.
3. Levine DP. Vancomycin: a history. Clin Infect Dis. 2006;42(Suppl. 1):S5–12.
4. Landman D, Gergescu C, Martin DA, Quale J. Polymyxins revisited. Clin Microbiol Rev.
2008;21:449–65.
5. Gillings MR. Integrons: past, present and future. Microbiol Mol Biol Rev. 2014;778:257–77.
6. Nathon C. Fresh approaches to anti-infective therapies. Sci Transl Med. 2012;4:140r2.
7. Hughes VM, Datta N. Conjugative plasmids in bacteria of the ‘pre-antibiotic’ era. Nature.
1983;302:725–6.
8. D’Costa VM, King CE, Kalan L, et al. Antibiotic resistance is ancient. Nature.
2011;477:457–61.
9. Deak D, Outterson K, Powers JH, Kesselhelm AS. Progress in the fight against multi-resistant
bacteria? A review of US Food and Drug Administration—approved antibiotics, 2010–2015.
Ann Intern Med. 2016;165:363–72.
10. Van Boeckel TP, Brower C, Gilbert M, et al. Global trends in antimicrobial use in food ani-
mals. Proc Natl Acad Sci U S A. 2015;112:5649–54.
11. Speksnijder DC, Mevius D, Bruscheke CJ, Wagenaar JA. Reduction of veterinary antimicro-
bial use in the Netherlands. The Dutch success model. Zoonoses Public Health. 2015;62(Suppl
1):79–87.
12. Arcilla MS, van Hattem JM, Haverkate MR, et al. Import and spread of extended-spectrum
β-lactamase-producing Enterobacteriaceae by international travelers [COMBAT study]: a pro-
spective, multicenter cohort study. Lancet Infect Dis. 2017;17:78–85.
13. Laxminarayan R, Matsoso P, Pant S, Brower C, Rottingen JA, Klugman K, Davies
S. Antimicrobials: access and sustainable effectiveness 1. Access to effective antimicrobials: a
worldwide challenge. Lancet. 2016;387:168–75.
14. Hamprecht A, Rohde AM, Behnke M, et al. Colonization with third-generation cephalosporin-­
resistant Enterobacteriaceae on hospital admission: prevalence and risk factors. J Chemother.
2016;71:2957–63.
15. Banerjee R, Teng CB, Cunningham SA, et al. Randomized trial of rapid multiplex polymerase
chain reaction—based blood culture identification and susceptibility testing. Clin Infect Dis.
2015;61:1071–80.
16. Wagner B, Filice GA, Drekonja D, et al. Antimicrobial stewardship programs in inpatient
hospital settings: a systematic review. Infect Control Hosp Epidemiol. 2014;35:1209–28.
17. Centers for Disease control and Prevention [CDC]. Office-related antibiotic prescribing for
persons aged ≤ 14 years—United States, 1993–1994 to 2007–2008. MMWR Morb Mortal
Wkly Rep. 2011;60:1153–6.
18. Lee GC, Reveles KR, Attridge RT, et al. Outpatient antibiotic prescribing in the United States:
2000 to 2010. BMC Med. 2014;12:96.
19. Jones BE, Sauer B, jones MM, et al. Variation in outpatient antibiotic prescribing for acute
respiratory infections in the veteran population. A cross-sectional study. Ann Intern Med.
2015;163:73–80.
20. Fleming-Dutra KE, Hersh AL, Shapiro DJ, et al. Prevalence of inappropriate antibiotic pre-
scriptions among US ambulatory care visits, 2010–2011. JAMA. 2016;315:1864–73.
10 I. W. Fong

21. Little P, Moore M, Kelly J, et al. the PIPS Investigators. Delayed antibiotic prescribing strate-
gies for respiratory tract infections in primary care: pragmatic, factorial, randomized con-
trolled trial. BMJ. 2014;348:g1606.
22. Dar OA, Hasa R, Schlundt J, et al. Antimicrobials: access and sustainable effectiveness.
Exploring the evidence base for national and regional policy interventions to combat resis-
tance. Lancet. 2016;387:285–95.
23. Spellberg B, Blaser M, Guidos RJ, et al. Infectious Disease Society of America [IDSA].
Combating antimicrobial resistance: policy recommendations to save lives. Clin Infect Dis.
2011;52(Suppl 5):S397-428.
24. Presidential Advisory Council on Combating Antibiotic-Resistant Bacteria. 2016. Accessed at
www.hhs.gov/sites/default/fioles/paccarb-final-report-03312016.pdf
25. WHO. Global priority list of antibiotic-resistant bacteria to guide research, discovery, and
development of new antibiotics. Accessed at http://www.int/medicines/publications/global-
priority-list-antibiotic-resistant-bacteria/en/. On 31 Mar 2017.
26. Slayton RB, Toth D, Lee BY, et al. Estimated effects of a coordinated approach for action to
reduce antibiotic-resistant infections in health care facilities—United States. MMWR Morb
Mortal Wkly Rep. 2015;64:826.
27. Cheng AC, Turnbridge J, Collignon P, Looke D, Barton M, Gottlieb T. Control of fluoroquino-
lone resistance through successful regulation, Australia. Emerg Infect Dis. 2012;18:1453–60.
28. Zhu YG, Zhao Y, Li B, et al. Continental-scale pollution of estuaries with antibiotic resistant
genes. Nat Microbiol. 2017;2:16270.
 Part I
Examples of Resistance
Chapter 2
Antimicrobial Resistance Among
Streptococcus pneumoniae

Catia Cillóniz, Carolina Garcia-Vidal, Adrian Ceccato, and Antoni Torres

2.1 Introduction

Antibiotic resistance is a direct result of antibiotic consumption [1, 2]. In the United
States, it is estimated that antibiotic resistance is responsible for more than 2 million
infections and 23,000 deaths each year, with a direct cost of $20 billion and addi-
tional productivity losses of $35 billion [3, 4]. Data from Europe showed that
approximately 25,000 deaths are attributable to antibiotic-resistant infections, with
a related cost of $1.5 billion annually [5]. The use of antibiotics in primary care is
high; the most frequent indications for their use are respiratory tract infections [6].
Streptococcus pneumoniae (pneumococcus) is the leading cause of community-­
acquired pneumonia and is considered to be a major cause of death of children
under 5 years old worldwide. In a recent report on global antibiotic resistance, pub-
lished by the World Health Organization (WHO) in 2014, pneumococcus was con-
sidered to be one of the nine bacteria of international concern [7]. Other infections
caused by pneumococcus include bacteremia, otitis media, and meningitis. In bacte-
rial meningitis, pneumococcus is associated with mortality rates ranging from 16%
to 37%. About 30–50% of adult survivors present permanent residual symptoms [8,
9]. The study by Van Boeckel et al. [10], regarding global antibiotic consumption
from 2000 to 2010, reported that it grew by more than 30%, from approximately 50

C. Cillóniz (*) · A. Torres

Department of Pneumology, Institut Clinic del Tórax, Hospital Clinic of Barcelona – Institut
d’Investigacions Biomèdiques August Pi i Sunyer (IDIBAPS), University of Barcelona
(UB) – SGR 911 – Ciber de Enfermedades Respiratorias (Ciberes), Barcelona, Spain
C. Garcia-Vidal
Infectious Disease Department, Hospital Clinic of Barcelona, Barcelona, Spain
A. Ceccato
Department of Pneumology, Hospital Nacional Alejandro Posadas, Palomar, Argentina

© Springer International Publishing AG, part of Springer Nature 2018 13

I. W. Fong et al. (eds.), Antimicrobial Resistance in the 21st Century, Emerging Infectious
Diseases of the 21st Century, https://doi.org/10.1007/978-3-319-78538-7_2
14 C. Cillóniz et al.

billion to 70 billion standard units. Penicillins, cephalosporins, and macrolides were

the three most consumed antibiotics in 2010. The three countries that consumed the
most antibiotics in 2010 were India with 13 billion standard units, China with 10
billion, and the United States with 7 billion standard units (a standard unit is the
number of doses sold; the IMS Health database identifies a dose as a pill, capsule,
or ampoule).
Resistance of pneumococcus against ß-lactams and macrolides is a major con-
cern worldwide. For example, in Southern European countries, the prevalence of
this resistance may be above 20% [11, 12]. The increased utilization of antibiotics,
the dissemination of several resistant clones, the ability of pneumococcus to undergo
serotype replacement and capsular switching, and the horizontal transmission of
antibiotic resistance genes make this pathogen very difficult to control. This chapter
summarizes currently available information regarding pneumococcal antibiotic
resistance.

2.2  asis of Antimicrobial Resistance in Streptococcus

B
pneumoniae

The nasopharyngeal carriage rate of pneumococcus is higher in children, mainly

during the first years of life (nasopharyngeal carriage rates range from 20% to
50% in healthy children). In contrast, with the healthy adult population, naso-
pharyngeal carriage rates range from 5% to 30%. Transmission of pneumococ-
cus from children to household contacts or adults is the principal cause of
nasopharyngeal carriage and the spread of antibiotic-resistant clones.
Pneumococcus undergoes genetic transformation and can acquire DNA from
other streptococci; during asymptomatic nasopharyngeal carriage, selection of
resistant pneumococcus occurs especially in children, because they carry pneu-
mococcus more often and for longer periods. Moreover, children are more fre-
quently exposed to antibiotics. Interestingly, the use of fluoroquinolones in
children is limited, because in animal models using young animals, develop-
ment of articular cartilage damage in weight-bearing joints has been described
[13, 14]. This adverse effect may explain why the rate of pneumococcus resis-
tance to fluoroquinolones remains low. A direct correlation has been reported
between the use of the fluoroquinolone antibiotics and prevalence of fluoroqui-
nolone resistance in pneumococcus [15–18]. Table 2.1 describes the principal
mechanisms of resistance to this antibiotic class by pneumococcus; Fig. 2.1
shows the timeline of antimicrobial resistance of pneumococcus.
2 Antimicrobial Resistance Among Streptococcus pneumoniae 15

Table 2.1 Basis of antimicrobial resistance in Streptococcus pneumoniae

Antibiotic Effect Mechanism resistance Risk factors
ß-lactam Inhibit the final steps Alteration of the cell wall Previous antibiotic
of peptidoglycan PBP, resulting in decreased use of ß-lactam
synthesis (cell wall) affinity for penicillin antibiotics in the last
by binding to 3–6 months
high-molecular-­ Prior hospitalization
weight penicillin-­ in the last 3 months
binding proteins Attendance in a
(PBPs) day-care center
Residence in
long-term care
facilities
Chronic pulmonary
disease mainly
chronic obstructive
pulmonary disease
(COPD)
Human
immunodeficiency
virus (HIV) infection
Macrolides Inhibit protein Target site (ribosomal) Previous hospital
alteration by an enzyme that
synthesis by binding admission
methylates 23S rRNA
23S ribosomal target Resistance to
sites in bacteria subunits and is encoded by penicillin
the ermB (erythromycin-­ Previous use of
resistance methylase) gene: macrolides
high level of macrolide
Recurrent otitis media
resistance and complete
cross-resistance to macrolide Cases related to
lincosamide streptogramin B serotypes such
type serotype 6A, 6B, 14,
Active efflux pumps encoded 23F, 19F
by the mefE or mefA Attendance in
(macrolid efflux) gene: day-care centers
low-level of resistance only to
macrolides
Fluoroquinolones Inhibit DNA Spontaneous point mutations Prior use of
synthesis by in the quinolone resistance-­ fluoroquinolones
interacting with determining region (QRDR) Chronic obstructive
intracellular drug pulmonary disease
targets, DNA gyrase, (COPD)
and topoisomerase Residence in a
IV long-term center
Elderly persons
Cerebrovascular
disease
16 C. Cillóniz et al.

Year of introduced antibiotic

Erythromycin 1953

Penicillin Levofloxacin PCV7 PCV13

Cephalosporin introduction
introduction

1943-45 1960 1962 1968 1980 1990 1996 1998 2000 2005 2010
Year of reported resistance

Levofloxacin resistant
Penicillin resistant S. pneumoniae
S. pneumoniae
Cephalosprin resistant
S. pneumoniae
MDR
Erythromycin resistant S. pneumoniae
S. pneumoniae

Abbreviation: MDR (multidrug-resistant); PCV (pneumococcal conjugate vaccine)

Fig. 2.1 Timeline of antibiotic resistance of Streptococcus pneumoniae. Abbreviation: MDR

multidrug-­resistant, PCV pneumococcal conjugate vaccine

2.2.1 Penicillin and ß-Lactam-Resistant Streptococcus

pneumoniae

β-lactam antibiotics include penicillins, cephalosporins, and carbapenems. These

compounds inhibit the final steps of peptidoglycan (cell wall) synthesis by binding
to high-molecular-weight penicillin-binding proteins (PBPs). These antibiotics
have a broad spectrum of activity against Gram-positive and Gram-negative bacte-
ria. β-lactam antibiotics are considered to be time-dependent killers, meaning that
increasing concentration significantly above the minimal inhibitory concentration
(MIC) does not increase killing. The compounds have efficacy when concentrations
are approximately four times the MIC of the microorganism. To determine the effi-
cacy of β-lactam antibiotics, the preferred pharmacodynamic parameter is time (T)
> MIC. For the majority of β-lactams, effectiveness is achieved at T > MIC for more
than 40–50% of the dosing interval [19].
Amino acid alterations of the cell wall PBP result in decreased affinity for peni-
cillin, which is the main mechanism of penicillin resistance. Several PBPs have
been identified, including 1a, 1b, 2x, 2a, 2b, and 3. Alterations to the properties of
PBPs are brought about by transfer of portions of the genes encoding the PBPs from
other streptococcal species, resulting in mosaic genes [20].
The Clinical and Laboratory Standards Institute (CLSI) and the European
Committee on Antimicrobial Susceptibility Testing (EUCAST) define penicillin
resistance of pneumococcus via empirical breakpoint determination [21].
Breakpoints established by the CLSI in 2012 for pneumococci defined penicillin
resistance as:
2 Antimicrobial Resistance Among Streptococcus pneumoniae 17

• Infections other than meningitis: susceptible < 2 μg/ml, intermediate < 4 μg/ml,
and resistant ≥ 8 μg/ml
• Meningitis: susceptible ≤ 0.06 μg/ml, intermediate ≥ 0.12 μg/ml, and resis-
tant ≥ 2 μg/ml
The breakpoints for penicillin susceptibility are based on three criteria: micro-
biological data, pharmacokinetic/pharmacodynamics of β-lactam antibiotics, and
clinical outcome of pneumococcal infections. In a patient treated with a dose of
intravenous penicillin, the levels achieved in the lung will be 100 times greater than
those reached in the brain. Thus, use of low concentrations of β-lactam for pneumo-
coccal infections, such as otitis media or meningitis, could lead to treatment failure.
In contrast, with pulmonary infections the levels of β-lactam reached are generally
sufficient to clear infection. Therefore, treating the same pathogen will require dif-
ferent doses of a given β-lactam depending on the site of infection. Likewise, we
must considerer pneumococcal resistance in different sites of infection differently,
and breakpoints for resistance will be different.

2.2.2 Macrolide Resistance in Streptococcus pneumoniae

Macrolides inhibit bacterial protein synthesis by binding to the 23S rRNA compo-
nent of the 50S ribosomal subunit in bacteria. There are two main mechanisms of
macrolide resistance in pneumococcus. One involves target-site (ribosomal) altera-
tion by an enzyme that methylates 23S rRNA, an enzyme that is encoded by the
ermB (erythromycin-resistance methylase) gene. The resistance phenotype is called
MLSB (macrolide, lincosamide, streptogramin B type) and is responsible for a high
level of macrolide resistance. In a low proportion of cases, ermB gene variation that
modifies the binding site for macrolides and lincosamides confers complete cross-­
resistance to clindamycin [22].
The second mechanism of resistance involves active efflux pumps encoded by
the mefE or mefA (macrolide efflux) genes. These mutations result in low-level
resistance to macrolides but not to the other two agents. The mefA gene is predomi-
nant in Europe, whereas mefB gene predominates in North America.
The relative frequency of the two macrolide resistance mechanisms varies by
geographic region [23–36] (Table 2.2): in European countries, approximately 90%
of the isolates of pneumococcus presented the MLSB phenotype, which is associ-
ated with high levels of macrolide resistance, whereas in North America between
50% and 65% of the resistant pneumococcus isolates contained efflux mutations
that were associated with lower levels of macrolide resistance [12]. In Asian coun-
tries, strains that showed both mechanism of resistance are a major concern, with
between 12% and 40% of the resistant isolates displaying both mechanisms [31,
37–40]. In South American countries, isolates reporting both mechanisms vary
between 4% and 20% [34, 41]. Worldwide resistance to macrolides in pneumococ-
18 C. Cillóniz et al.

Table 2.2 Worldwide genotype distribution of macrolide resistance in Streptococcus pneumoniae

No of isolates % genotype
Country/year of study tested distribution Reference
Europe
Turkey – 2008–2009 80 44% ermB Sirekbasan et al. [23]
11% mefA
44% ermB + mefA
Spain – 1999–2007 187 90% ermB Calatayud et al. [24]
9% mefE
1% mefA
Greece – 2005–2009 1105 (carriers) 29% ermB Grivea et al. [25]
24% ermB + mefE
42% mefE
5% mefA
Belgium – 2007–2009 249 90% ermB Lismond et al. [26]
2% mefE
3% ermB + mefE
North America
Canada– 1998–2004 865 47% mefA Wierzbowsk et al. [27]
43% ermB
6% ermB + mefA
USA – 2007 4535 18% ermB Hawkins et al. [28]
62% mef (A/E) gene
15% mef (A/E) + ermB
Asia
Lebanon − 2010–2015 132 38% ermB El Ashkar et al. [29]
29% mef (A/E)
31% mef (A/E) + ermB
Japan – 2013–2014 960 76% ermB Kawaguchiya et al.
32% mef (A/E) [30]
11% mef (A/E) + ermB
Iran 186 44% ermB Azadegan et al. [31]
16% mef (A/E)
40% mef (A/E) + ermB
South 2184 49% ermB Kim et al. [32]
Korea – 2008–2009 20% mefA
30% mefA + ermB
South America
Colombia – 1994–2011 225 98% ermB Ramos et al. [33]
2% ermB + mef E
Argentina – 2009–2010 126 77% mefA Reijtman et al. [34]
19% ermB
4% mefA + ermB
Africa
Morocco – 2007–2014 655 90% ermB Diawara et al. [35]
6% mef E
35 ermB + mefE
Tunisia – 1998–2004 100 88% ermB Rachdi et al. [36]
12% mefA
2 Antimicrobial Resistance Among Streptococcus pneumoniae 19

cus has increased recently and is associated with the extensive global use of macro-
lides, principally for community-acquired respiratory tract infections.

2.2.3  luoroquinolone Resistance in Streptococcus

F
pneumoniae

Fluoroquinolones inhibit DNA synthesis by forming drug-enzyme-DNA complexes

with DNA gyrase and topoisomerase IV. The main mechanism of resistance to fluo-
roquinolones is mediated by amino acid substitutions in these two essential enzymes.
As with other bacteria, resistant pneumococcus exhibits spontaneous point muta-
tions in a region of GyrA (gyrase) and ParC (topoisomerase IV) called the quino-
lone resistance-determining region (QRDR). Some pneumococci may also exhibit
an efflux-mediated mechanism, although the clinical significance is unclear. The
presence of dual mechanisms of resistance has been reported in strains having high
levels of resistance, often from cases of treatment failure [12, 42, 43]. In some cases
multiple mutations in the target proteins accumulate [44], which supports the idea
that repeated antimicrobial challenge gradually erodes the effectiveness of
fluoroquinolones.

2.2.4 Resistance to Other Antibiotics

Currently, the European Respiratory Society (ERS)/European Society of Clinical

Microbiology and Infectious Diseases (ESCMID) guidelines recommend the use of
tetracyclines (broad-spectrum bacteriostatic antibiotics that act by binding to the
30S ribosomal subunit and thereby inhibit bacterial protein synthesis) as a first
choice for treatment of lower respiratory infections [45]. On the other hand, the
American Thoracic Society (ATS)/Infectious Disease Society of America (IDSA)
[46] recommends doxycycline for healthy patients with pneumococcal community-­
acquired pneumonia with low risk of drug-resistant pneumococcus and for patients
with penicillin allergy [46].
The ribosomal protection protein (RRP), which binds to the ribosome and forces
the drug from its binding site, is the main resistance mechanism of pneumococcus
to tetracycline and doxycycline. This form of resistance is mediated by an alteration
in the tetM gene. In 2012 a study by Dönhöfer et al. showed that TetM can directly
remove and release tetracycline from the bacterial ribosome by an interaction
between domain IV of the 16S rRNA and the tetracycline binding site [47].
Due to the increase in resistance of pneumococcus to several antibiotics over the
last decade and several reported cases of treatment failure, vancomycin, a glycopep-
tide antibiotic that acts by inhibiting proper cell wall synthesis, was added to the
standard antibiotic treatment for pneumococcal meningitis. There are several reports
20 C. Cillóniz et al.

about treatment failure with vancomycin due to the emergence of vancomycin-­

tolerant pneumococcus. However, there is no report of vancomycin-resistant
pneumococcus.
Tolerant pneumococcus survives but does not replicate during therapy with anti-
biotics. When antibiotic therapy is finished, pneumococci are able to resume growth.
This phenomenon is associated with a reduction of autolysin activity, which is part
of an endogenous bacterial cell-death pathway [48].

2.2.5 Multidrug-Resistant (MDR) Streptococcus pneumoniae

It is estimated that the worldwide prevalence of multidrug-resistant (MDR) S. pneu-

moniae is high, ranging from 36% in Asia to 15% in Europe [12, 49], although the
prevalence is geographically variable. Multidrug resistance in pneumococcus is
defined as resistance to three or more antibiotic classes. Pneumococcus MDR gen-
erally involves reduced susceptibility to β-lactams, macrolides, tetracyclines, and
sulfonamides; resistance to quinolones in MDR pneumococcus is less frequent.
The majority of MDR strains of pneumococcus are derived from resistant genetic
clones, with a few clones dominating the pneumococcus isolates on a worldwide
basis [49]. Data from European studies show that the MDR phenotype is most fre-
quent among serotypes 1, 14, 15A, 19A, 19F, and 23F [50]. In the United States and
Canada, however, the most frequent serotypes associated with MDR pneumococcus
are 15A, 15B, 15C, 22F, 23A, 33F, and 35B [51–54]. Studies from Asian countries
report that 11A, 15A, 19A, and 19F are the serotypes most frequently associated
with MDR pneumococcus [53, 55, 56]. In African countries, 19A and 19F are the
most frequently associated with MDR pneumococcus [57]. Collectively these data
indicate that the spread of MDR pneumococcus globally has high variability among
countries. The introduction of conjugate pneumococcal vaccines contributed to the
large reduction of the burden of pneumococcal disease and the reduction of antimi-
crobial resistance in S. pneumoniae. Nevertheless, the emergence of non-vaccine
serotypes that show multidrug resistance is a major concern.

2.3  isk Factors for Infection by Drug-Resistant

R
Pneumococcus

Several studies identify factors associated with an increased risk of infection by

pneumococcus resistant to the most frequently used antibiotics. The three main fac-
tors are host factors (age, comorbidities), environmental factors (geographic regions
with high population density and proximity to high-resistance regions, day-care
centers with children, long-term nursing facilities with elderly persons), and factors
related to the use of antibiotics (previous antibiotic therapy, duration of a­ ntibiotic
therapy).
2 Antimicrobial Resistance Among Streptococcus pneumoniae 21

2.3.1 Risk Factors Related to Penicillin Resistance

The use of a β-lactam antibiotic in the previous 3–6 months is the main risk factor
associated with penicillin-resistant pneumococcal infection [12, 21, 58–60]. A
study by Ruhe et al. [61] regarding the duration of previous antibiotic treatment and
its association with penicillin-resistant bacteremic infection revealed that the risk
depends on the class of prior antibiotic exposure and the duration of therapy. The
study analyzed 303 patients with pneumococcal bacteremia. In 98 (32%) cases of
bacteremia caused by penicillin-non-susceptible S. pneumoniae, statistical analysis
showed that the use of β-lactams, sulfonamides, and macrolides within the last
1–6 months before presentation was associated with penicillin-non-susceptible S.
pneumoniae bacteremia (p < 0.05). In a second study with the same bacteremic
population, Ruhe et al. [62] identified 33 (11%) cases of bacteremia caused by high-­
level resistant S. pneumoniae. In these cases, three risk factors for high-level
penicillin-­resistant pneumococcal infection were identified: β-lactam antibiotic use
in the previous 6 months, previous residence in a risk area (defined as stays in day-­
care facilities, prisons, homeless shelters, nursing homes, or other long-term care
facilities), and respiratory tract infection in the previous year.
Age extremes (<5 years or > 65 years) are a recognized risk factor for penicillin-­
resistant pneumococcal infections [12, 17, 63]. As pointed out above, nasopharyn-
geal carriage of pneumococcus in healthy children ranges from 20% to 50%, and in
the healthy adult population, nasopharyngeal carriage rates range from 5% to 30%
[64, 65]. Consequently, it is not difficult to understand why several studies have
shown that day-care centers are a risk factor for colonization and infection of chil-
dren due to penicillin-resistant pneumococcus [66–68]. Similarly, institutionalized
adults, especially those older than 65 years of age, have increased risk for penicillin-­
resistant pneumococcal infections [69]. Moreover, the presence of specific comor-
bidities, such as human immunodeficiency virus (HIV) and chronic pulmonary
disease, especially chronic obstructive pulmonary disease (COPD), is a recognized
risk factor for penicillin-resistant pneumococcal infection [58].
Several studies have addressed the association between antibiotic consumption
and resistance selection. A study by van Eldere et al. [70], concerning the impact of
antibiotic usage in ambulatory patients in Belgium, involved 14,448 Streptococcus
pneumoniae isolates collected between 1994 and 2004. This work showed a modest
relationship between consumption and resistance; additional factors were high pop-
ulation density and proximity to high-resistance regions, particularly for the devel-
opment of multiple resistances in pneumococcus. In this Belgian population, the
highest levels of resistance were to erythromycin, followed by resistance to tetracy-
cline and penicillin; the highest prevalence of co-resistance to two antibiotics was
for erythromycin-tetracycline.
In 2001 the prevalence of non-susceptibility to erythromycin in the Belgium
study peaked at 36.7% and stayed mostly stable until 2004. Prevalence of non-­
susceptibility to tetracycline reached its highest level (31.7%) in 2000; penicillin
non-susceptibility hit 17.7% in 2000 and declined to 11.6% in 2004. The prevalence
22 C. Cillóniz et al.

of co-resistance to erythromycin-tetracycline was 26.7% in the period 2002–2003

and decreased slightly to 25.9% in 2004.
The overall antibiotic consumption in Belgium was 26.4 DID (daily doses
per 1000 inhabitants per day) in 1995 and decreased slightly to 23.3 DID in
2004. The most frequently consumed antibiotics were broad-spectrum penicil-
lins (9 DID in 2000 to 6.4 DID in 2004). Macrolides showed a similar pattern (6
DID in 2000 to 4.5 DID in 2004) as did cephalosporins (4.7 DID in 2000 to 3.7
DID in 2004). Tetracycline was the second most prescribed class in 1995, but
usage declined in 2004 to 1.9 DID. Overall, consumption and resistance were
roughly parallel.
Another study concerned antimicrobial drug use in ambulatory care and resis-
tance trends in Europe [71] for 21 countries during the period 2000–2005. The work
showed that variation in consumption coincided with the prevalence of resistance at
the country level [71]. Antimicrobial drug use decreased (>15%) in Bulgaria, Czech
Republic, France, and Germany, but it increased (>15%) in Croatia, Denmark,
Greece, and Ireland. The most widely used antibiotics were penicillins (including
broad-spectrum penicillins). Macrolides were the second most widely used cate-
gory; the third consisted of cephalosporins, monobactams, and carbapenems.
Fluoroquinolones occupied the fourth position. Four (France, Luxemburg, Belgium,
and Portugal) of the six countries reporting the highest antimicrobial usage (Greece,
France, Luxembourg, Portugal, Croatia, and Belgium) also reported the highest
resistance proportions.
An interesting, small, case-controlled study about penicillin dust exposure
with pharmaceutical workers in Tehran (Iran) reported that the workers in the
penicillin production line carried a greater percentage of resistant pneumococcus
[72]. The study included 60 cases (workers on a penicillin production line) and
60 controls (workers in food production), and data were obtained via survey, air
sampling, and throat swab. In the penicillin production line arm of the study, the
mean overall concentrations of penicillin dust were 6.6 mg/m3, while it was
4.3 mg/m3 in the food production line (p = 0.001). S. pneumoniae was detected
in 45% (27) individuals in the dust-exposed group, 92.6% of which showed peni-
cillin resistance. In the control group, S. pneumoniae was detected in 35% of the
subjects, while 71.4% of the S. pneumoniae-positive cases were drug resistant
(p = 0.014).

2.3.2 Risk Factors Related to Macrolide Resistance

Recent therapy by macrolides is the main risk factor for macrolide-resistant nasal
colonization and pneumococcal infection [1, 12, 73, 74]. The study by Dias et al.
[75], which evaluated the role of antimicrobial and vaccine use in the trends of
resistance to penicillin and erythromycin in Portugal from 1994 to 2004, found
that the use of macrolides was the main factor associated with an increase of
2 Antimicrobial Resistance Among Streptococcus pneumoniae 23

penicillin and erythromycin non-susceptible isolates among adults (p < 0.01) and
erythromycin non-susceptible isolates among children (p = 0.006). The study also
suggested that the heptavalent vaccine is failing to reduce antimicrobial resis-
tance, possibly due to the increased consumption of azithromycin (p = 0.04).
Other works showed that there is an increased risk of macrolide-resistant infec-
tion in cases related to certain pneumococcus serotypes, in particular 6A, 6B,
11A, 14, 23F, and 19F [76, 77].
Other important risk factors are age below 5 years [78–81], attendance in a day-­
care center [82–84], middle ear infection [85–87], and nosocomial acquisition [26].
As with β-lactams, there is strong evidence correlating the prevalence of macrolide
resistance of pneumococcus and overall macrolide consumption within specific
geographic areas [70, 71, 88].

2.3.3 Risk Factors Related to Fluoroquinolone Resistance

Previous exposure to fluoroquinolones is considered the main risk factor for fluo-
roquinolone resistance [89–92]. Other risk factors, reported worldwide, are
COPD, nosocomial acquisition, and residence in a nursing home [43, 93, 94]. A
retrospective review of cases of invasive pneumococcal infections in adults in
Spain reported that residence in public shelters (OR 26.13, p = 0.002), previous
hospitalization (OR 61.77, p < 0.001), human immunodeficiency virus (HIV)
infection (OR 28.14, p = 0.009), and heavy smoking (OR 14.41, p = 0.016) are
risk factors associated with acquiring an infection by levofloxacin-resistant pneu-
mococci [95–97].

2.3.4 Risk Factors Related to Multidrug-Resistance

The reported risk factors for multidrug-resistant pneumococcal infection are

extremes in age (< 2 years and > 65 years), presence of co-morbidities, such as
chronic heart disease, chronic lung disease, chronic liver disease, chronic renal
disease, prior exposure, especially repeated exposure, to antibiotic therapy in
the previous 3 months, and being an immunosuppressed host [21, 46, 49, 66,
98–100]. Also, infections with pneumococcal serotypes such as 6A/B, 19A,
19F, 15A, 35B, 23A, 22F, and 33F were risk factors. Of these, the strongest risk
factor is repeated exposure to antibiotic therapy. Figure 2.2 provides a sche-
matic explanation for how antibiotic resistance arises and spreads in bacterial
populations. We conclude that the increasing prevalence of multidrug-resistant
strains complicates treatment options for S. pneumoniae infections and in some
cases resistance leads to treatment failure.
24 C. Cillóniz et al.

Fig. 2.2 How antibiotic resistance arises and spreads in bacterial population. (Figure reproduced
by the permission of the author: Laura Piddock and Victoria Wells. Longevity Bulletin:
Antimicrobial Resistance, Chapter 3: How antimicrobial resistance emerges. Issue 8, May 2016)
2 Antimicrobial Resistance Among Streptococcus pneumoniae 25

2.4 Pneumococcal Serotypes and Antibiotic Resistance

There are 98 reported pneumococcal serotypes (capsule type); 92 were identified

using the Quellung method, and the additional serotypes were identified using
molecular techniques [101–103]. These serotypes are grouped into 48 serogroups
based on their antigenic similarities [104]. Several epidemiological studies suggest
that relationships exist between specific serotypes/serogroups and the age of the
host, site of infection, comorbidities, geographic region, pneumococcal invasive-
ness, and disease severity [105–108]. As pointed out above, serotype differences
also relate to antimicrobial resistance. The differing behavior among serotypes may
reflect differences in nasopharyngeal carriage, with the highest rates in children,
especially in the first year of life. As pointed out above, risk factors for nasopharyn-
geal carriage in children include winter season, age below 6 years, presence of
younger siblings, and attendance in day-care centers. In adults, risk factors for naso-
pharyngeal carriage include cigarette smoking, asthma, and acute upper respiratory
infection [64, 109, 110].
Colonization in children may persist for a mean of 4 months, but it is much
shorter in adults, usually 2–4 weeks [16]. This long period of carriage and the fre-
quent exposure to antibiotics by children explain why they are considered the main
source of resistant strains of pneumococcus [99].

2.5 Global Resistance Trends

In recent decades there has been a global acceleration in pneumococcal antibiotic

resistance that coincides with the increased use of antibiotics [2]. The report,
Antibiotic Resistance Threats in the United States, 2013 [111], highlights the impor-
tance of drug-resistant pneumococcus. This report covers bacteria causing severe
human infections and the antibiotics used to treat those infections. The main objec-
tive of this report was to provide an overview of the complex problem of antibiotic
resistance and to encourage immediate action to keep the situation from getting
worse. In this report the CDC prioritized bacteria into one of three categories: urgent
threats, serious threats, and concerning threats. Drug-resistant S. pneumoniae was
considered to be a serious threat. Pathogens in this category require prompt and
sustained action.
Navarro et al. [112], in a 2010 surveillance report on invasive pneumococcal
disease in 26 EU/EEA countries, considered isolates with MIC ≥ 0.12 mg/L as non-­
susceptible to penicillin (this cutoff value is for meningeal isolates and is the most
widely used for surveillance studies). The highest rates of non-susceptibility to
penicillin were found in Romania (42.2%), Cyprus (36.4%), and France (27.5%).
The highest rates of non-susceptibility to cefotaxime were found in Romania
(23.8%) and Ireland (9.3%).
26 C. Cillóniz et al.

The European Antimicrobial Surveillance, published in 2014, showed that of

the 10,456 invasive pneumococcal disease cases reported by 28 EU/EEA coun-
tries, Romania, Spain, and Croatia showed the highest rates of non-susceptibility
to penicillin (47%, 28%, and 26%, respectively, for these countries). The lowest
rates were reported for Cyprus, Belgium, and the Netherlands, at 0%, 1.3%, and
2.1%, respectively [95]. We note that these surveillance data might not be strictly
comparable among all countries, as the clinical breakpoints used to determine
penicillin susceptibility differed, depending on guidelines used and the site of
infection [113]. Nevertheless, the striking differences likely reveal key differ-
ences in antimicrobial use.
Rates of macrolide resistance range widely, from 20% to 90%. This variability
is likely related to geographical differences [114–116]. A US surveillance study
by Jones et al. [114] reported that 56% of isolates (from 19,000 samples ana-
lyzed) showed macrolide resistance. The 2014 European Report of antimicrobial
resistance showed that Romania, Slovakia, and Malta (48%, 41%, and 38%,
respectively) reported the highest rates of non-susceptibility to macrolides; the
lowest rates were reported for Cyprus (0%), Latvia (4.1%), and the Netherlands
(4.3%) [113]. A recent Spanish study of 643 patients with community-acquired
pneumonia found that 22% had macrolide-resistant pneumococcus and 98% of
those showed high-level resistance [117].
The rate of fluoroquinolone resistance of pneumococcus in the United States
and Europe remains low (<1% and <3%, respectively) [12, 113, 118, 119]. One
study, the Antimicrobial Resistance Surveillance in Europe [113], reported resis-
tance data for 30 European countries from the period 2009 to 2012 for 8 bacterial
pathogens as invasive isolates (blood and cerebrospinal fluid). Twenty-four
European countries reported susceptibility data for fluoroquinolones in 6263 iso-
lates (57% of all reported pneumococcus isolates). Among these, 5.2% were
resistant to fluoroquinolones, and 4.4% of the fluoroquinolone-resistant isolates
were also penicillin non-­susceptible. Similarly, an American study by Jones et al.
[114], which was a 14-year longitudinal (1998–2011) survey of S. pneumoniae
that analyzed 18,911 isolates (collected from community-acquired respiratory
tract infections, bacteremias, and pneumonia), reported only 1.2% non-suscepti-
bility to fluoroquinolones (levofloxacin). In contrast, Asian countries reported
higher levels, from 10% to 12%. For example, a study from Hong Kong that
analyzed antimicrobial resistance data for S. pneumoniae from the period 2001–
2007, using samples from respiratory tissue, wounds, blood, and other fluids,
reported that 11% had reduced susceptibility to levofloxacin [32, 120]. Similarly,
a prospective surveillance study of 2184 S. pneumoniae isolates collected from
patients with pneumococcal infections from 60 hospitals in 11 Asian countries
from 2008 to 2009 reported resistance to fluoroquinolones at 1.7%, 0.4%, 1.5%,
and 13.4% for levofloxacin, moxifloxacin, gatifloxacin, and ciprofloxacin,
respectively (Kim et al. [91]). Isolates from Taiwan (6.5%) and South Korea
(4.6%) showed the highest rates of levofloxacin resistance.
2 Antimicrobial Resistance Among Streptococcus pneumoniae 27

2.6 Impact of Vaccines on Resistance

Two types of pneumococcal vaccines are currently available: the polyvalent pneu-
mococcal polysaccharide vaccine (PPV) and the pneumococcal conjugate vaccine
(PCV).
The PPV23 vaccine includes 23 purified capsular polysaccharide antigens of
Streptococcus pneumoniae (1, 2, 3, 4, 5, 6B, 7F, 8, 9N, 9V, 10A, 11A, 12F, 14, 15B,
17F, 18C, 19A, 19F, 20, 22F, 23F, and 33F); it was licensed in the United States in
1983. PPV23 induces antibodies primarily through a T-cell-independent immune
response that enhances phagocytosis, thereby killing the bacterium [121]. The
immune system of young children does not produce an adequate response to the
polysaccharide capsule; consequently, the vaccine is not used in this age group.
The pneumococcal conjugate vaccine 7-valent (PCV7), which included seven
pneumococcal sertotypes (4, 6B, 9V, 14, 18C, 19F, and 23F), was introduced in the
United States in 2000. It is recommended for infants and young children. This vac-
cine is highly effective in preventing invasive disease, with percentages of efficacy
of about 90%. The routine use of PCV7 has resulted not only in a tremendous reduc-
tion in invasive pneumococcal infections in children but also decreased rates of
pneumococcal disease in adults.
Also, after 4 years of the introduction of PCV7 in the United States, the inci-
dence of invasive pneumococcal disease caused by penicillin-non-susceptible S.
pneumoniae and multidrug-resistant S. pneumoniae decreased. In 1999, the rate of
invasive disease caused by penicillin-non-susceptible strains was 6.3 cases per
100,000 – it decreased to 2.7 cases per 100,000 in 2004. Similarly, in 1999 the rate
of cases caused by strains not susceptible to multiple antibiotics was 4.1 cases per
100,000 and decreased to 1.7 cases per 100,000 in 2004 [122]. The study by Whitney
et al. [123] demonstrated that the PCV7 vaccine prevents invasive disease in both
healthy and chronically ill children. Despite the success of PCV7, studies have
noted an increase in the incidence of invasive pneumococcal disease (IPD) caused
by non-vaccine serotypes, such as 1, 3, 5, 6A, 6C, 7F, 12F, 19A, and 22F [124].
These serotypes are related to penicillin-non-susceptible clones. The emergence of
serotype 19A, which correlates with high-level penicillin and multidrug resistance,
is a main concern globally [125]. This serotype presents a dual macrolide-resistance
phenotype (erm B and mefA).
A new 13-valent pneumococcal polysaccharide-protein conjugate vaccine
(PCV13) was approved by the Food and Drug Administration in February 2010 for
the prevention of IPD in infants and young children. PCV13 contains capsular poly-
saccharides from serotypes 1, 3, 4,6A, 7F, 9 V, 14, 18C, 19A, 19F, and 23F. In
March 2010, the Advisory Committee for Immunization Practices (ACIP) recom-
mended that PCV13 replace PCV7 for the vaccination of children. New studies
show a similar reduction in IPD following the introduction of the PCV13 vaccine,
as seen previously with the PCV7 vaccine. The study by Moore et al. [126] analyzed
IPD cases (33,688 cases, of which 89% contained serotyping results) during July
2004–June 2013 and classified as being caused by the PCV13 serotypes against
28 C. Cillóniz et al.

which PCV7 has no effect (PCV13/nonPCV7). The work found a reduction in IPD
in adults associated with PCV13 introduction in children. In all adult age groups,
PCV13/nonPCV7-type IPD (especially serotypes 19A and 7F) declined by 58–72%,
which was comparable to that observed early after PCV7 introduction. The PCV13
led to overall reductions of IPD of 12–32% [126]. However, the phenomenon of
serotype replacement, which is thought to be caused by non-vaccine serotypes
(NVT) that occupy nasopharyngeal natural niches vacated after pneumococcal vac-
cination, is again observed with pneumococcal serotypes 11A, 15A, 23B, and 35B,
the most frequent serotypes. Serotypes 15A and 23B show a high proportion of
penicillin non-susceptibility [127].

2.7 Impact of Antibiotic Resistance on Outcome

The relationship between antibiotic resistance of pneumococci and clinical outcome

is an important consideration for clinicians, because treatment failure related to
antimicrobial resistance is not clear-cut. There are several factors that influence
clinical outcome in pneumococcal infections, such as comorbidities (host factors)
and invasiveness of the pneumococcus serotype (virulence of the microorganism)
that contribute to poor outcome [12].

2.7.1 β-Lactam Resistance and Clinical Implications

The relevance of penicillin-resistant S. pneumoniae to clinical outcome in cases of

pneumococcal community-acquired pneumonia (CAP) is controversial. Several
studies showed that treatment failure in CAP cases does not occur when appropriate
therapy and doses are used, even in those patients infected with non-susceptible
strains and treated with β-lactams. For example, in 2010 a Spanish study analyzed
1041 patients with pneumococcal pneumonia in which 114 (11%) presented septic
shock. The main risk factors were current smoking, chronic corticosteroid therapy,
and serotype 3 pneumococcus. No difference was found regarding genotypes or pat-
terns of antibiotic resistance between patients with or without septic shock [128].
Similarly, a study by Morgandon et al. [129], concerning severe pneumococcal
pneumonia in patients admitted to intensive care units (ICU), reported that risk fac-
tors for mortality were age, male sex, and renal replacement therapy. Comorbidities,
macrolide administration, concomitant bacteremia, or penicillin susceptibility did
not influence outcome in these cases. These studies suggest that the outcome with
community-acquired pneumococcal pneumonia is probably associated with the
clinical presentation of pneumonia rather than the antibiotic resistance of the pneu-
mococcus strain. A plausible explanation is that antibiotic concentrations achieved
in the lung are usually higher than the pneumococcal MIC for more than 40–50% of
the dosing interval, even with resistant strains. It will now be interesting to
2 Antimicrobial Resistance Among Streptococcus pneumoniae 29

determine whether infection by strains having a very high level of resistance to

β-lactams (MIC ≥ 16 μg/ml) correlates with clinical failure with pneumonia
patients.
A different situation is seen with pneumococcal otitis media or meningitis when
treated with a β-lactam – treatment failure is associated with resistant strains. The
speculation is that treatment failure is due to the difficulty in obtaining sufficiently
high antibiotic levels at these sites of infection. For this reason, most guidelines
recommend the use of concomitant vancomycin for patients with pneumococcal
meningitis until the pneumococcal MIC for a β-lactam is known [130].

2.7.2 Macrolide Resistance and Clinical Implications

The high rate of macrolide resistance in pneumococcus is a major concern world-

wide. Reports of treatment failure in cases of otitis media, meningitis, pneumonia,
and bacteremic pneumonia are in the literature [12, 131] for patients who had infec-
tions caused by macrolide-resistant strains. For this reason, monotherapy with mac-
rolides is not recommended as an empirical treatment in any infection caused by
pneumococcus.
Much less information is available for the relationship between macrolide-­
resistant Streptococcus pneumoniae and clinical outcome than with patients treated
with β-lactams. A recent work by Cillóniz et al. [117] concerning the effect of mac-
rolide resistance on the presentation and outcome of 643 patients with CAP reported
that 22% were macrolide resistant. They found no evidence suggesting that patients
hospitalized for macrolide-resistant S. pneumoniae pneumonia were more severely
ill upon presentation or had worse clinical outcomes if they were treated with
guideline-­compliant regimens, including β-lactams, versus noncompliant regimens.
A randomized prospective trial is needed to determine whether there is a relation-
ship between macrolide resistance and poor outcome in patients with severe
community-­acquired pneumonia with whom β-lactam-macrolide combination ther-
apy might improve outcome.

2.7.3 Fluoroquinolone Resistance and Clinical Implications

Treatment failure has been observed with patients treated with fluoroquinolones
who had infections caused by fluoroquinolone-resistant strains [43, 132]. However,
the global rates of fluoroquinolone resistance remain low [32, 99, 114, 118, 133],
making correlation between resistance and outcome statistically marginal. In a 2013
study, Kang et al. [134] evaluated the impact of levofloxacin resistance on 136 adult
patients with invasive pneumococcal disease (IPD). In this work, pneumonia was
the most frequent disease (68%), followed by primary bacteremia (11%) and men-
ingitis (11%). The rate of levofloxacin resistance in invasive pneumococcal isolates
30 C. Cillóniz et al.

was 3.7% (5/136) of the isolates. The overall 30-day mortality rate was 26.5%
(36/136). In univariate analysis, the factors associated with 30-day mortality in
patients with IPD were corticosteroid use, presentation with septic shock, and
development of acute respiratory distress syndrome (ARDS). The authors found an
association between levofloxacin resistance and increased mortality, although sta-
tistical significance was not reached (p = 0.083). However, multivariate analysis
revealed that presentation with septic shock, corticosteroid use, development of
ARDS, and levofloxacin resistance were independent factors associated with 30-day
mortality.
Several worldwide reports about antimicrobial resistance in pneumococcus
noted that in countries where the rates of β-lactam resistance and macrolide resis-
tance are high, the prevalence of fluoroquinolone resistance is also high [70, 114].
It may be that in those situations the consumption of fluoroquinolones is also high.

2.8 Future Considerations

Pneumococcal infections and antimicrobial resistance remain a global health

problem.
Since global antibiotic consumption contributes to the emergence of antibiotic-­
resistant bacteria such as S. pneumoniae, one approach for reducing the problem is
to reduce the need for antibiotics through better public health. Changing social
norms about how and when to use antibiotics is central to preserving antibiotic
effectiveness in all countries. For example we should avoid the use of antibiotics in
agriculture and the food industry. The study by Boeckel el al., concerning global
trends in antimicrobial use in food animals, reported that the demand for meat glob-
ally has led to antibiotic consumption in animals to rise by 70% over the past decade.
The pneumococcus is unusual because vaccines are available. The pediatric
pneumococcal conjugate vaccine has had a striking effect on vaccinated children
and even non-vaccinated children and adults for the pneumococcal serotypes
included in the vaccine. However, non-vaccine serotypes have emerged and are now
associated with high-level antimicrobial resistance. Therefore, continuous surveil-
lance programs are needed to determine optimal empiric treatment for a given local-
ity. Surveillance programs are also needed to control the impact of pneumococcal
campaigns on serotype distribution, emergence of non-vaccine serotypes, and anti-
microbial resistance.
Not all members of an antibiotic class are equally effective against the pneumo-
coccus. Some have a lower MIC than others, and some kill more rapidly. At approved
doses, some reach infected tissues more effectively than others. These properties
need to be carefully defined to guide clinical use. For example, with compounds that
induce mutagenic responses, rapid killing is likely to be important. Additional
insight may emerge from geographical locations that use particular derivatives and
have very high rates of resistance. A clear example of this is the resistance of pneu-
mococcus to macrolides. In Europe the main resistance mechanism is the ribosomal
2 Antimicrobial Resistance Among Streptococcus pneumoniae 31

mutation that confers high resistance to macrolides, whereas in the United States,
the dominant mechanism of resistance is active efflux, which confers low levels of
resistance to macrolides. These data suggest the importance of clinical studies in
different geographical areas before recommending particular antibiotics. A com-
pletely different question is how to slow transmission among young children and
elderly persons in long-term care facilities. Solutions may involve reducing antimi-
crobial consumption, the main driver of newly acquired resistance.
Continued surveillance to quantify pneumococcal resistance is also needed to
detect the emergence of new strains exhibiting high-level resistance to penicillin.
Moreover, we need to better understand the clinical relevance and impact of antibi-
otic resistance on pneumococcal infections, since there is not always a clear rela-
tionship between resistance and treatment failure.

Major Points
• Streptococcus pneumoniae remains an important pathogen worldwide.
Pneumococcal infections are related to high rates of morbidity and mortality
especially in young children, older adults, and immunocompromised persons.
• Worldwide pneumococcal infections remain a big challenge for physicians
because of its resistance to penicillin and increasing resistance to macrolides.
• Efforts to reduce antibiotic consumption should be encouraged by educational
programs and guidelines for healthcare professionals.
• The best way to prevent pneumococcal infection is by the implementation of
conjugate pneumococcal vaccinations.
• It is important to monitor the evolution of pneumococcal disease, focusing on
serotype replacement.
• Studies focusing on the development of new vaccine designs should be addressed
in order to avoid serotype replacement.

References

1. Goossens H. Antibiotic consumption and link to resistance. Clin Microbiol Infect.

2009;15(Suppl 3):12–5.
2. Bell BG, Schellevis F, Stobberingh E, Goossens H, Pringle M. A systematic review and meta-­
analysis of the effects of antibiotic consumption on antibiotic resistance. BMC Infect Dis.
2014;14:13.
3. Frieden T, Center for Disease Control. Antibiotic Resistance Threats in the United States
2013. US: U.S. Department of Health and Human Service, Center for Disease Control and
Prevention; 2013.
4. Solomon SL, Oliver KB. Antibiotic resistance threats in the United States: stepping back
from the brink. Am Fam Physician. 2014;89(12):938–41.
5. Anon. The bacterial challenge: time to react. EMEA-ECDC Technical Report 2009. 2009.
6. Adriaenssens N, Coenen S, Versporten A, Muller A, Vankerckhoven V, Goossens H. European
Surveillance of Antimicrobial Consumption (ESAC): quality appraisal of antibiotic use in
Europe. J Antimicrob Chemother. 2011;66(Suppl 6):vi71–7.
32 C. Cillóniz et al.

7. World Health Organization (WHO). Antimicrobial resistance: global report on surveil-

lance 2014, Antimicrobial resistance global surveillance report. Geneva: World Health
Organization; 2014.
8. Weisfelt M, van de Beek D, Spanjaard L, Reitsma JB, de Gans J. Clinical features, complica-
tions, and outcome in adults with pneumococcal meningitis: a prospective case series. Lancet
Neurol. 2006;5(2):123–9.
9. van de Beek D, de Gans J, Tunkel AR, Wijdicks EF. Community-acquired bacterial meningi-
tis in adults. N Engl J Med. 2006;354(1):44–53.
10. Van Boeckel TP, Gandra S, Ashok A, et al. Global antibiotic consumption 2000 to 2010: an
analysis of national pharmaceutical sales data. Lancet Infect Dis. 2014;14(8):742–50.
11. Linares J, Ardanuy C, Pallares R, Fenoll A. Changes in antimicrobial resistance, serotypes
and genotypes in Streptococcus pneumoniae over a 30-year period. Clin Microbiol Infect.
2010;16(5):402–10.
12. Lynch JP III, Zhanel GG. Streptococcus pneumoniae: does antimicrobial resistance matter?
Semin Respir Crit Care Med. 2009;30(2):210–38.
13. Choi SH, Kim EY, Kim YJ. Systemic use of fluoroquinolone in children. Korean J Pediatr.
2013;56(5):196–201.
14. Bradley JS, Jackson MA. The use of systemic and topical fluoroquinolones, e1034. Pediatrics.
2011;128(4):–e1045.
15. Arason VA, Kristinsson KG, Sigurdsson JA, Stefansdottir G, Molstad S, Gudmundsson S. Do
antimicrobials increase the carriage rate of penicillin resistant pneumococci in children?
Cross sectional prevalence study. BMJ. 1996;313(7054):387–91.
16. Dagan R. Impact of pneumococcal conjugate vaccine on infections caused by antibiotic-­
resistant Streptococcus pneumoniae. Clin Microbiol Infect. 2009;15(Suppl 3):16–20.
17. Hogberg L, Geli P, Ringberg H, Melander E, Lipsitch M, Ekdahl K. Age- and serogroup-­
related differences in observed durations of nasopharyngeal carriage of penicillin-resistant
pneumococci. J Clin Microbiol. 2007;45(3):948–52.
18. Fenoll A, Granizo JJ, Aguilar L, et al. Temporal trends of invasive Streptococcus pneumoniae
serotypes and antimicrobial resistance patterns in Spain from 1979 to 2007. J Clin Microbiol.
2009;47(4):1012–20.
19. Owens RC Jr, Shorr AF. Rational dosing of antimicrobial agents: pharmacokinetic and phar-
macodynamic strategies. Am J Health Syst Pharm. 2009;66(12 Suppl 4):S23–30.
20. Jensen A, Valdorsson O, Frimodt-Moller N, Hollingshead S, Kilian M. Commensal strep-
tococci serve as a reservoir for beta-lactam resistance genes in Streptococcus pneumoniae.
Antimicrob Agents Chemother. 2015;59(6):3529–40.
21. Cilloniz C, Ardanuy C, Vila J, Torres A. What is the clinical relevance of drug-resistant pneu-
mococcus? Curr Opin Pulm Med. 2016;22(3):227–34.
22. Shortridge VD, Doern GV, Brueggemann AB, Beyer JM, Flamm RK. Prevalence of macro-
lide resistance mechanisms in Streptococcus pneumoniae isolates from a multicenter antibi-
otic resistance surveillance study conducted in the United States in 1994–1995. Clin Infect
Dis. 1999;29(5):1186–8.
23. Sirekbasan L, Gonullu N, Sirekbasan S, Kuskucu M, Midilli K. Phenotypes and genotypes of
macrolide-resistant streptococcus pneumoniae. Balkan Med J. 2015;32(1):84–8.
24. Calatayud L, Ardanuy C, Tubau F, et al. Serotype and genotype replacement among
macrolide-­resistant invasive Pneumococci in adults: mechanisms of resistance and associa-
tion with different transposons. J Clin Microbiol. 2010;48(4):1310–6.
25. Grivea IN, Sourla A, Ntokou E, Chryssanthopoulou DC, Tsantouli AG, Syrogiannopoulos
GA. Macrolide resistance determinants among Streptococcus pneumoniae isolates from car-
riers in Central Greece. BMC Infect Dis. 2012;12:255.
26. Lismond A, Carbonnelle S, Verhaegen J, et al. Antimicrobial susceptibility of Streptococcus
pneumoniae isolates from vaccinated and non-vaccinated patients with a clinically con-
firmed diagnosis of community-acquired pneumonia in Belgium. Int J Antimicrob Agents.
2012;39(3):208–16.
2 Antimicrobial Resistance Among Streptococcus pneumoniae 33

27. Wierzbowski AK, Nichol K, Laing N, et al. Macrolide resistance mechanisms among
Streptococcus pneumoniae isolated over 6 years of Canadian Respiratory Organism
Susceptibility Study (CROSS) (1998 2004). J Antimicrob Chemother. 2007;60(4):733–40.
28. Hawkins PA, Chochua S, Jackson D, Beall B, McGee L. Mobile elements and chromosomal
changes associated with MLS resistance phenotypes of invasive pneumococci recovered in
the United States. Microb Drug Resist. 2015;21(2):121–9.
29. El Ashkar S, Osman M, Rafei R, et al. Molecular detection of genes responsible for macro-
lide resistance among Streptococcus pneumoniae isolated in North Lebanon. J Infect Public
Health. 2017;10(6):745–8.
30. Kawaguchiya M, Urushibara N, Aung MS, et al. Emerging non-PCV13 serotypes of nonin-
vasive Streptococcus pneumoniae with macrolide resistance genes in northern Japan. New
Microbes New Infect. 2016;9:66–72.
31. Azadegan A, Ahmadi A, Lari AR, Talebi M. Detection of the efflux-mediated erythromycin
resistance transposon in Streptococcus pneumoniae. Ann Lab Med. 2015;35(1):57–61.
32. Kim SH, Song JH, Chung DR, et al. Changing trends in antimicrobial resistance and sero-
types of Streptococcus pneumoniae isolates in Asian countries: an Asian Network for
Surveillance of Resistant Pathogens (ANSORP) study. Antimicrob Agents Chemother.
2012;56(3):1418–26.
33. Ramos V, Duarte C, Diaz A, Moreno J. Mobile genetic elements associated with
erythromycin-­ resistant isolates of Streptococcus pneumoniae in Colombia. Biomedica.
2014;34(Suppl 1):209–16.
34. Reijtman V, Gagetti P, Faccone D, et al. Macrolide resistance in Streptococcus pneumoniae
isolated from Argentinian pediatric patients suffering from acute otitis media. Rev Argent
Microbiol. 2013;45(4):262–6.
35. Diawara I, Zerouali K, Katfy K, et al. Phenotypic and genotypic characterization of
Streptococcus pneumoniae resistant to macrolide in Casablanca, Morocco. Infect Genet Evol.
2016;40:200–4.
36. Rachdi M, Boutiba-Ben Boubaker I, Moalla S, et al. Phenotypic and genotypic character-
ization of macrolide resistant Streptococcus pneumoniae in Tunisia. Pathol Biol (Paris).
2008;56(3):125–9.
37. Reinert RR, Filimonova OY, Al-Lahham A, et al. Mechanisms of macrolide resistance
among Streptococcus pneumoniae isolates from Russia. Antimicrob Agents Chemother.
2008;52(6):2260–2.
38. Ma X, Yao KH, Xie GL, et al. Characterization of erythromycin-resistant Streptococcus
pneumoniae isolates causing invasive diseases in Chinese children. Chin Med J (Engl).
2013;126(8):1522–7.
39. Zhou L, Yu SJ, Gao W, Yao KH, Shen AD, Yang YH. Serotype distribution and antibiotic
resistance of 140 pneumococcal isolates from pediatric patients with upper respiratory infec-
tions in Beijing, 2010. Vaccine. 2011;29(44):7704–10.
40. Geng Q, Zhang T, Ding Y, et al. Molecular characterization and antimicrobial susceptibility
of Streptococcus pneumoniae isolated from children hospitalized with respiratory infections
in Suzhou, China. PLoS One. 2014;9(4):e93752.
41. Caierao J, Hawkins P, Sant’anna FH, et al. Serotypes and genotypes of invasive Streptococcus
pneumoniae before and after PCV10 implementation in southern Brazil. PLoS One.
2014;9(10):e111129.
42. Fuller JD, McGeer A, Low DE. Drug-resistant pneumococcal pneumonia: clinical relevance
and approach to management. Eur J Clin Microbiol Infect Dis. 2005;24(12):780–8.
43. Fuller JD, Low DE. A review of Streptococcus pneumoniae infection treatment failures asso-
ciated with fluoroquinolone resistance. Clin Infect Dis. 2005;41(1):118–21.
44. Urban C, Rahman N, Zhao X, et al. Fluoroquinolone-resistant Streptococcus pneumoniae
associated with levofloxacin therapy. J Infect Dis. 2001;184(6):794–8.
45. Woodhead M, Blasi F, Ewig S, et al. Guidelines for the management of adult lower respira-
tory tract infections--full version. Clin Microbiol Infect. 2011;17(Suppl 6):E1–59.
34 C. Cillóniz et al.

46. Mandell LA, Wunderink RG, Anzueto A, et al. Infectious Diseases Society of America/
American Thoracic Society Consensus guidelines on the management of community-­
acquired pneumonia in adults. Clin Infect Dis. 2007;44(Suppl 2):S27–72.
47. Donhofer A, Franckenberg S, Wickles S, Berninghausen O, Beckmann R, Wilson
DN. Structural basis for TetM-mediated tetracycline resistance. Proc Natl Acad Sci U S A.
2012;109(42):16900–5.
48. Moscoso M, Domenech M, Garcia E. Vancomycin tolerance in Gram-positive cocci. Environ
Microbiol Rep. 2011;3(6):640–50.
49. Van BF, Reinert RR, Appelbaum PC, Tulkens PM, Peetermans WE. Multidrug-resistant
Streptococcus pneumoniae infections: current and future therapeutic options. Drugs.
2007;67(16):2355–82.
50. Sheppard C, Fry NK, Mushtaq S, et al. Rise of multidrug-resistant non-vaccine serotype
15A Streptococcus pneumoniae in the United Kingdom, 2001 to 2014. Euro Surveill.
2016;21(50):3042.
51. Richter SS, Diekema DJ, Heilmann KP, Dohrn CL, Riahi F, Doern GV. Changes in pneumo-
coccal serotypes and antimicrobial resistance after introduction of the 13-valent conjugate
vaccine in the United States. Antimicrob Agents Chemother. 2014;58(11):6484–9.
52. Golden AR, Rosenthal M, Fultz B, et al. Characterization of MDR and XDR Streptococcus
pneumoniae in Canada, 2007–13. J Antimicrob Chemother. 2015;70(8):2199–202.
53. Golden AR, Adam HJ, Zhanel GG. Invasive Streptococcus pneumoniae in Canada, 2011–
2014: Characterization of new candidate 15-valent pneumococcal conjugate vaccine sero-
types 22F and 33F. Vaccine. 2016;34(23):2527–30.
54. Olarte L, Kaplan SL, Barson WJ, et al. Emergence of Multidrug-resistant Pneumococcal
Serotype 35B among U.S. Children. J Clin Microbiol. 2017;55(3):724–34.
55. Nakano S, Fujisawa T, Ito Y, et al. Serotypes, antimicrobial susceptibility, and molecular
epidemiology of invasive and non-invasive Streptococcus pneumoniae isolates in paediatric
patients after the introduction of 13-valent conjugate vaccine in a nationwide surveillance
study conducted in Japan in 2012–2014. Vaccine. 2016;34(1):67–76.
56. Elshafie S, Taj-Aldeen SJ. Emerging resistant serotypes of invasive Streptococcus pneu-
moniae. Infect Drug Resist. 2016;9:153–60.
57. Bahy RH, Hamouda HM, Shahat AS, Yassin AS, Amin MA. Emergence of neoteric sero-
types among multidrug resistant strains of Streptococcus pneumoniae prevalent in Egypt.
Jundishapur J Microbiol. 2016;9(4):e30708.
58. Aspa J, Rajas O, Rodríguez de Castro F, et al. Drug-resistant pneumococcal pneumonia:
clinical relevance and related factors. Clin Infect Dis. 2004;38(6):787–98.
59. Pallares R, Gudiol F, Linares J, et al. Risk factors and response to antibiotic therapy in adults
with bacteremic pneumonia caused by penicillin-resistant pneumococci. N Engl J Med.
1987;317:18–22.
60. Tsai HY, Lauderdale TL, Wang JT, et al. Updated antibiotic resistance and clinical spectrum
of infections caused by Streptococcus pneumoniae in Taiwan: emphasis on risk factors for
penicillin nonsusceptibilities. J Microbiol Immunol Infect. 2013;46(5):345–51.
61. Ruhe JJ, Hasbun R. Streptococcus pneumoniae bacteremia: duration of previous antibiotic
use and association with penicillin resistance. Clin Infect Dis. 2003;36(9):1132–8.
62. Ruhe JJ, Myers L, Mushatt D, Hasbun R. High-level penicillin-nonsusceptible Streptococcus
pneumoniae bacteremia: identification of a low-risk subgroup. Clin Infect Dis.
2004;38(4):508–14.
63. Mollendorf CV, Cohen C, De Gouveia L, et al. Factors associated with ceftriaxone non-
susceptibility of Streptococcus pneumoniae: analysis of South African national surveillance
data, 2003 to 2010. Antimicrob Agents Chemother. 2014;58(6):3293–305.
64. Bogaert D, de Groot R, Hermans PW. Streptococcus pneumoniae colonisation: the key to
pneumococcal disease. Lancet Infect Dis. 2004;4(3):144–54.
65. Gritzfeld JF, Wright AD, Collins AM, et al. Experimental human pneumococcal carriage.
J Vis Exp. 2013;(72):50115.
2 Antimicrobial Resistance Among Streptococcus pneumoniae 35

66. Vasoo S, Singh K, Hsu LY, et al. Increasing antibiotic resistance in Streptococcus pneumoniae
colonizing children attending day-care centres in Singapore. Respirology. 2011;16(8):1241–8.
67. Jain A, Kumar P, Awasthi S. High nasopharyngeal carriage of drug resistant Streptococcus
pneumoniae and Haemophilus influenzae in North Indian schoolchildren. Trop Med Int Heal.
2015;10:234–9.
68. Dunais B, Bruno-Bazureault P, Carsenti-Dellamonica H, Touboul P, Pradier C. A decade-­
long surveillance of nasopharyngeal colonisation with Streptococcus pneumoniae among
children attending day-care centres in south-eastern France: 1999–2008. Eur J Clin Microbiol
Infect Dis. 2011;30(7):837–43.
69. Nuorti JP, Butler JC, Crutcher JM, et al. An outbreak of multidrug-resistant pneumococ-
cal pneumonia and bacteremia among unvaccinated nursing home residents. N Engl J Med.
1998;338:1861–8.
70. Van Eldere J, Mera RM, Miller LA, Poupard JA, Amrine-Madsen H. Risk factors for devel-
opment of multiple-class resistance to Streptococcus pneumoniae Strains in Belgium over
a 10-year period: antimicrobial consumption, population density, and geographic location.
Antimicrob Agents Chemother. 2007;51(10):3491–7.
71. van de Sande-Bruinsma N, Grundmann H, Verloo D, et al. Antimicrobial drug use and resis-
tance in Europe. Emerg Infect Dis. 2008;14(11):1722–30.
72. Farshad AA, Enferadi M, Bakand S, Jamshidi OR, Mirkazemi R. Penicillin dust exposure
and penicillin resistance among pharmaceutical workers in Tehran, Iran. Int J Occup Environ
Health. 2016:22(3): 218–23.
73. Hare KM, Leach AJ, Morris PS, et al. Impact of recent antibiotics on nasopharyngeal car-
riage and lower airway infection in Indigenous Australian children with non-cystic fibrosis
bronchiectasis. Int J Antimicrob Agents. 2012;40(4):365–9.
74. Metlay JP, Fishman NO, Joffe MM, Kallan MJ, Chittams JL, Edelstein PH. Macrolide
resistance in adults with bacteremic pneumococcal pneumonia. Emerg Infect Dis.
2006;12(8):1223–30.
75. Dias R, Canica M. Trends in resistance to penicillin and erythromycin of invasive pneumo-
cocci in Portugal. Epidemiol Infect. 2008;136(7):928–39.
76. Baek JY, Kim SH, Kang CI, et al. Prevalence of antimicrobial resistant Streptococcus pneu-
moniae serotype 11A isolates in Korea, during 2004–2013, due to the increase of multidrug-­
resistant clone, CC166. Infect Genet Evol. 2016;38:122–5.
77. Zhou L, Ma X, Gao W, et al. Molecular characteristics of erythromycin-resistant
Streptococcus pneumoniae from pediatric patients younger than five years in Beijing, 2010.
BMC Microbiol. 2012;12:228.
78. Pan F, Han L, Huang W, et al. Serotype distribution, antimicrobial susceptibility, and molecu-
lar epidemiology of Streptococcus pneumoniae Isolated from children in Shanghai, China.
PLoS One. 2015;10(11):e0142892.
79. Hadjipanayis A, Efstathiou E, Alexandrou M, et al. Nasopharyngeal Pneumococcal Carriage
among Healthy Children in Cyprus Post Widespread Simultaneous Implementation of
PCV10 and PCV13 Vaccines. PLoS One. 2016;11(10):e0163269.
80. Vanderkooi OG, McConnell A, Church DL, Kellner JD. Antimicrobial susceptibility of inva-
sive and lower respiratory tract isolates of Streptococcus pneumoniae, 1998 to 2007. Can
J Infect Dis Med Microbiol. 2009;20(4):e139–44.
81. Matsumoto A, Hosoya M, Kawasaki Y, Katayose M, Kato K, Suzuki H. The emergence of
drug-resistant Streptococcus pneumoniae and host risk factors for carriage of drug-resistant
genes in northeastern Japan. Jpn J Infect Dis. 2007;60(1):10–3.
82. Grivea IN, Priftis KN, Giotas A, et al. Dynamics of pneumococcal carriage among day-care
center attendees during the transition from the 7-valent to the higher-valent pneumococcal
conjugate vaccines in Greece. Vaccine. 2014;32(48):6513–20.
83. Zuccotti G, Mameli C, Daprai L, et al. Serotype distribution and antimicrobial susceptibili-
ties of nasopharyngeal isolates of Streptococcus pneumoniae from healthy children in the
13-valent pneumococcal conjugate vaccine era. Vaccine. 2014;32(5):527–34.
36 C. Cillóniz et al.

84. Korona-Glowniak I, Siwiec R, Malm A. Resistance determinants and their association with
different transposons in the antibiotic-resistant Streptococcus pneumoniae. Biomed Res Int.
2015;2015:836496.
85. Ding Y, Geng Q, Tao Y, et al. Etiology and epidemiology of children with acute otitis media
and spontaneous otorrhea in Suzhou, China. Pediatr Infect Dis J. 2015;34(5):e102–6.
86. Abdelnour A, Arguedas A, Dagan R, et al. Etiology and antimicrobial susceptibility of
middle ear fluid pathogens in Costa Rican children with otitis media before and after the
introduction of the 7-valent pneumococcal conjugate vaccine in the National Immunization
Program: acute otitis media microbiology in Costa Rican children. Medicine (Baltimore).
2015;94(2):e320.
87. Pumarola F, Mares J, Losada I, et al. Microbiology of bacteria causing recurrent acute oti-
tis media (AOM) and AOM treatment failure in young children in Spain: shifting patho-
gens in the post-pneumococcal conjugate vaccination era. Int J Pediatr Otorhinolaryngol.
2013;77(8):1231–6.
88. Arason VA, Sigurdsson JA, Erlendsdottir H, Gudmundsson S, Kristinsson KG. The role of
antimicrobial use in the epidemiology of resistant pneumococci: a 10-year follow up. Microb
Drug Resist. 2006;12(3):169–76.
89. Ho PL, Tse WS, Tsang KW, et al. Risk factors for acquisition of levofloxacin-resistant
Streptococcus pneumoniae: a case-control study. Clin Infect Dis. 2001;32(5):701–7.
90. Kang CI, Song JH, Kim SH, et al. Risk factors for levofloxacin-nonsusceptible Streptococcus
pneumoniae in community-acquired pneumococcal pneumonia: a nested case-control study.
Eur J Clin Microbiol Infect Dis. 2014;33(1):55–9.
91. Kim ES, Hooper DC. Clinical importance and epidemiology of quinolone resistance. Infect
Chemother. 2014;46(4):226–38.
92. Kuster SP, Rudnick W, Shigayeva A, et al. Previous antibiotic exposure and antimicrobial
resistance in invasive pneumococcal disease: results from prospective surveillance. Clin
Infect Dis. 2014;59(7):944–52.
93. Wolter N, Du Plessis M, Von Gottberg A, De Gouveia L, Klugman KP. Molecular charac-
terization of emerging non-levofloxacin-susceptible pneumococci isolated from children in
South Africa. J Clin Microbiol. 2009;47(5):1319–24.
94. Ben-David D, Schwaber MJ, Adler A, et al. Persistence and complex evolution
of fluoroquinolone-­ resistant Streptococcus pneumoniae clone. Emerg Infect Dis.
2014;20(5):799–805.
95. Isea-Pena MC, Sanz-Moreno JC, Esteban J, Fernandez-Roblas R, Fernandez-Guerrero
ML. Risk factors and clinical significance of invasive infections caused by levofloxacin-­
resistant Streptococcus pneumoniae. Infection. 2013;41(5):935–9.
96. Kupronis BA, Richards CL, Whitney CG. Invasive pneumococcal disease in older adults resid-
ing in long-term care facilities and in the community. J Am Geriatr Soc. 2003;51(11):1520–5.
97. Vanderkooi OG, Low DE, Green K, Powis JE, McGeer A. Predicting antimicrobial resistance
in invasive pneumococcal infections. Clin Infect Dis. 2005;40(9):1288–97.
98. Thabit AK, Crandon JL, Nicolau DP. Antimicrobial resistance: impact on clinical and
economic outcomes and the need for new antimicrobials. Expert Opin Pharmacother.
2015;16(2):159–77.
99. Song JH, Dagan R, Klugman KP, Fritzell B. The relationship between pneumococcal sero-
types and antibiotic resistance. Vaccine. 2012;30(17):2728–37.
100. Thummeepak R, Leerach N, Kunthalert D, Tangchaisuriya U, Thanwisai A, Sitthisak S. High
prevalence of multi-drug resistant Streptococcus pneumoniae among healthy children in
Thailand. J Infect Public Health. 2015;8(3):274–81.
101. Jin P, Kong F, Xiao M, et al. First report of putative Streptococcus pneumoniae serotype 6D
among nasopharyngeal isolates from Fijian children. J Infect Dis. 2009;200(9):1375–80.
102. Oliver MB, van der Linden MP, Kuntzel SA, Saad JS, Nahm MH. Discovery of Streptococcus
pneumoniae serotype 6 variants with glycosyltransferases synthesizing two differing repeat-
ing units. J Biol Chem. 2015;290(44):26474–5.
2 Antimicrobial Resistance Among Streptococcus pneumoniae 37

103. Calix JJ, Porambo RJ, Brady AM, et al. Biochemical, genetic, and serological characteriza-
tion of two capsule subtypes among Streptococcus pneumoniae Serotype 20 strains: discov-
ery of a new pneumococcal serotype. J Biol Chem. 2012;287(33):27885–94.
104. Jauneikaite E, Tocheva AS, Jefferies JM, et al. Current methods for capsular typing of
Streptococcus pneumoniae. J Microbiol Methods. 2015;113:41–9.
105. Brueggemann AB, Peto TE, Crook DW, Butler JC, Kristinsson KG, Spratt BG. Temporal
and geographic stability of the serogroup-specific invasive disease potential of Streptococcus
pneumoniae in children. J Infect Dis. 2004;190(7):1203–11.
106. Browall S, Backhaus E, Naucler P, et al. Clinical manifestations of invasive pneumococcal
disease by vaccine and non-vaccine types. Eur Respir J. 2014;44(6):1646–57.
107. Sjostrom K, Spindler C, Ortqvist A, et al. Clonal and capsular types decide whether pneumo-
cocci will act as a primary or opportunistic pathogen. Clin Infect Dis. 2006;42(4):451–9.
108. Henriques-Normark B, Blomberg C, Dagerhamn J, Battig P, Normark S. The rise and fall of
bacterial clones: Streptococcus pneumoniae. Nat Rev Microbiol. 2008;6(11):827–37.
109. Meropol SB, Stange KC, Jacobs MR, Weiss JK, Bajaksouzian S, Bonomo RA. Bacterial
colonization and antibiotic resistance in a prospective cohort of newborn infants during the
first year of life. Open Forum Infect Dis. 2016;3(4):ofw221.
110. Boken DJ, Chartrand SA, Goering RV, Kruger R, Harrison CJ. Colonization with
penicillin-­resistant Streptococcus pneumoniae in a child-care center. Pediatr Infect Dis
J. 1995;14(10):879–84.
111. Centers for Disease Control and Prevention (CDC). Antibiotic Resistance Threats in the
United States. 2013.
112. Navarro TA, Dias JG, Quinten C, et al. European enhanced surveillance of invasive pneumo-
coccal disease in 2010: data from 26 European countries in the post-heptavalent conjugate
vaccine era. Vaccine. 2014;32(29):3644–50.
113. European Centre for Disease Prevention and Control (ECDC). Antimicrobial resistance sur-
veillance in Europe 2012. 2012.
114. Jones RN, Sader HS, Mendes RE, Flamm RK. Update on antimicrobial susceptibility trends
among Streptococcus pneumoniae in the United States: report of ceftaroline activity from
the SENTRY Antimicrobial Surveillance Program (1998–2011). Diagn Microbiol Infect Dis.
2013;75(1):107–9.
115. Jenkins SG, Farrell DJ. Increase in pneumococcus macrolide resistance, United States.
Emerg Infect Dis. 2009;15(8):1260–4.
116. Xiao Y, Wei Z, Shen P, et al. Bacterial-resistance among outpatients of county hospitals
in China: significant geographic distinctions and minor differences between central cities.
Microbes Infect. 2015;17(6):417–25.
117. Cilloniz C, Albert RK, Liapikou A, et al. The effect of macrolide-resistance on the presenta-
tion and outcome of patients hospitalized for Streptococcus pneumoniae pneumonia. Am
J Respir Crit Care Med. 2015;191(11):1265–72.
118. Pletz MW, van der Linden M, von Baum H, Duesberg CB, Klugman KP, Welte T. Low preva-
lence of fluoroquinolone resistant strains and resistance precursor strains in Streptococcus
pneumoniae from patients with community-acquired pneumonia despite high fluoroquino-
lone usage. Int J Med Microbiol. 2011;301(1):53–7.
119. Jacobs E, Dalhoff A, Korfmann G. Susceptibility patterns of bacterial isolates from hospital-
ised patients with respiratory tract infections (MOXIAKTIV Study). Int J Antimicrob Agents.
2009;33(1):52–7.
120. Ho PL, Cheng VC, Chow KH. Decreasing prevalence of levofloxacin-resistant Streptococcus
pneumoniae in Hong Kong, 2001 to 2007. J Antimicrob Chemother. 2009;63(4):836–8.
121. Vila-Corcoles A, Ochoa-Gondar O. Preventing pneumococcal disease in the elderly: recent
advances in vaccines and implications for clinical practice. Drugs Aging. 2013;30(5):263–76.
122. Kyaw MH, Lynfield R, Schaffner W, et al. Effect of introduction of the pneumococcal conjugate
vaccine on drug-resistant Streptococcus pneumoniae. N Engl J Med. 2006;354(14):1455–63.
38 C. Cillóniz et al.

123. Whitney CG, Pilishvili T, Farley MM, et al. Effectiveness of seven-valent pneumococcal con-
jugate vaccine against invasive pneumococcal disease: a matched case-control study. Lancet.
2006;368(9546):1495–502.
124. Gertz RE Jr, Li Z, Pimenta FC, et al. Increased penicillin nonsusceptibility of nonvaccine-­
serotype invasive pneumococci other than serotypes 19A and 6A in post-7-valent conjugate
vaccine era. J Infect Dis. 2010;201(5):770–5.
125. Reinert R, Jacobs MR, Kaplan SL. Pneumococcal disease caused by serotype 19A: review of
the literature and implications for future vaccine development. Vaccine. 2010;28(26):4249–59.
126. Moore MR, Link-Gelles R, Schaffner W, et al. Effect of use of 13-valent pneumococcal conju-
gate vaccine in children on invasive pneumococcal disease in children and adults in the USA:
analysis of multisite, population-based surveillance. Lancet Infect Dis. 2015;15(3):301–9.
127. van der Linden M, Perniciaro S, Imohl M. Increase of serotypes 15A and 23B in IPD in
Germany in the PCV13 vaccination era. BMC Infect Dis. 2015;15:207.
128. Garcia-Vidal C, Ardanuy C, Tubau F, et al. Pneumococcal pneumonia presenting with septic
shock: host- and pathogen-related factors and outcomes. Thorax. 2010;65(1):77–81.
129. Mongardon N, Max A, Bougle A, et al. Epidemiology and outcome of severe pneumococcal
pneumonia admitted to intensive care unit: a multicenter study. Crit Care. 2012;16(4):R155.
130. Kaplan SL, Mason EO Jr. Management of infections due to antibiotic-resistant Streptococcus
pneumoniae. Clin Microbiol Rev. 1998;11(4):628–44.
131. Schroeder MR, Stephens DS. Macrolide resistance in Streptococcus pneumoniae. Front Cell
Infect Microbiol. 2016;6:98.
132. Endimiani A, Brigante G, Bettaccini AA, Luzzaro F, Grossi P, Toniolo AQ. Failure of levo-
floxacin treatment in community-acquired pneumococcal pneumonia. BMC Infect Dis.
2005;5:106.
133. Kim L, McGee L, Tomczyk S, Beall B. Biological and epidemiological features of antibiotic-­
resistant Streptococcus pneumoniae in pre- and post-conjugate vaccine eras: a United States
perspective. Clin Microbiol Rev. 2016;29(3):525–52.
134. Kang CI, Song JH, Kim SH, et al. Association of levofloxacin resistance with mortality in
adult patients with invasive pneumococcal diseases: a post hoc analysis of a prospective
cohort. Infection. 2013;41(1):151–7.
Chapter 3
Emergence of MRSA in the Community

Lacey P. Gleason, David C. Ham, Valerie Albrecht, and Isaac See

3.1 Introduction: Staphylococcus aureus and Antimicrobial

Resistance

Staphylococcus aureus has been recognized as a cause of human infection for over
100 years, and its role in causing clinical syndromes, such as sepsis and abscesses,
was first described by Ogston in the late nineteenth century [119]. S. aureus can
colonize human hosts without causing disease, but infections with S. aureus, espe-
cially antimicrobial-resistant varieties such as methicillin-resistant S. aureus
(MRSA), contribute significantly to the burden of infectious diseases in humans.
Penicillin was first introduced to treat patients with bacterial infections in 1941,
and resistance to penicillin was first reported in S. aureus within 1–2 years [87].
These resistant strains were first found in hospitals after the Second World War,
where patients were exposed to this new antimicrobial agent [7]. S. aureus had
quickly acquired the ability to produce penicillinase, an enzyme that inactivates
penicillin. An “epidemic strain” of penicillin-resistant S. aureus, which was charac-
teristically lysed by bacteriophage 80 and 81, was noted to cause hospital outbreaks
in Australia, Canada, and the United States in the 1950s, particularly in hospitalized
children and otherwise healthy young adults [50].
In Denmark in the late 1960s, the first large-scale study of penicillin-resistant
S. aureus discovered that not only was the majority of S. aureus found in hospitals
resistant to penicillin but also the resistance gene had spread to a majority of
S. aureus strains collected from patients in community settings [79]. Within a
decade, the majority of community S. aureus strains in the United States were
penicillin-­resistant [132]. New drug development provided a solution to penicillin-

L. P. Gleason · D. C. Ham · V. Albrecht · I. See (*)

Division of Healthcare Quality Promotion, Centers for Disease Control and Prevention,
Atlanta, GA, USA
e-mail: [email protected]

© Springer International Publishing AG, part of Springer Nature 2018 39

I. W. Fong et al. (eds.), Antimicrobial Resistance in the 21st Century, Emerging Infectious
Diseases of the 21st Century, https://doi.org/10.1007/978-3-319-78538-7_3
40 L. P. Gleason et al.

resistant strains with the release of the semisynthetic penicillins (e.g., methicillin,
oxacillin, nafcillin) that resisted penicillinases produced by the majority of S. aureus
strains.
The first S. aureus isolates resistant to methicillin were isolated from patients in
England within months of the introduction of methicillin in 1959 [80]. Reports of
MRSA in the United States soon followed [8]. As with penicillin-resistant S. aureus,
MRSA strains were first seen in hospitals, prompting concerns that MRSA would
spread outside the hospital. Over 50 years later, MRSA has established itself as a
common cause of infections in community settings.
In this chapter, we discuss the epidemiology and mechanisms of resistance of
community-associated MRSA (CA-MRSA), an approach to management of
CA-MRSA infections, recommendations for prevention of MRSA in the commu-
nity, and future directions for research, focusing mostly on CA-MRSA in the United
States.

3.2  pidemiology and Mechanisms of Resistance in MRSA

E
in the Community

3.2.1  ecognition of Emergence of CA-MRSA in the United

R
States

In 1999, the Centers for Disease Control and Prevention (CDC) published a report
concerning four children from 12 months to 13 years of age who had died from
MRSA infections [17]. None of the children had risk factors for healthcare-­
associated MRSA (HA-MRSA), which at that time included recent hospitalization
or surgery, residence in a long-term care facility, or a history of injection drug use.
Although CA-MRSA had been recently reported in children [1, 71], most previ-
ously described MRSA infections in the United States were associated with health-
care settings or injection drug use in adults [93, 135]. These four cases demonstrated
that not only could patients develop MRSA disease outside of the hospital but also
that CA-MRSA disease could be severe or fatal. In addition to the four pediatric
deaths, CA-MRSA infections were reported in other populations, such as prisoners
[18] and military personnel [81]. In response to these reports, CDC initiated active
surveillance to describe the epidemiology and drug resistance patterns of MRSA
isolates in the community in the United States [56].
This and other early studies demonstrated that there were CA-MRSA strains
with unique characteristics compared to methicillin-susceptible S. aureus (MSSA)
and traditional healthcare-associated MRSA strains. Infections from these
CA-MRSA strains often had different epidemiologic risk factors, clinical manifes-
tations, and microbiological characteristics than HA-MRSA strains.
In the literature, the designations CA-MRSA and HA-MRSA have been used to
describe both distinct MRSA strains and infections with different epidemiologic
3 Emergence of MRSA in the Community 41

risk factors. CA-MRSA infection is defined epidemiologically as an MRSA

infection with onset in the community in an individual lacking established
­
HA-MRSA risk factors. Specifically, for surveillance conducted through the CDC-
sponsored Emerging Infections Program, an MRSA infection is classified as com-
munity associated if S. aureus isolates are cultured from an outpatient or less than
2 days after hospitalization and the patient has had no hospitalization, surgery, dial-
ysis, or residence in a long-term care facility within the previous year and no
indwelling central vascular catheter in the 2 calendar days prior to infection [26].
The remainder of the chapter will primarily focus on CA-MRSA infections as
defined epidemiologically by the CDC Emerging Infections Program. In this chap-
ter, we will specifically use the term “CA-MRSA strains” to refer to the strains most
closely associated with MRSA infections in the community.

3.2.2 Microbiology and Mechanisms of Resistance

S. aureus is a non-motile, non-spore-forming Gram-positive coccus that appears in

grape-like clusters. The bacteria are catalase-positive, facultative aerobes that are
usually unencapsulated. They can survive on fomites in the environment for months.
Colonization with S. aureus (i.e., growth of the organism in or on the body with-
out disease) is common in humans. Although colonization may occur in many parts
of the body (including the axillae, perineum, groin, rectum, skin, and umbilical
stump in neonates), the anterior nares are the most consistent site of colonization
[4].
Most individuals are either transiently or persistently colonized by S. aureus at
some point during their lives. Staphylococcal carriage studies have found that
16–36% of individuals are persistently colonized, 15–70% are intermittently colo-
nized, and 6–47% are never colonized [158, 168]. Colonization with S. aureus is
usually thought of as a precursor of S. aureus infection; individuals colonized with
MRSA are more likely than non-colonized individuals to develop infection [47,
161].
Microbiologic differences between MRSA isolated from patients with CA- and
HA-MRSA infections have been identified based on molecular typing, antimicro-
bial susceptibility testing, and identification of methicillin resistance and toxin
genes (Table 3.1). However, these differences are becoming less distinct as MRSA
strains that emerged in the community develop resistance to additional classes of
antimicrobial agents and enter healthcare settings [151, 156].

3.2.2.1 Mechanisms of Resistance

Penicillin resistance in S. aureus is conferred by a plasmid-associated gene (blaZ)

that codes for beta-lactamase. Methicillin resistance is usually conferred by an
altered chromosomally-encoded penicillin-binding protein (PBP2a) that causes
42 L. P. Gleason et al.

Table 3.1 Molecular characteristics of methicillin-resistant Staphylococcus aureus strains

typically considered community-associated and healthcare-associated, 2017
Characteristic Community-associated Healthcare-associated
Pulsed-field gel USA300 commonly, USA400, USA100 commonly,
electrophoresis (PFGE) type USA1000, USA1100 less commonly USA200 less
commonly
SCCmec type IV II
Presence of Panton-Valentine Common Rare
Leukocidin (PVL) toxin
Antimicrobial susceptibilitya Generally susceptible to Generally resistant to
antimicrobials other than β-lactams multiple agents
and erythromycin
Clindamycin Often susceptible Usually resistant
Erythromycin Usually resistant Usually resistant
Fluoroquinolone Variable Usually resistant
Trimethoprim-­ Usually susceptible Usually susceptible
sulfamethoxazole
a
Antimicrobial susceptibility patterns may change over time

resistance to all beta-lactam antimicrobial agents (including penicillin) and cepha-

losporins. PBP2a is encoded by the mecA gene that is carried on a distinct mobile
genetic element, the staphylococcal chromosomal cassette (SCCmec). SCCmec can
be mobilized for transfer between organisms in vitro [84], although this has histori-
cally been thought to be a rare occurrence [27]. SCCmec contains two genes (cas-
sette chromosome recombinase A and B [ccrA and ccrB]) that encode recombinases
that integrate the cassette into its chromosomal locus.
Eleven types of SCCmec have been described. Types II and IV are the primary
types seen in the United States. SCCmec Type IV has been identified in MRSA
strains from CA-MRSA infections in the United States and worldwide. MRSA
strains classically associated with healthcare transmission in the United States most
commonly contain SCCmec Type II and less commonly Types I and III. SCCmec
Type IV is also typical in some healthcare-associated strains, such as USA800 (see
Sect. 3.2.2.4 on molecular typing below). Types II and III often carry genes confer-
ring resistance to other antimicrobial agents (e.g., aminoglycosides, tetracyclines,
erythromycin, and clindamycin), whereas Type IV typically does not. This differ-
ence in community- and healthcare-associated SCCmec types often leads to differ-
ent antimicrobial agent susceptibility patterns between healthcare- and
community-associated MRSA infections.

3.2.2.2 Mechanisms of Virulence

Virulence factors enhance the ability of bacteria to cause infection by evading the
host’s defenses, increasing adherence to tissues, or spreading through tissues.
Examples of virulence factors in S. aureus include production of coagulase, toxins,
3 Emergence of MRSA in the Community 43

and proteins intrinsic to the cell wall. S. aureus produces coagulase, which interacts
with fibrinogen causing plasma to clot. This clumping creates a loose polysaccha-
ride capsule that can interfere with phagocytosis. The combination of these viru-
lence factors may cause localization of an infection, such as in an abscess, a common
clinical manifestation of CA-MRSA infection.
Panton-Valentine leukocidin (PVL) is a cytotoxin (coded by the lukS-PV and
lukF-PV genes) first identified in methicillin-susceptible S. aureus [121]. PVL kills
leukocytes by creating pores in the cell membrane of affected cells or by activating
apoptosis pathways. Pore formation leads to increased cell wall permeability and
leakage of protein from the cell causing cell death and tissue necrosis. PVL genes
have been associated with severe abscesses, necrotizing pneumonia, and increased
complications in osteomyelitis [96, 102]. PVL genes are found in most CA-MRSA
strains, such as with pulsed-field gel electrophoresis type USA300. PVL, however,
is not limited to CA-MRSA, as the toxin is also found in the majority of MSSA
strains isolated from patients with community-acquired skin and soft tissue infec-
tions (SSTIs) [64]. While PVL is rare in other S. aureus strain collections such as
colonization or clinical isolates from bloodstream infections, it is highly associated
with SSTIs [144].
In addition to PVL, other toxins may be produced by S. aureus: α-toxin, which
causes tissue necrosis and acts on cell membranes; exfoliative toxins ETA, ETB,
and ETD, which are encoded on different genetic elements and cause skin separa-
tion in diseases such as bullous impetigo and staphylococcal scalded skin syndrome;
enterotoxins A–E, G–J, and R–T (SEA-SEE, SEG-SEJ, SER-SET) which can cause
vomiting and diarrhea associated with food poisoning; and toxic shock syndrome
toxin I (TSST-1) which induces production of interleukin-1 and tumor necrosis fac-
tor leading to shock [64]. Peptidoglycans, which comprise 50% by weight of the
cell wall of staphylococci, can have endotoxin properties as well. Other cell wall
polymers (e.g., teichoic acid) and cell surface proteins (e.g., protein A and fibronec-
tin- and collagen-binding proteins) may also be virulence factors for S. aureus [64].
Recent DNA sequencing of the most common molecular type of CA-MRSA
(USA300) suggests that encoded gene products might enhance the ability of the
strain to live on the host’s skin [41].

3.2.2.3 Antimicrobial Susceptibility Testing

Antimicrobial agent susceptibility testing is commonly used in clinical laboratories

to guide the clinical treatment of S. aureus infection. Disk diffusion and broth
microdilution are the most standardized and accurate testing methods. In disk diffu-
sion tests, a disk impregnated with an antimicrobial agent is placed on an agar plate
containing a lawn of bacteria to test whether the antimicrobial agent inhibits the
growth of bacteria. (However, the vancomycin disk diffusion test does not detect
vancomycin-intermediate S. aureus isolates.) One variation of the disk diffusion test
is the E-test, a plastic strip with a gradient of antimicrobial agent concentrations
used to determine the minimal inhibitory concentration (MIC) of specific
44 L. P. Gleason et al.

antimicrobial agents. Broth microdilution tests determine the lowest concentration

of antimicrobial agents that inhibit bacterial growth in a broth medium using a stan-
dard inoculum size. In an agar screen test, a standardized suspension of the micro-
organism is inoculated directly onto an agar plate impregnated with an antimicrobial
agent. Rapid automated instrumentation, such as with devices offered by Vitek™,
Microscan™, and others, are most commonly used in laboratories to determine the
susceptibility pattern of S. aureus.
Clindamycin resistance may be constitutive or inducible, and testing for this
resistance can impact clinical treatment decisions. Resistance to clindamycin shares
some common mechanisms with resistance to erythromycin, the latter of which is
encoded by two different genes: msrA and erm [145]. The msrA gene encodes an
adenosine triphosphate (ATP)-dependent efflux pump that confers resistance to
erythromycin but not clindamycin. The erm (or erythromycin ribosomal methylase)
gene confers constitutive resistance to erythromycin and either constitutive or
inducible clindamycin resistance. MRSA isolates with inducible clindamycin resis-
tance are resistant to erythromycin and sensitive to clindamycin on routine testing
but can be induced to express resistance to clindamycin in vitro.
Rates of inducible clindamycin resistance among strains of MRSA vary widely
across the United States, from less than 10% [136] to greater than 90% [54]. A
population-based analysis of MRSA in the United States found the rate of inducible
clindamycin resistance to be almost 18%. Inducible clindamycin resistance was
much higher among USA100 isolates compared to USA300 isolates (25.9% vs
2.9%) [94]. Inducible clindamycin resistance can be identified with the D-zone test,
a double-disk diffusion test in which the zone of inhibition is measured around both
erythromycin and clindamycin disks [52]. The “D” is formed when the zone of
inhibition around the clindamycin disk is blunted on the side adjacent to the eryth-
romycin disk. A positive D-zone test indicates inducible clindamycin resistance.
The Clinical and Laboratory Standards Institute (CLSI), formerly the National
Committee for Clinical Laboratory Standards (NCCLS), recommends performing a
D-zone test on all erythromycin-resistant, clindamycin-susceptible S. aureus iso-
lates before reporting clindamycin susceptibility results [28].

3.2.2.4 Molecular Typing of MRSA

Molecular typing of MRSA strains is used to link cases in a cluster, locate sources
of specific outbreaks, and conduct macroepidemiology and evolutionary studies.
Using the antimicrobial agent susceptibility profile to determine genetic related-
ness of strains of S. aureus is unreliable. Historically, pulsed-field gel electropho-
resis (PFGE) was one of the most commonly used methods for MRSA strain typing
in outbreak investigations. Pulsed-field types are still commonly recognized in the
United States and around the world. Sequence-based typing methods such as
multi-­locus sequence typing (MLST) and spa (Staphylococcal protein A) typing
have been used in more recent years for the analysis of long-term epidemiology
3 Emergence of MRSA in the Community 45

and evolution of MRSA. MLST and spa typing have been found to be highly con-
cordant, and both typing methods are easier and less costly than performing PFGE,
providing unambiguous typing results that can be compared between laboratories
and over time [120]. An early application of whole genome sequencing was to
characterize MRSA outbreaks in healthcare settings [89].The requirement for bio-
informatics expertise is one factor that has limited adoption by healthcare institu-
tions. Nonetheless, advances in WGS and the accessibility of a large number of
assembled bacterial genomes have made possible new methods for strain-level
epidemiologic tracking of isolates. This includes the ability to apply MLST
schemes on a genome-­ wide scale [37]. Phylogenetic analysis, derived from
sequencing results, has been used to study the population structures of outbreaks
and to build transmission networks and help identify factors associated with
USA300 strains [127].
In the United States, a limited number of MRSA strains have been implicated in
most community outbreaks. A recent description of clinical MRSA isolates from 43
centers across the United States indicated that the USA300 pulsed-field type was the
most common type in all regions and from all specimen sources [40]. Other com-
munity MRSA genotypes (USA400, USA1000, and USA1100) also cause disease
[103]. Molecular typing has further classified the USA300 strain as ST8 and most
commonly spa type t008. The highly conserved USA300 strain (USA300–0114)
has been implicated in multiple outbreaks across the United States in diverse popu-
lations that are not epidemiologically related, such as athletes, prisoners, and chil-
dren [85].

3.2.2.5 Molecular Origins of MRSA

Phenotypic and molecular characterization of CA-MRSA isolates demonstrate that

they are different from major circulating HA-MRSA clones. The microbiological
differences between strains of MRSA isolated from CA-MRSA and HA-MRSA
infections give us clues as to the origins of MRSA in communities, and a number of
studies have addressed the genomic evolution of USA300. Recent findings provide
support for the idea that there has been clonal emergence of USA300 and conse-
quent spread throughout the United States, rather than evolutionary convergence
[156]. However, many questions remain regarding the evolution of USA300 and
how it has adapted to the community. Some investigators have suggested that an
arginine catabolic mobile element in the USA300 genome was transferred recently
(from an evolutionary sense) from coagulase-negative staphylococcus and may be
important for the success of this MRSA clone. The speG ACME gene product con-
fers increased survival to innate immune defenses on the skin [123, 124]. However,
ACME is absent from some USA300 isolates from household clusters, leaving
some uncertainty about the necessity of this element for evolutionary success [3].
46 L. P. Gleason et al.

3.2.3 Epidemiology of CA-MRSA

3.2.3.1 Geographic Characteristics

Cases of CA-MRSA infection have been reported worldwide. The predominant

MRSA strains (i.e., MLST sequence types) responsible for community-associated
disease vary around the world as does presence or absence of PVL in those strains,
though SCCmec type IV predominates among strains causing CA-MRSA world-
wide [38]. Reports of USA300 MRSA causing community-associated and
healthcare-­associated disease outside the Americas have been published [61, 131].
However, despite evidence of multiple introductions of USA300 in some countries,
it has not achieved the same type of epidemic success outside of the Americas [118,
153]; the reasons for this are not known. The incidence of CA-MRSA infections
varies severalfold across the United States [88].
Within the United States, CA-MRSA is found in both urban and rural settings.
However, the USA300 MRSA strain was previously reported to be rare as a cause
of CA-MRSA infections in rural Alaska [5, 33]. Rural areas have been described as
having lower rates of invasive CA-MRSA compared to urban areas [141]. However,
as injection drug use is a significant risk factor for invasive S. aureus infection and
there are concerns that injection drug use is increasing in non-urban areas in the
United States recently [167], these trends may change in the future.

3.2.3.2 Age and Sex

As discussed earlier, some of the first reports of MRSA infection without traditional
healthcare-associated risk factors were in children [17, 71]. The age distribution for
noninvasive CA-MRSA syndromes, such as SSTI, is not well described. However,
from population-based surveillance data for MRSA in the United States, invasive
(i.e., isolated from a normally sterile body site) CA-MRSA infections are most
common in persons aged 50 or older and in children less than 1 year of age [26].
Age differences exist in nasal colonization rates as well. A national study of S.
aureus nasal colonization among noninstitutionalized persons at least 1 year of age
showed that persons age 60 years or older had the highest odds of MRSA nasal car-
riage; however, this does not specifically refer to carriage of USA300 MRSA or
other community MRSA strains. For example, the same study reported that among
persons colonized with MRSA, younger persons most often had USA300 or
USA400 community MRSA strains [62]. In addition, older age was associated with
decreased odds of USA300 MRSA nasal colonization in one study [55].
Trends for MRSA colonization by sex have varied. S. aureus nasal carriage has
been reported to be more common in men than women in most studies [62, 76].
Some studies have shown higher prevalence of MRSA nasal carriage in men [76],
but a large population study of nasal carriage in the United States in 2001–2004 did
not find significantly higher prevalence of MRSA carriage in men [62]. Most studies
3 Emergence of MRSA in the Community 47

have reported higher rates of MRSA bloodstream infections in men vs women;

reasons for this are unclear. It has been postulated that these reasons might relate to
differences in behavior (e.g., studies showing that hygiene differs by sex) or to
physiological factors (e.g., hormonal differences) (reviewed in [76]).

3.2.3.3 Race, Ethnicity, and Socioeconomic Status

Rates of CA-MRSA infection vary between racial and ethnic groups. Compared
with Australians of European descent, high rates of CA-MRSA infection have been
noted in Maori and Pacific Islanders in Australia [72]. Similarly, Pacific Islanders in
Hawaii have much higher rates of CA-MRSA infection than Hawaiians of Asian
descent. In one investigation of MRSA in Hawaii, 76% of patients with CA-MRSA
were Pacific Islanders, but only 35% of patients who received care in the facility
were Pacific Islanders [48]. Rates of CA-MRSA SSTIs are high among Alaskan
natives [128].
Population-based surveillance data from the continental United States have con-
sistently shown higher incidence rates of invasive CA-MRSA infection among
black versus white persons, in both adult and pediatric populations [65, 67, 78, 88,
129]. African-American race has also been described as an independent risk factor
for colonization with the USA300 strain [55]. Analysis of public health surveillance
has demonstrated that when socioeconomic factors such as income and crowding
are accounted for, no significant differences in invasive community-associated
MRSA rates by race remain [141]. However, the specific mechanisms by which dif-
ferences in socioeconomic status lead to racial disparities in invasive community-­
associated MRSA have not yet been elucidated. It may occur through differences in
concurrent diseases such as diabetes that result from socioeconomic disparities or
from factors such as limited access to healthcare and crowded housing conditions,
which are more common in some racial groups [48]. For example, a study in urban
Chicago has described incarceration and public housing as risk factors for
CA-MRSA SSTIs [74]. The importance of socioeconomic factors as a contributor
to CA-MRSA rates is underscored by a study showing transient residence and sub-
stance abuse as factors associated with CA-MRSA compared to patients with
MSSA [159].

3.2.3.4 Other Epidemiologic Risk Factors for Colonization

HIV infection, illicit drug use, temporary housing, and incarceration are associated
with colonization with USA300 MRSA [126]. In addition, antimicrobial use is also
a risk factor for MRSA colonization in the community [47, 105, 106].
48 L. P. Gleason et al.

3.3  onsiderations in Management of MRSA Infection

C
in the Community

Clinical treatment guidelines for MRSA were published by the Infectious Disease
Society of America (IDSA) in 2011 [97]. This section will review major recommen-
dations from those guidelines as well as new developments that have occurred since
then.

3.3.1 Clinical Presentation

CA-MRSA infections can occur as any one of a myriad of clinical syndromes.

Among the most common are SSTIs, pneumonia, and invasive infections such as
osteomyelitis, endocarditis, and other bloodstream infections [56]. Wound, skin,
and soft tissue MRSA infections account for approximately 90% of CA-MRSA
infections [56]. The majority of SSTIs are abscesses or cellulitis, but up to one quar-
ter are superficial infections such as impetigo [112]. MRSA and other S. aureus
SSTIs have often been misdiagnosed as “spider bites,” particularly by patients [30,
42] even when the relevant spiders (e.g., brown recluse) are not present in those
regions [160]. Whether invasive CA-MRSA infections from the USA300 strain or
other PVL-positive strains are more severe than other MRSA infections remains
unclear [92, 113, 144]. MRSA pneumonia has been reported as a bacterial superin-
fection after influenza infection in both adults and children [11, 25, 66, 125]. MRSA
can be a cause of otitis media in children [134] and accounted for as much as 12%
of cases of otorrhea in one series [77]. MRSA has also been a reported cause of
pyomyositis [82] and Waterhouse-Friderichsen syndrome in children [2]. In addi-
tion, cases of CA-MRSA necrotizing fasciitis have been reported [104].
CA-MRSA may also cause recurrent infections, as there is no known natural
immunity to S. aureus after infection. Severity of disease can vary by site of infec-
tion. In one study, approximately one quarter of patients with CA-MRSA infection
were hospitalized for their infection [56], with a greater percentage among those
with severe or invasive disease [112].

3.3.2 Management of MRSA Skin or Soft Tissue Infections

MRSA has increased in prevalence as a cause of purulent SSTIs since the early
2000s and is the most common cause in many communities [151]. Some of the
major considerations for treatment of these infections outlined in the IDSA guide-
lines, along with a brief discussion of the literature, are listed below:
3 Emergence of MRSA in the Community 49

1. Incision and drainage should be routine for skin lesions that can be drained.
Some clinicians suggest the use of ultrasonography to distinguish if there is a
drainable collection [148]. Incision and drainage has been the primary mode of
treatment for skin and soft tissue abscesses for many centuries and is essential.
Although the 2011 IDSA guidelines suggest that additional data are needed to
describe whether antimicrobials are needed for simple abscesses/boils, a ran-
domized trial at five US emergency departments showed that trimethoprim-­
sulfamethoxazole (TMP-SMX) treatment led to higher cure rates among patients
with a drained cutaneous abscess compared with patients who received placebo
[152]. Antimicrobial treatment may also reduce recurrent infections [45, 139,
152] and the need for subsequent procedures [152]. Available literature suggests
that clinicians should weigh the costs and benefits of prescribing antimicrobial
agents: for example, considering an individual patient’s risk for recurrent
infections.
2. Collect diagnostic specimens for culture. Cultures should be obtained from
patients with both draining and non-draining purulent lesions. Obtaining isolates
helps guide treatment of individual patients and monitor the antimicrobial agent
susceptibility patterns in the community. Cultures are also recommended by
some experts when there is a severe local infection, signs of systemic illness,
inadequate response to initial treatment, or concerns about a cluster or outbreak
of cases [148].
3. Use of antimicrobial agents. For adult patients, the IDSA guidelines suggest
using clindamycin, TMP-SMX, a tetracycline, or linezolid for empiric coverage
of CA-MRSA in outpatients with SSTI when empiric coverage is needed, with
the additional suggestion that if coverage for β-hemolytic streptococci is also
needed, a beta-lactam be added if TMP-SMX or a tetracycline is prescribed. For
hospitalized patients with complicated SSTI, empiric therapy is suggested to
include vancomycin, linezolid, daptomycin, telavancin, or clindamycin. Local
antimicrobial susceptibility data for outpatient S. aureus SSTIs should guide
empiric treatment decisions. The 2011 IDSA guidelines specifically suggest
empiric therapy for CA-MRSA be prescribed for outpatients with purulent cel-
lulitis, hospitalized patients with complicated SSTI, and abscess associated with
certain conditions—severe or extensive disease or rapid progression with cellu-
litis, signs/symptoms of systemic illness, immunosuppression, extremes of age,
lack of response to initial treatment, septic phlebitis, or anatomically difficult-to-­
drain location. The guidelines state that the role of CA-MRSA in outpatient non-
purulent cellulitis is unknown, and despite use of state-of-the-art techniques for
molecular identification, the microbiologic etiology of nonpurulent cellulitis in
the community and role of MRSA in particular remain unknown [32]. Notably a
recent study has shown no difference in outcomes with the use of cephalexin
alone compared with the use of cephalexin plus TMP-SMX in patients with
uncomplicated cellulitis [109].
4. Perform careful and thorough personal and environmental hygiene. MRSA can
be transmitted from person to person or to the environment through contact with
draining skin and soft tissue lesions. After incision and drainage, wounds should
50 L. P. Gleason et al.

be adequately covered, bandages should be appropriately disposed of, and hand

hygiene should be continued to prevent further spread of MRSA. In addition,
items that have contacted infected skin such as towels should not be reused or
shared.

3.3.3 Management of Severe or Invasive MRSA Infections

Severe or invasive MRSA infections include sepsis, pneumonia, endocarditis, osteo-

myelitis, and the progression of localized infections such as of the skin or soft tis-
sue. Empiric therapy for MRSA is recommended in IDSA guidelines for severe
cases of hospitalized community-acquired pneumonia, defined as (1) requiring ICU
admission, (2) having necrotizing or cavitary infiltrates, or (3) being associated with
empyema [97]. Recent studies of community-acquired pneumonia have found
MRSA to be an uncommon pathogen, and though such cases are severe, it is unclear
whether clinical presentation of MRSA significantly differs from that of other
community-­acquired pneumonia pathogens to distinguish it on clinical ground
alone [108, 143]. In addition to antimicrobial agent therapy, incision and drainage is
mandatory for drainable SSTIs and should be considered for severe or deep infec-
tions, such as septic joints or osteomyelitis. In addition, for infections related to an
implanted device, the device should be removed if feasible [97].
Vancomycin and daptomycin are recommended as first-line agents for adults
with MRSA bacteremia. Duration of therapy depends on whether bacteremia is
determined to be complicated or uncomplicated and whether or not associated with
endocarditis [97].

3.3.4  ewer Developments in Antimicrobial Treatment

N
of MRSA Infections

Several antimicrobials have been developed more recently with in vitro activity
against MRSA, many of which also have indications for skin infections or other
syndromes that may be CA-MRSA infections (Table 3.2). These include dalba-
vancin, oritavancin, ceftaroline, and tedizolid [14]. Some clinicians have suggested
combination therapy (e.g., either vancomycin or daptomycin, in conjunction with a
β-lactam) for complicated MRSA bacteremia [39]. Clinical trials are underway to
further define the role of combination therapy [154]. In addition, a new clinical
treatment guideline for S. aureus bacteremia is planned by IDSA and may clarify
the role of newer agents or combination therapy for bacteremia for treatment of
CA-MRSA syndromes.
3 Emergence of MRSA in the Community 51

Table 3.2 Description of major antimicrobials or antimicrobial classes that have been used to treat
methicillin-resistant Staphylococcus aureus infections
Antimicrobial Mechanism of action Other comments
Ceftaroline Binding to penicillin-binding High affinity for penicillin-binding protein
proteins (fifth-generation 2A (PBP2A), leading to greater in vitro
cephalosporin) activity against MRSA than other
cephalosporin antibiotics
Clindamycin Inhibition of bacterial protein Inducible resistance can occur
synthesis
Dalbavancin Same as vancomycin Dosing for skin infections approved as
either single dose or two doses 1 week
apart
Daptomycin Cyclic lipopeptide; binds to Inactivated by surfactant and not
bacterial cell membranes recommended for treatment of MRSA
pneumonia
Linezolid Inhibition of ribosomal Myelosuppression reported; could cause
protein synthesis serotonin syndrome in conjunction with
some medications (e.g., antidepressants)
Oritavancin Inhibits cell wall synthesis Once/week dosing
(similar to vancomycin) and
disrupts cell membrane
barrier function
Quinupristin-­ Inhibits peptide bond Arthralgias/myalgias are common adverse
dalfopristin formation in ribosome events
Tedizolid Same as linezolid Approved for skin infections; advantage
over linezolid is once/day dosing
Telavancin Inhibits cell wall synthesis Synthetic derivative of vancomycin but
(similar to vancomycin) and once/day dosing
disrupts cell membrane
barrier function
Tetracyclines (class) Inhibition of bacterial protein Not recommended in children because of
synthesis effects on bones and teeth
Tigecycline Inhibits protein translation FDA boxed warning about increased risk
of death
Trimethoprim-­ Blocks production of folic
sulfamethoxazole acid
Vancomycin Inhibits cell wall synthesis Monitoring of serum levels recommended
by IDSA guidelines; dosing in obese
patients controversial; often drug of choice
for patients with bacteremia

3.3.5 Strategies to Eliminate S. aureus Colonization as Part

of Treatment for Infected Patients

Decolonization has been suggested as a component of treatment for recurrent or

persistent MRSA infections, but its value is unclear. Decolonization regimens have
been effective in reducing colonization in the short term, but recolonization is com-
mon [90, 122]. Carriage at sites other than the nares and reports of resistance to
52 L. P. Gleason et al.

mupirocin may limit the effectiveness of nasal decolonization [98]. Proposed regi-
mens for decolonization include intranasal mupirocin twice a day for 5–10 days and
antiseptic body wash, such as with chlorhexidine, for 5–14 days [97, 150]. Dilute
bleach baths have been suggested for patients with recurrent MRSA infections, but
children who underwent routine hygienic measures plus twice weekly bleach baths
for 3 months did not experience a significant reduction in recurrent SSTI requiring
medical attention within a year of treatment compared with those using routine
hygienic measures [83]. A randomized controlled trial testing the effect of skin
cleaning with chlorhexidine gluconate (CHG) three times per week for 6 months in
a jail showed no difference in MRSA carriage between groups cleaning with CHG
cloths compared with water-soaked cloths [35]. When household transmission is
suspected, implementation of personal and environmental hygiene measures and
evaluation of symptomatic contacts are recommended in IDSA guidelines, which
state that decolonization may be considered [97]. Decolonization with a 5-day regi-
men of hygiene, nasal mupirocin, and chlorohexidine body washes of all household
members was associated with a significant decrease in self-reported recurrent SSTI
at 12-month follow-up compared to a group in which only the index patient was
decolonized [57]. However, SSTI still recurred in the majority of cases [57]. The
use of oral antimicrobial therapy is generally not recommended for decolonization
in current guidelines [97].

3.4  ransmission and Outbreaks of MRSA

T
in the Community

3.4.1 Transmission of CA-MRSA

The following information describes some of what is known about transmission of

MRSA in the community from studies in non-outbreak settings. In general, direct
person-to-person transmission of S. aureus is believed to occur via contact, as opposed
to respiratory droplets or aerosols [147]. Whole genome single-nucleotide polymor-
phism (SNP) analysis of community MRSA isolates from a hospital system in
New York demonstrated that isolates from the same household are much more closely
related than community isolates from different households, suggesting transmission
and possibly persistence within households [155]. Analysis of USA300 isolates from
Los Angeles and Chicago also showed through whole genome sequencing that iso-
lates often clustered into closely related groups by household [3]. In addition, the
study suggested that introduction of MRSA colonization within the household often
preceded the first symptomatic infection of the household member. Household studies
have suggested that factors associated with household transmission of MRSA include
persons in the household requiring assistance for daily activities or sharing topical
products or bath towels [114, 116]. Colonized and infected individuals can also con-
taminate the household environment [58]. Therefore in summary, environmental sur-
faces, shared items, and hands/skin may serve as vehicles for transmission.
3 Emergence of MRSA in the Community 53

3.4.2 Outbreaks of MRSA in the Community

In addition to the above studies, a significant amount of information has been

learned about CA-MRSA transmission from outbreaks in the community. Although
some experts have questioned whether risk factors for transmission might differ in
outbreak versus non-outbreak settings, the lessons learned from outbreaks can still
be useful.
Since the early 1980s when MRSA was recognized as a pathogen that can cause
outbreaks in the community in groups such as intravenous drug users, outbreaks of
CA-MRSA have been reported in a number of diverse groups: Native American,
Alaskan Native, and Pacific Islander communities [5, 6, 48, 63, 72]; prisoners [18,
22]; amateur and professional sports participants, such as football players, wres-
tlers, rugby players, fencers, and divers [9, 21, 85, 149, 163]; child care center
attendees [1]; military personnel [13, 170]; men who have sex with men [91]; meth-
amphetamine and injection drug users [29, 53]; survivors of natural disasters [23];
recipients of tattoos [24]; and isolated religious communities [31]. MRSA can cause
infections in animals and pets and has been reported to cause infections in humans
who have had contact with infected animals [165].
Although these groups are diverse, they have common factors that may underlie
the transmission of MRSA in the community. Based on investigations of commu-
nity outbreaks, five factors that contribute to transmission of MRSA in the commu-
nity can be characterized as the “Five Cs” as described below. However, it is
important to note that in a study at an urban emergency department in the United
States, most patients with MRSA skin infections had none of these characteristic
risk factors associated with CA-MRSA outbreaks [107]:
1. Crowding. Outbreaks have occurred in populations living in crowded quarters
such as prisons and military barracks. Living in a house with more than one per-
son per bedroom has been independently associated with developing a CA-­
MRSA skin or soft tissue infection [29].
2. Contact, skin-to-skin. Participants in contact sports have frequent skin-to-skin
contact, which may act as a method of transmitting MRSA SSTI. Outbreaks
among professional and college football teams have been attributed to frequent
skin-to-skin contact [9, 85]. Similarly, wrestlers who have significant skin-to-­
skin contact have experienced outbreaks of MRSA infection [21]. High-risk
sexual behavior [91] and sexual contact with someone with a skin infection [29,
91] have both been associated with CA-MRSA SSTIs. These factors have been
described both in rural and urban communities.
3. Cut or compromised skin. Breaks in the skin are a portal for MRSA bacteria to
enter the body. For example, in an outbreak of MRSA infections among a college
football team, MRSA infections were associated with abrasions from artificial
grass (“turf burns”) and cosmetic body shaving [9]. In an outbreak among mili-
tary recruits, most of the MRSA SSTIs were on exposed skin of the arms, legs,
and knees, where abrasions are common during field training [170]. Skin-­picking
behavior has also been associated with MRSA SSTIs [29]. Injection drug use,
54 L. P. Gleason et al.

where the skin is compromised by insertion of contaminated needles, has been

associated with MRSA infections [104, 169], but injection may not be the only
method by which MRSA is transmitted among drug users [29].
4. Contaminated surfaces and shared items. Although environmental transmission
of MRSA may not be the most common mode of transmission, the environment
may have played a role in some outbreaks of MRSA in the community. Outbreaks
have been associated with whirlpools [9] and MRSA contaminated sauna
benches [6]. An outbreak among fencers was unusual because there is typically
little skin-to-skin contact in that sport; however, investigators surmised that the
cluster of cases was due to shared fencing equipment [21]. In a correctional facil-
ity in Mississippi, sharing personal items such as linens was associated with
infection [18], while sharing bars of soap was implicated in an outbreak among
members of a college football team [117].
5. Cleanliness. Cleanliness includes both personal bathing and laundering of cloth-
ing, linens, and towels, all of which have been noted as potential contributing
factors to CA-MRSA infection among prison inmates [18]. Investigations of
MRSA transmission in prisons suggest that lack of access to basic hygiene is a
contributing factor [22]. Homelessness has also been associated with MRSA
SSTIs [169].
In addition to the “Five Cs,” previous use of antimicrobial agents has also been
shown to be a factor in the development of CA-MRSA [5, 85]. An outbreak of
CA-MRSA skin infections in southwestern Alaska found that patients with skin
infections received significantly more antimicrobial agents in the year before the
outbreak compared to community members without skin infections [6]. In an out-
break of MRSA in a closed religious community in the United States, investigators
found the use of antimicrobial agents was associated with infection [31].

3.5 Prevention of MRSA in the Community

Prevention strategies for CA-MRSA need to include public health officials, medical
providers and infection control practitioners, and patients and community members.
The following are considerations for these different groups for prevention of
CA-MRSA.

3.5.1 Public Health Officials

1. Consider initiating public health investigations when MRSA is detected in a

group of individuals in the community who are linked epidemiologically. When
considering whether to investigate, public health officials should weigh the num-
ber and clustering of time and space of cases, the setting of the cluster, the sever-
3 Emergence of MRSA in the Community 55

ity of illness, the presence of ongoing transmission, and the likelihood that an
intervention could be successfully implemented.
2. Enhance surveillance. Both prospective and retrospective surveillance are impor-
tant to identify cases of MRSA in the community and intervene in outbreak set-
tings. Consider notifying contacts of patients with MRSA infection to identify
new cases in outbreak settings and to ensure that they are receiving proper
treatment.

3.5.2 Medical Providers and Infection Control Practitioners

1. Use appropriate treatments for infections. Treatment considerations for MRSA

infections were discussed in an earlier section.
In addition, since CA-MRSA infections have been associated with previous anti-
microbial use, antimicrobials should be used appropriately, both when treating
patients with CA-MRSA infections and when prescribing antimicrobials for other
conditions in the community.
2. Educate providers to assess for additional symptomatic contacts. Clinicians
should ask patients with MRSA infections if other contacts and household mem-
bers also have suspicious lesions or infections, so that contacts can be appropri-
ately treated and further transmission limited.
3. Prevention of MRSA in healthcare settings. Recommendations for preventing
MRSA in healthcare settings are reviewed below, although it is not clear if these
have significant impact on preventing CA-MRSA infections. Recommendations
from CDC’s Healthcare Infection Control Practices Advisory Committee for
preventing MRSA infections in the healthcare setting where MRSA is consid-
ered an epidemiologically important multidrug-resistant organism generally
include (1) promotion of appropriate hand hygiene, (2) use of contact precau-
tions for MRSA-colonized and MRSA-infected patients, (3) appropriate clean-
ing and disinfection of patient equipment and environmental surfaces, and (4)
educating healthcare personnel about MRSA prevention and transmission [146].
However, the use of contact precautions has recently been questioned, and addi-
tional research would be helpful to define the added benefits of contact precau-
tions to other measures for controlling transmission of MRSA in acute care
hospitals [110]. Additional measures are suggested for consideration if MRSA is
not adequately controlled using standard measures. Active surveillance, i­ solation,
and cohorting of MRSA-positive patients can also be implemented [12, 146]. In
addition, based on results from a positive clinical trial, the Society for Healthcare
Epidemiology of America also recommends universal decolonization of ICU
patients with chlorhexidine and nasal mupirocin in hospital locations that con-
tinue to have high MRSA rates after the implementation of basic MRSA control
strategies [12, 75].
56 L. P. Gleason et al.

3.5.3 Patients and Members of the Community

1. Educate patients on treatment and prevention. Patients should be encouraged to

keep wounds covered, to maintain good personal and hand hygiene, to avoid
sharing potentially contaminated items, and to seek care early if they believe
they might have an infection (CDC MRSA website https://www.cdc.gov/mrsa/
community/index.html). Prevention recommendations may need to be tailored
for high-risk groups (e.g., athletes).
2. Care for and contain wounds. Wounds should be covered with clean, dry dress-
ings until healed. Patients with open skin wounds, such as draining SSTIs that
cannot be covered, may need to be excluded from activities that could lead to
transmission. For example, if sport-specific rules do not exist, in general, athletes
should be excluded if wounds cannot be properly covered during participation,
and athletes with open wounds or infections should not use common-use water
facilities like swimming pools or therapy pools. Patients should be encouraged
to seek care from a medical provider and not treat wounds themselves by picking
or popping sores (CDC MRSA website https://www.cdc.gov/mrsa/community/
index.html).
3. Encourage personal hygiene, especially hand hygiene. Patients should wash
hands regularly especially after dressing changes. Use soap and water or alcohol-­
based hand gels to clean hands and encourage regular bathing or showering,
especially after exercise. Do not share personal items that may transmit infection
such as towels, washcloths, razors, and clothing. Launder contaminated clothes
and linens with detergent, soap, or bleach, and dry thoroughly. Athletic uniforms
should be washed and dried after each use (CDC MRSA website https://www.
cdc.gov/mrsa/community/index.html [142]).
4. Maintain a clean environment. For example, when MRSA skin infections occur,
cleaning and disinfection should be performed on surfaces likely to contact
uncovered or poorly covered infections, such as in facilities where patrons and
staff have close contact (e.g., homeless shelters) or shared equipment or surfaces
(e.g., gyms). Surfaces should be cleaned with detergent-based cleaners or
Environmental Protection Agency (EPA)-registered disinfectants (­https://www.
epa.gov/pesticide-registration/selected-epa-registered-disinfectants).

3.6 Future Directions

3.6.1 Vancomycin Resistance

Vancomycin is a primary treatment for severe, invasive MRSA infections. Clinical

isolates of S. aureus with intermediate resistance to vancomycin (MICs of 8–16
mug/mL) were first reported in Japan in the late 1990s [73]. Intermediate resistance
3 Emergence of MRSA in the Community 57

to vancomycin in strains of S. aureus may be due to the development of thicker cell

walls in the bacteria.
Resistance to vancomycin is conferred by the presence of a vanA operon, which
is thought to be transferred from vancomycin-resistant enterococci [166]. In 2002,
reports of S. aureus resistant to vancomycin (MICs ≥ 32 ug/mL) came from two
states (Michigan and Pennsylvania) in the United States [19, 20]. To date, 14
vancomycin-­resistant S. aureus (VRSA) isolates have been identified in the United
States [162]. All have occurred in patients with significant prior healthcare encoun-
ters, such as for chronic wounds and dialysis. In addition, all except for the 13th
isolate have belonged to lineages traditionally associated with healthcare; the 13th
isolate was a USA1100 MRSA [95]. In addition, VRSA from a community MRSA
lineage has been reported in Brazil, where vancomycin resistance appears to have
been acquired while a patient was on vancomycin treatment [133].

3.6.2  accine and Other Novel Prevention/Treatment

V
Approaches

New mechanisms of preventing and treating S. aureus infection are being studied.
Vaccines against S. aureus are being developed. Although significant difficulties
have been encountered with previous studies of potential S. aureus vaccines, there
are clinical trials of vaccine candidates ongoing [10, 59, 100, 138]. In addition,
some studies of antibodies to treat or prevent S. aureus infections are also underway
[59]. Another research area that has attracted some interest recently is the use of
bacteriophages for prevention or treatment of S. aureus infections, though the appli-
cation of such an approach is likely distant [86].

3.6.3 Novel Potential Sources for MRSA in the Community

3.6.3.1 Pets and MRSA

MRSA is also an important pathogen in veterinary medicine [60], and some

researchers have explored the relationship between animal and human MRSA infec-
tions. There have been case reports of suspected transmission of MRSA between
owners and pets [16, 43, 44, 101], and models of MRSA acquisition in dogs identify
contact with humans as an influential source [69, 70]. Little is known about the
potential for companion animals to serve as reservoirs of human infections, but a
small percentage of healthy cats and dogs have been shown to carry multidrug-­
resistant staphylococci [36]. Approximately 6–9% of pets living in the same house-
hold as a patient infected with MRSA carried genetically concordant strains of
MRSA [49, 51, 111], but the odds of obtaining a positive culture from pets rapidly
decrease from the time of the patient’s MRSA diagnosis [111]. Additional studies
58 L. P. Gleason et al.

might clarify the dynamics of, risk factors for, and importance of this potential hori-
zontal transmission between pets and humans.

3.6.3.2 Livestock and MRSA

The use of antimicrobials in industrial agriculture is associated with the presence of

antimicrobial-resistant bacteria [68]. In Europe, livestock-associated MRSA has
been described as a unique MRSA clone (MLST sequence type 398) associated with
livestock exposure, that can cause a range of infections including bacteremia, pneu-
monia, osteomyelitis, endocarditis, and SSTIs [46, 99, 157]. Some have raised con-
cerns in the United States for occupational and environmental exposures, as MRSA
has been found in dust and surface samples within industrial hog operations, air and
soil samples in the surrounding environment, and surface waters near industrial hog
operation spray fields [68]. Among patients admitted to a rural, tertiary care hospital
without occupational exposure to industrial hog operations, individuals living in
areas with higher swine density have higher odds of MRSA carriage [137]. Proximity
to swine crop fields is also associated with CA-MRSA and SSTIs [15]. However, in
one study, individuals who have occupational exposure to industrial hog operations
were not at increased risk of MRSA colonization compared to community referents
[115]. In addition, a study of individuals in the United States found that livestock
exposure was a risk factor for colonization with S. aureus but not for SSTIs overall,
and only one participant with livestock exposure was found to have the livestock-
associated MRSA strain [164]. The ST398 MRSA strain has not been detected to
date in multisite public health surveillance in the United States for invasive MRSA
(Centers for Disease Control and Prevention, unpublished data). At this point, it is
unclear whether observed differences in MRSA colonization in areas with concen-
trated livestock operations are reproducible and, most importantly, if they meaning-
fully contribute to the burden of CA-MRSA infections in the United States.

3.6.4 Changes in Molecular Epidemiology of MRSA Strains

CA-MRSA strains have been increasingly described as the cause of disease and
outbreaks in healthcare settings. USA300 MRSA was the most common MRSA
strain causing bloodstream infections in a Chicago hospital and has been reported to
make up an increasing proportion of hospital-onset MRSA infections [34, 130].
Surveillance from CDC’s EIP suggests that the increase in the proportion of
hospital-­
onset MRSA infections caused by USA300 has occurred principally
because the incidence of MRSA infections caused by other strains in hospitals has
declined [140]. Data from the same surveillance system show that the incidence of
invasive USA300 infections has not decreased over the past decade, and additional
strategies for preventing USA300 MRSA infections in both healthcare and the com-
munity are needed.
3 Emergence of MRSA in the Community 59

The epidemiology of CA-MRSA infections continues to evolve. We will need to

be vigilant in our identification and treatment of MRSA infections in the future to
prevent further spread and development of new resistant strains. Innovative methods
of preventing CA-MRSA disease would provide a large benefit to public health.

Acknowledgments/Disclaimer The findings and conclusions in this report are those of the
authors and do not necessarily represent the official position of the Centers for Disease Control and
Prevention. Use of trade names and commercial sources is for identification only and does not
imply endorsement by the Centers for Disease Control and Prevention, the Public Health Service,
or the US Department of Health and Human Services. The authors would like to thank the authors
of the previous version of this chapter (Adam Cohen, Daniel Jernigan, and Rachel Gorwitz) for
permission to include some materials from the previous edition.

References

1. Adcock PM, Pastor P, Medley F, Patterson JE, Murphy TY. Methicillin-resistant

Staphylococcus aureus in two child care centers. J lnfect Dis. 1998;178:577–80.
2. Adem PV, Montgomery CP, Husain AN, Koogler TK, Arangelovich V, Humilier M, Boyle-­
Vavra S, Daum RS. Staphylococcus aureus Sepsis and the Waterhouse-Friderichsen syn-
drome in children. N Engl J Med. 2005;353:1245–51.
3. Alam MT, Read TD, Petit RA, Boyle-Vavra S, Miller LG, Eells SJ, Daum RS, David
MZ. Transmission and microevolution of USA300 MRSA in US households: evidence from
whole genome sequencing. MBio. 2015;6:e00054-15.
4. Albrecht VS, Limbago BM, Moran GJ, Krishnadasan A, Gorwitz RJ, McDougal LK, Talan
DA, The EMERGency ID NET Study Group. Staphylococcus aureus colonization and strain
type at various body sites among patients with a closed abscess and uninfected controls at US
emergency departments. J Clin Microbiol. 2015;53:3478–84.
5. Baggett HC, Hennessy TW, Leman R, Hamlin C, Bruden D, Reasonover A, Martinez P,
Butler JC. An outbreak of community-onset methicillin-resistant Staphylococcus aureus skin
infections in southwestern Alaska. lnfect Control Hosp Epidemiol. 2003;24:397–402.
6. Baggett HC, Hennessy TW, Rudolph K, Bruden D, Reasonover A, Parkinson A, Sparks R,
Donlan RM, Martinez P, Mongkolrattanothai K, Butler JC. Community-onset methicillin-­
resistant Staphylococcus aureus associated with antibiotic use and the cytotoxin
­Panton-­Valentine leukocidin during a furunculosis outbreak in rural Alaska. J Infect Dis.
2004;189:1565–73.
7. Barber M, Rozwadowska-Dowzenko M. Infection by penicillin-resistant staphylococci.
Lancet. 1948;1:641–4.
8. Barrett FF, McGehee RF Jr, Finland M. Methicillin-resistant Staphylococcus aureus at Boston
City Hospital. Bacteriologic and epidemiologic observations. N Engl J Med. 1968;279:441–8.
9. Begier EM, Frenette K, Barrett NL, Mshar P, Petit S, Boxrud DJ, Watkins-Colwell K,
Wheeler S, Cebelinski EA, Glennen A, Nguyen D, Hadler JL, Connecticut Bioterrorism
Field Epidemiology Response Team. A high-morbidity outbreak of methicillin-resistant
Staphylococcus aureus among players on a college football team, facilitated by cosmetic
body shaving and turf burns. Clin Infect Dis. 2004;39:1446–53.
10. Begier E, Seiden DJ, Patton M, Zito E, Severs J, Cooper D, Eiden J, Gruber WC, Jansen KU,
Anderson AS, Gurtman A. SA4g, a 4-antigen Staphylococcus aureus vaccine, rapidly induces
high levels of bacteria-killing antibodies. Vaccine. 2017;35:1132–9.
11. Bhat N, Wright JG, Broder KR, Murray EL, Greenberg ME, Glover MJ, Likos AM, Posey
DL, Klimov A, Lindstrom SE, Balish A, Medina MJ, Wallis TR, Guarner J, Paddock CD,
60 L. P. Gleason et al.

Shieh WJ, Zaki SR, Sejvar JJ, Shay DK, Harper SA, Cox NJ, Fukuda K, Uyeki TM, Influenza
Special Investigations Team. Influenza-associated deaths among children in the United
States, 2003-2004. N Engl J Med. 2005;353:2559–67.
12. Calfee DP, Salgado CD, Milstone AM, Harris AD, Kuhar DT, Moody J, Aureden K, Huang
SS, Maragakis LL, Yodoe DS, Society for Healthcare Epidemiology of America. Strategies to
prevent methicillin-resistant Staphylococcus aureus transmission and infection in acute care
hospitals: 2014 update. Infect Control Hosp Epidemiol. 2014;35:s108–32.
13. Campbell KM, Vaughn AF, Russell KL, Smith B, Jimenez DL, Barrozo CP, Minarcik JR, Crum
NF, Ryan MA. Risk factors for community-associated methicillin-resistant Staphylococcus
aureus infections in an outbreak of disease among military trainees in San Diego, California,
in 2002. J Clin Microbiol. 2004;42:4050–43.
14. Cardona AF, Wilson SE. Skin and soft-tissue infections: a critical review and the role of tela-
vancin in their treatment. Clin Infect Dis. 2015;61:S69–78.
15. Casey JA, Curriero FC, Cosgrove SE, Nachman KE, Schwartz BS. High-density livestock
operations, crop field application of manure, and risk of community-associated methicillin-­
resistant Staphylococcus aureus infection in Pennsylvania. JAMA Int Med. 2013;173:1980.
16. Cefai C, Ashurst S, Owens C. Human carriage of methicillin-resistant Staphylococcus aureus
linked with pet dog. Lancet. 1994;344:539–40.
17. Centers for Disease Control and Prevention. Four pediatric deaths from community-acquired
methicillin-resistant Staphylococcus aureus—Minnesota and North Dakota, 1997-1999.
MMWR Morb Mortal Wkly Rep. 1999;48:707–10.
18. Centers for Disease Control and Prevention. Methicillin-resistant Staphylococcus aureus skin
or soft tissue infections in a state prison—Mississippi, 2000. MMWR Morb Mortal Wkly
Rep. 2001;50:919–22.
19. Centers for Disease Control and Prevention. Staphylococcus aureus Resistant to vancomy-
cin—United States, 2002. MMWR Morb Mortal Wkly Rep. 2002;51:565–7.
20. Centers for Disease Control and Prevention. Vancomycin-resistant Staphylococcus aureus—
Pennsylvania, 2002. MMWR Morb Mortal Wkly Rep. 2002;51:902.
21. Centers for Disease Control and Prevention. Methicillin-resistant Staphylococcus aureus
infections among competitive sports participants—Colorado, Indiana, Pennsylvania, and Los
Angeles County, 2000-2003. MMWR Morb Mortal Wkly Rep. 2003;52:793.
22. Centers for Disease Control and Prevention. Methicillin-resistant Staphylococcus aureus
infections in correctional facilities—Georgia, California, and Texas, 2001-2003. MMWR
Morb Mortal Wkly Rep. 2003;52:992–6.
23. Centers for Disease Control and Prevention. Infectious disease and dermatologic conditions
in evacuees and rescue workers after hurricane Katrina—multiple states, august-September,
2005. MMWR Morb Mortal Wkly Rep. 2005;54:961–4.
24. Centers for Disease Control and Prevention. Methicillin-resistant Staphylococcus aureus
skin infections among tattoo recipients—Ohio, Kentucky, and Vermont, 2004-2005. MMWR
Morb Mortal Wkly Rep. 2006;55:677–9.
25. Centers for Disease Control and Prevention. Severe co-infection with seasonal influenza a
(H3N2) virus and Staphylococcus aureus – Maryland, February-march 2012. MMWR Morb
Mortal Wkly Rep. 2012;61:289–91.
26. Centers for Disease Control and Prevention. Emerging infections program network report:
methicillin-resistant Staphylococcus aureus, 2015. Published February 28, 2017. Available
online at: https://www.cdc.gov/hai/eip/pdf/2015-MRSA-Report.pdf. 2017.
27. Chambers HF. The changing epidemiology of Staphylococcus aureus? Emerg Infect Dis.
2001;7:178–82.
28. Clinical and Laboratory Standards Institute. Performance standards for antimicrobial suscep-
tibility testing; 16th informational supplement. Wayne: The Institute; 2006. p. M100–S16.
29. Cohen AL, Shuler C, McAllister S, Fosheim GE, Brown MG, Abercrombie D, Anderson
K, McDougal LK, Drenzek C, Arnold K, Jernigan D, Gorwitz R. Methamphetamine
3 Emergence of MRSA in the Community 61

use and methicillin-resistant Staphylococcus aureus skin infections. Emerg Infect Dis.
2007;13:1707–13.
30. Cohen PR. Community-acquired methicillin-resistant Staphylococcus aureus skin infections:
implications for patients and practitioners. Am J Clin Dermatol. 2007;8:259–70.
31. Coronado F, Nicholas JA, Wallace BJ, Kohlerschmidt DJ, Musser K, SchoonmakerBopp DJ,
Zimmerman SM, Boller AR, Jernigan DB, Kacica MA. Community-associated methicillin-­
resistant Staphylococcus aureus skin infections in a religious community. Epidemiol Infect.
2006;135:492–501.
32. Crisp JG, Takhar SS, Moran GJ, Krishnadasan A, Dowd SE, Finegold SM, Summanen PH,
Talan DA, EMERGency ID Net Study Group. Inability of polymerase chain reaction, pyro-
sequencing, and culture of infected and uninfected site skin biopsy specimens to identify the
cause of cellulitis. Clin Infect Dis. 2015;61:1679–87.
33. David MZ, Rudolph KM, Hennessey TW, Boyle-Vavra B, Daum RS. Molecular epidemiol-
ogy of methicillin-resistant Staphylococcus aureus, rural southwestern Alaska. Emerg Infect
Dis. 2008;14:1693–9.
34. David MZ, Cadilla A, Boyle-Vavra S, Daum RS. Replacement of HA-MRSA by CA-MRSA
infections at an academic medical center in the Midwestern United States, 2004-2005 to
2008. PLoS One. 2014;9:e92760.
35. David MZ, Siegel JD, Henderson J, Leos G, Lo K, Iwuora J, Porsa E, Schumm LP, Boyle-­
Vavra S, Daum RS. A randomized, controlled trial of Chlorhexidine-soaked cloths to reduce
methicillin-resistant and methicillin-susceptible Staphylococcus aureus carriage prevalence
in an urban jail. Infect Control Hosp Epidemiol. 2014;35:1466–73.
36. Davis J, Jackson C, Fedorka-Cray P, Barrett J, Brousse J, Gustafson J, Kucher M. Carriage
of methicillin-resistant staphylococci by healthy companion animals in the US. Lett Appl
Microbiol. 2014;59:1–8.
37. Dekker JP, Frank KM. Next-generation epidemiology: using real-time core genome multilo-
cus sequence typing to support infection control policy. J Clin Microbiol. 2016;54:2850–3.
38. DeLeo FR, Otto M, Kreiswirth BN, Chambers HF. Community-associated methicillin-­
resistant Staphylococcus aureus. Lancet. 2010;375:1557–68.
39. Dhand A, Sakoulas G. Daptomycin in combination with other antibiotics in the treat-
ment of complicated methicillin-resistant Staphylococcus aureus bacteremia. Clin Ther.
2014;36:1303–16.
40. Diekema DJ, Richter SS, Heilmann KP, Dohrn CL, Riahi F, Tendolkar S, JS MD, Doern
GV. Continued emergence of USA300 methicillin-resistant Staphylococcus aureus in
the United States: results from a nationwide study. Infect Control Hosp Epidemiol.
2014;35:285–92.
41. Diep BA, Gill SR, Chang RF, Phan TH, Chen JH, Davidson MG, Lin F, Lin J, Carleton
HA, Mongodin EF, Sensabaugh GF, Perdreau-Remington F. Complete genome sequence
of USA300, an epidemic clone of community-acquired meticillin-resistant Staphylococcus
aureus. Lancet. 2006;367:731–9.
42. Dominguez TJ. It’s not a spider bite, it’s community-acquired methicillin-resistant
Staphylococcus aureus. Am Board Fam Med. 2004;17:220–6.
43. Duijekeren EV, Wolfhagen MJ, Box AT, Heck ME, Wannet WJ, Fluit AC. Human-to-dog trans-
mission of methicillin-resistant Staphylococcus aureus. Emerg Infect Dis. 2004;10:2235–7.
44. Duijekeren E, Wolfhagen MJ, Heck ME, Wannet WJ. Transmission of a panton-valentine
leucocidin-positive, methicillin-resistant Staphylococcus aureus strain between humans and
a dog. J Clin Microbiol. 2005;43:6209–11.
45. Duong M, Markwell S, Peter J, Barenkamp S. Randomized, controlled trial of antibiotics
in the Management of Community-Acquired Skin Abscesses in the pediatric patient. Ann
Emerg Med. 2010;55:401–7.
46. Ekkelenkamp MB, Sekkat M, Carpaij N, Troelstra A, Bonten MJ. Endocarditis due to
methicillin-resistant Staphylococcus aureus originating from pigs. Ned Tijdschr Geneeskd.
2006;150:2442–7.
62 L. P. Gleason et al.

47. Ellis MW, Hospenthal DR, Dooley DP, Gray PJ, Murray CK. Natural history of community-­
acquired methicillin-resistant Staphylococcus aureus colonization and infection in soldiers.
Clin Infect Dis. 2004;39:971–9.
48. Estivariz CF, Park SY, Hageman JC, Dvorin J, Melish MM, Arpon R, Coon P, Slavish S,
Kim M, McDougal LK, Jensen B, McAllister S, Lonsway D, Killgore G, Effler PE, Jernigan
DB. Emergence of community-associated methicillin-resistant Staphylococcus aureus in
Hawaii, 2001-2003. J Infect. 2007;54:349–57.
49. Faires MC, Tater KC, Weese JS. An investigation of methicillin-resistant Staphylococcus
aureus colonization in people and pets in the same household with an infected person or
infected pet. J Am Vet Med Assoc. 2009;235:540–3.
50. Fekety FR, Bennett IL Jr. The epidemiological virulence of staphylococci. Yale J Biol Med.
1959;32:23–32.
51. Ferreira JP, Anderson KL, Correa MT, Lyman R, Ruffin F, Reller LB, Fowler VG. Transmission
of MRSA between companion animals and infected human patients presenting to outpatient
medical care facilities. PLoS One. 2011;6:e26978.
52. Fiebelkorn KR, Crawford SA, McElmeel ML, Jorgensen JH. Practical disk diffusion method
for detection of inducible clindamycin resistance in Staphylococcus aureus and coagulase-­
negative staphylococci. J Clin Microbiol. 2003;41:4740–4.
53. Fleisch F, Zbinden R, Vanoli C, Ruef C. Epidemic spread of a single clone of methicillin-­
resistant Staphylococcus aureus among injection drug users in Zurich, Switzerland. Clin
Infect Dis. 2001;32:581–6.
54. Frank AL, Marcinak JF, Mangat PD, Tjhio JT, Kelkar S, Schreckenberger PC, Quinn
JP. Clindamycin treatment of methicillin-resistant Staphylococcus aureus infections in chil-
dren. Pediatr Infect Dis J. 2002;21:530–4.
55. Freitas EAF, Harris RM, Blake RK, Salgado CD. Prevalence of USA300 strain type of
methicillin-­resistant Staphylococcus aureus among patients with nasal colonization identi-
fied with active surveillance. Infect Control Hosp Epidemiol. 2010;31:469–75.
56. Fridkin SK, Hageman JC, Morrison M, Sanza LT, Como-Sabetti K, Jernigan JA, Harriman K,
Harrison LH, Lynfield R, Farley MM. Methicillin-resistant Staphylococcus aureus disease in
three communities. N Engl J Med. 2005;352:1436–44.
57. Fritz SA, Hogan PG, Hayek G, Eisenstein KA, Rodriguez M, Epplin EK, Garbutt J, Fraser
VJ. Household versus individual approaches to eradication of community-associated
Staphylococcus aureus in children: a randomized trial. Clin Infect Dis. 2011;54:743–51.
58. Fritz SA, Hogan PG, Singh LN, Thompson RM, Wallace MA, Whitney K, Al-Zubeidi D,
Burnham CD, Fraser VJ. Contamination of environmental services with Staphylococcus
aureus in households with children infected with methicillin-resistant S aureus. JAMA Ped.
2014;168:1030–8.
59. Giersing BK, Dastgheyb SS, Modjarrad K, Moorthy V. Status of vaccine research and devel-
opment of vaccines for Staphylococcus aureus. Vaccine. 2016;34:2962–6.
60. Gingrich EN, Kurt T, Hyatt DR, Lappin MR, Ruch-Gallie R. Prevalence of methicillin-­resistant
staphylococci in northern Colorado shelter animals. J Vet Diagn Investig. 2011;23:947–50.
61. Glaser P, Martins-Simoes P, Villain A, Barbier M, Tristan A, Bouchier C, Ma L, Bes M,
Laurent F, Guillemot D, Wirth T, Vandenesch F. Demography and Intercontinental spread of
the USA300 community acquired methicillin-resistant Staphylococcus aureus lineage. mBio.
2016;7:e02183–15.
62. Gorwitz RJ, Kruszon-Moran D, McAllister SK, McQuillan G, McDougal LK, Fosheim
GE, Jensen BJ, Killgore G, Tenover FC, Kuehnert MJ. Changes in the prevalence of nasal
colonization with Staphylococcus aureus in the United States, 2001-2004. J Infect Dis.
2008;197:1226–34.
63. Groom AV, Wolsey DH, NaimI TS, Smith K, Johnson S, Boxrud D, Moore KA, Cheek
JE. Community-acquired methicillin-resistant Staphylococcus aureus in a rural American
Indian community. JAMA. 2001;286:1201–5.
3 Emergence of MRSA in the Community 63

64. Grumann D, Nübel U, Bröker BM. Staphylococcus aureus Toxins--their functions and genet-
ics. Infect Genet Evol. 2014;21:583–92.
65. Gualandi NR, Mu Y, Dumyati G, Harrison L, Lesher L, Nadle L, Petit S, Ray SM, Schaffner W,
Townes J, Fridkin S, See I. Racial disparities in invasive methicillin-resistant Staphylococcus
aureus infections, 2005–2012. 2015 Council of State and Territorial Epidemiologists Annual
Conference, Boston, MA. Presented June 15, 2015. Abstract #4203; 2015.
66. Hageman JC, UyekI TM, Francis JS, Jernigan DB, Wheeler JG, Bridges CB, Barenkamp SJ,
Sievert DM, Srinivasan A, Doherty MC, McDougal LK, Killgore GE, Lopatin UA, Coffman
R, MacDonald JK, McAllister SK, Fosheim GE, Patel JB, McDonald C. Severe community-­
acquired pneumonia caused by an emerging strain of Staphylococcus aureus during the 2003-­
04 influenza season. Emerg Infect Dis. 2006;12:894–9.
67. Ham DC, Jackson K, Bamberg W, Dumyati G, Harrison L, Lynfield R, Nadle J, Petit S, Ray
SM, Schaffner W, Townes J, Kallen AJ, See I. Community associated invasive methicillin-­
resistant Staphylococcus aureus (MRSA), 2005–2014. Spring 2017 Conference for the
Society for Healthcare Epidemiology of America, St Louis, MO. Presented March 31, 2017.
Abstract #8707; 2017.
68. Hatcher SM, Rhodes SM, Stewart JR, Silbergeld E, Pisanic N, Larsen J, Jiang S, Krosche A,
Hall D, Carroll KC, Heaney CD. The prevalence of antibiotic-resistant Staphylococcus aureus
nasal carriage among industrial hog operation workers, community residents, and children
living in their households: North Carolina, USA. Environ Health Perspect. 2017;125:560–9.
69. Heller J, Innocent G, Denwood M, Reid S, Kelly L, Mellor D. Assessing the probability of
acquisition of meticillin-resistant Staphylococcus aureus (MRSA) in a dog using a nested
stochastic simulation model and logistic regression sensitivity analysis. Prev Vet Med.
2011;99:211–24.
70. Heller J, Kelly L, Reid SW, Mellor DJ. Qualitative risk assessment of the acquisition of
Meticillin-resistant Staphylococcus aureus in pet dogs. Risk Anal. 2010;30:458–72.
71. Herold BC, lmmergluck LC, Maranan MC, Lauderdale DS, Gaskin RE, BoyleVavra S, Leitch
CD, Daum RS. Community-acquired methicillin-resistant Staphylococcus aureus in children
with no identified predisposing risk. JAMA. 1998;279:593–8.
72. Hill PC, Birch M, Chambers S, Drinkovic D, Ellis-Pegler RB, Everts R, Murdoch D,
Pottumarthy S, Roberts SA, Swager C, Taylor SL, Thomas MG, Wong CG, Morris
AJ. Prospective study of 424 cases of Staphylococcus aureus bacteraemia: determination of
factors affecting incidence and mortality. Intern Med J. 2001;31:97–103.
73. Hiramatsu K, Hanaki H, Ino T, Yabuta K, Oguri T, Tenover FC. Methicillinresistant
Staphylococcus aureus clinical strain with reduced vancomycin susceptibility. J Antimicrob
Chemother. 1997;40:135–6.
74. Hota B, Ellenbogen C, Hayden MK, Aroutcheva A, Rice TW, Weinstein RA. Community
associated methicillin-resistant Staphylococcus aureus skin and soft tissue infections at a
public hospital: do public housing and incarceration amplify transmission? Arch Int Med.
2007;167:1026–33.
75. Huang SS, Septimus E, Kleinman K, Moody J, Hickok J, Avery TR, Lankiewicz J, Bombosev
A, Terpstra L, Hartford F, Hayden MK, Jernigan JJ, Weinstein RA, Fraser VJ, Haffenreffer K,
Cui E, Kaganov RE, Lolans K, Perline JB, Platt R. Targeted versus universal decolonization
to prevent ICU infection. N Engl J Med. 2013;368:2255–65.
76. Humphreys H, Fitzpatrick F, Harvey BJ. Gender differences in rates of carriage and blood-
stream infection caused by methicillin-resistant Staphylococcus aureus: are they real, do they
matter and why? Clin Infect Dis. 2015;61:1708–14.
77. Hwang JH, Tsai HY, Liu TC. Community-acquired methicillin-resistant Staphylococcus
aureus infections in discharging ears. Acta Otolaryngol. 2002;122:827–30.
78. Iwamoto M, Mu Y, Lynfield R, Bulens SN, Nadle J, Aragon D, Petit S, Ray SM, Harrison LH,
Dumyati G, Townes JM, Schaffner W, Gorwitz RJ, Lessa FC. Trends in invasive methicillin-­
resistant Staphylococcus aureus infections. Pediatrics. 2013;132:e817.
64 L. P. Gleason et al.

79. Jessen O, Rosendal K, Bulow P, Faber V, Eriksen KR. Changing staphylococci and staphy-
lococcal infections. A ten-year study of bacteria and cases of bacteremia. N Engl J Med.
1969;281:627–35.
80. Jevons MP. “Celbenin”-resistance staphylococci. BMJ. 1961;1:124.
81. Kallen AJ, Driscoll TJ, Thornton S, Olson PE, Wallace MR. Increase in community-acquired
methicillin-resistant Staphylococcus aureus at a Naval Medical Center. Infec Control Hosp
Epidemiol. 2000;21:223–6.
82. Kaplan SL. Implications of methicillin-resistant Staphylococcus aureus as a communityac-
quired pathogen in pediatric patients. lnfect Dis Clin North Am. 2005;19:747–57.
83. Kaplan SL, Forbes A, Hammerman WA, Lamberth L, Hulten KG, Minard CG, Mason
EO. Randomized trial of “bleach baths” plus routine hygienic measures vs routine hygienic
measures alone for prevention of recurrent infections. Clin Infect Dis. 2013;58:679–82.
84. Katayama Y, Ito T, Hiramatsu K. A new class of genetic element, staphylococcus cassette
chromosome mec, encodes methicillin-resistance in Staphylococcus aureus. Antimicrob
Agents Chemother. 2000;44:1549–55.
85. Kazakova SV, Hageman JC, Matava M, Srinivasan A, Phelan L, Garfinkel B, Boo T,
McAllister S, Anderson J, Jensen B, Dodson D, Lonsway D, McDougal LK, Arduino M,
Fraser VJ, Killgore G, Tenover FC, Cody S, Jernigan DB. A clone of methicillin-resistant
Staphylococcus aureus among professional football players. N Engl J Med. 2005;352:468–75.
86. Kazmierczak Z, Gorski A, Dabrowska K. Facing antibiotic resistance: Staphylococcus aureus
phages as a medical tool. Virus. 2014;6:2551–70.
87. Kirby WMM. Extraction of a highly potent penicillin inactivator from penicillin resistant
staphylococci. Science. 1944;99:452–3.
88. Klevens RM, Morrison MA, Nadle J, Petit S, Gershman K, Ray S, Harrison LH, Lynfield
R, Dumyati G, Townes JM, Craig AS, Zell ER, Fosheim GE, McDougal LK, Carey RB,
Fridkin SK. Invasive methicillin-resistant Staphylococcus aureus infections in the United
States. JAMA. 2007;298:1763–71.
89. Köser CU, Holden MT, Ellington MJ, Cartwright EJ, Brown NM, Ogilvy-Stuart AL, Hsu
LY, Chewapreecha C, Croucher NJ, Harris SR, Sanders M, Enright MC, Dougan G, Bentley
SD, Parkhill J, Fraser LJ, Betley JR, Schulz-Trieglaff OB, Smith GP, Peacock SJ. Rapid
whole-genome sequencing for investigation of a neonatal MRSA outbreak. N Engl J Med.
2012;366:2267–75.
90. Laupland KB, Conly JM. Treatment of Staphylococcus aureus colonization and prophylaxis
for infection with topical intranasal mupirocin: an evidence-based review. Clin lnfect Dis.
2003;37:933–8.
91. Lee NE, Taylor MM, Bancroft E, Ruane PJ, Morgan M, McCoy L, Simon PA. Risk factors
for community-associated methicillin-resistant Staphylococcus aureus skin infections among
HIV-positive men who have sex with men. Clin Infect Dis. 2005;40:1529–34.
92. Lessa FC, Mu Y, Ray SM, Dumyati G, Bulens S, Gorwitz RJ, Fosheim G, AS DV, Schaffner
W, Nadle J, Gershman K, Fridkin SK, Active Bacterial Core surveillance (ABCs); MRSA
Investigators of the Emerging Infections Program. Impact of USA300 methicillin-resistant
Staphylococcus aureus on clinical outcomes of patients with pneumonia or central line asso-
ciated bloodstream infections. Clin Infect Dis. 2012;55:232–41.
93. Levine DP, Cushing RD, Jui J, Brown WJ. Community-acquired methicillin-resistant
Staphylococcus aureus endocarditis in the Detroit Medical Center. Ann Intern Med.
1982;97:330–8.
94. Limbago B, Fosheim GE, Schoonover V, Crane CE, Nadle J, Petit S, Heltzel D, Ray SM,
Harrison LH, Lynfield R, Dumyati G, Townes JM, Schaffner W, Mu Y, Fridkin SK, Active
Bacterial Core Surveillance MRSA Investigators. Characterization of methicillin-resistant
Staphylococcus aureus isolates collected in 2005 and 2006 from patients with invasive dis-
ease: a population-based analysis. J Clin Microbiol. 2009;47:1344–51.
95. Limbago BM, Kallen AJ, Zhu W, Eggers P, McDougal LK, Albrecht VS. Report of the
13th vancomycin-resistant Staphylococcus aureus isolates from the United States. J Clin
Microbiol. 2014;52:998–1002.
3 Emergence of MRSA in the Community 65

96. Lina G, Piemont Y, Godail-Gamot F, Bes M, Peter MO, Gauduchon V, Vandenesch F, Etienne
J. Involvement of Panton-Valentine leukocidin-producing Staphylococcus aureus in primary
skin infections and pneumonia. Clin lnfect Dis. 1999;29:1128–32.
97. Liu C, Bayer A, Cosgrove SE, Daum RS, Fridkin SK, Gorwitz RJ, Kaplan SL, Karchmer
AW, Levine DP, Murray BE, J Rybak M, Talan DA, Chambers HF. Clinical practice guide-
lines by the infectious diseases society of America for the treatment of methicillin-resistant
Staphylococcus aureus infections in adults and children. Clin Infect Dis. 2011;52:e18–55.
98. Loeb M, Main C, Walker-Dilks C, Eady A. Antimicrobial drugs for treating methicillin-­resistant
Staphylococcus aureus colonization. Cochrane Database Syst Rev. 2003;(4):CD003340.
99. Mammina C, Calà C, Plano MR, Bonura C, Vella A, Monastero R, Palma DM. Ventilator-­
associated pneumonia and MRSA ST398, Italy. Emerg Infect Dis. 2010;16:730–1.
100. Mancini F, Monaci E, Lofano G, Torre A, Bacconi M, Tavarini S, Sammicheli C, Arcidiacono
L, Galletti B, Laera D, Pallaoro M, Tuscano G, Fontana MR, Bensi G, Grandi G, Rossi-­
Paccani S, Nuti S, Rappuoli R, De Gregorio E, Bagnoli F, Soldaini E, Bertholet S. One dose
of Staphylococcus aureus 4C-staph vaccine formulated with a novel TLR7-dependent adju-
vant rapidly protects mice through antibodies, effector CD4+ T cells, and IL-17A. PLoS One.
2016;11:e0147767.
101. Manian FA. Asymptomatic nasal carriage of mupirocin-resistant, methicillin-resistant
Staphylococcus aureus (MRSA) in a pet dog associated with MRSA infection in household
contacts. Clin Infect Dis. 2003;36:e26–8.
102. Martinez-Aguilar G, Avalos-Mishaan A, Hulten K, Hammerman W, Mason EO Jr, Kaplan
SL. Community-acquired, methicillin-resistant and methicillin-susceptible Staphylococcus
aureus musculoskeletal infections in children. Pediatr lnfect Dis J. 2004;23:701–6.
103. McDougal LK, Steward CD, Killgore GE, Chaitram JM, McAllister SK, Tenover FC. Pulsed-­
field gel electrophoresis typing of oxacillin-resistant Staphylococcus aureus isolates from the
United States: establishing a national database. J Clin Microbiol. 2003;41:5113–20.
104. Miller LG, Perdreau-Remington F, Rieg G, Mehdi S, Perlroth J, Bayer AS, Tang AW, Phung
TO, Spellberg B. Necrotizing fasciitis caused by communityassociated methicillin-resistant
Staphylococcus aureus in Los Angeles. N Engl J Med. 2005;352:1445–53.
105. Miller LG, Eells SJ, Taylor AR, David MZ, Ortiz N, Zychowski D, Kumar N, Cruz D,
Boyle-Vavra S, Daum RS. Staphylococcus aureus Colonization among household contacts
of patients with skin infections: risk factors, strain discordance, and complex ecology. Clin
Infect Dis. 2012;54:1523–35.
106. Miller MB, Weber DJ, Goodrich JS, Popowitch EB, Poe MD, Nguyen V, Shope TR,
Foster DT, Miller JR, Kotch J. Prevalence and risk factor analysis for methicillin-resistant
Staphylococcus aureus nasal colonization in children attending child care centers. J Clin
Microbiol. 2011;49:1041–7.
107. Moran GJ, Amii RN, Abrahamian FM, Talan DA. Methicillin-resistant Staphylococcus
aureus in community-acquired skin infections. Emerg Infect Dis. 2005;11:928–30.
108. Moran GJ, Krishnadasan A, Gorwitz RJ, Fosheim GE, Albrecht V, Limbago B, Talan DA,
EMERGEncy ID NET Study Group. Prevalence of methicillin-resistant Staphylococcus aureus
as an etiology of community-acquired pneumonia. Clin Infect Dis. 2012;54(8):1126–33.
109. Moran GJ, Krishnadasan A, Mower WR, Abrahamian FM, LoVecchio F, Steele MT, Rothman
RE, Karras DJ, Hoagland R, Pettibone S, Talan DA. Effect of cephalexin plus trimethoprim-­
sulfamethoxazole vs. cephalexin alone on clinical cure of uncomplicated cellulitis: a random-
ized clinical trial. JAMA. 2017;317:2088–96.
110. Morgan DJ, Murthy R, Munoz-Price LS, Barnden M, Camins BC, Johnston BL, Rubin Z,
Sullivan KV, Shane AL, Dellinger EP, Rupp ME, Bearman G. Reconsidering contact pre-
cautions for endemic methicillin-resistant Staphylococcus aureus and vancomycin-resistant
Enterococcus. Infect Control Hosp Epidemiol. 2015;36:1163–72.
111. Morris DO, Lautenbach E, Zaoutis T, Leckerman K, Edelstein PH, Rankin SC. Potential
for pet animals to harbour methicillin-resistant Staphylococcus aureus when residing with
human MRSA patients. Zoonoses Public Health. 2012;59:286–93.
66 L. P. Gleason et al.

112. Naimi TS, Ledell KH, Boxrud DJ, Groom AV, Steward CD, Johnson SK, Besser JM, O'Boyle
C, Danila RN, Cheek JE, Osterholm MT, Moore KA, Smith KE. Epidemiology and clonal-
ity of community-acquired methicillin-resistant Staphylococcus aureus in Minnesota, 1996-­
1998. Clin Infect Dis. 2001;33:990–6.
113. Nair R, Ammann E, Rysavy M, Schweizer ML. Mortality among patients with methicillin-­
resistant Staphylococcus aureus USA300 versus non-USA300 invasive infections: a meta-­
analysis. Infect Control Hosp Epidemiol. 2014;35:31–41.
114. Nerby JM, Gorwitz R, Lesher L, Juni B, Jawahir S, Lynfield R, Harriman K. Risk factors
for household transmission of community associated methicillin-resistant Staphylococcus
aureus. Pediatr Infect Dis J. 2011;30:927–32.
115. Neyra RC, Frisancho JA, Rinsky JL, Resnick C, Carroll KC, Rule AM, Ross T, You Y, Price
LB, Silbergeld EK. Multidrug-resistant and methicillin-resistant Staphylococcus aureus
(MRSA) in hog slaughter and processing plant workers and their community in North
Carolina (USA). Environ Health Perspect. 2014;122:471–7.
116. Ng W, Faheem A, McGeer A, Simor AE, Gelosia A, Willey BM, Watt C, Richardson
DC, Wong H, Ostrowska K, Vernich L, Muller MP, Gnanasuntharam P, Porter V, Katz
K. Community and healthcare associated methicillin-resistant Staphylococcus aureus strains:
an investigation into household transmission, risk factors, and environmental contamination.
Infect Control Hosp Epidemiol. 2017;38:61–7.
117. Nguyen DM, Mascola L, Brancoft E. Recurring methicillin-resistant Staphylococcus aureus
infections in a football team. Emerg Infect Dis. 2005;11:526–32.
118. Nimmo GR. USA300 abroad: global spread of a virulent strain of community associated
methicillin-resistant Staphylococcus aureus. Clin Microbiol Infect. 2012;18:725–34.
119. Ogston A. Micrococcus poisoning. J Anat. 1882;17:24–58.
120. O’Hara FP, Suaya JA, Ray GT, Baxter R, Brown ML, Mera RM, Close NM, Thomas E,
Amrine-Madsen H. Spa typing and multilocus sequence typing show comparable perfor-
mance in a macroepidemiologic study of Staphylococcus aureus in the United States. Microb
Drug Resist. 2016;22:88–96.
121. Panton PN, Valentine FC. Staphylococcal toxin. Lancet. 1932;222:506–8.
122. Perl TM, Cullen JJ, Wenzel RP, Zimmerman MB, Pfaller MA, Sheppard D, Twombley J,
French PP, Herwaldt LA. Intranasal mupirocin to prevent postoperative Staphylococcus
aureus infections. N Engl J Med. 2002;346:1871–7.
123. Planet PJ, LaRussa SJ, Dana A, Smith H, Xu A, Ryan C, Uhlemann AC, Boundy S, Goldberg J,
Narechania A, Kulkarni R, Ratner AJ, Geoghegan JA, Kolokotronis SO, Prince A. Emergence
of the epidemic of methicillin-resistant Staphylococcus aureus strain USA300 coincides with
horizontal transfer of the arginine catabolic mobile element and speG-­mediated adaptations
for survival on skin. mBio. 2013;4:e00889–13.
124. Planet PJ, Diaz L, Kolokotronis S, Narechania A, Reyes J, Xing G, Rincon S, Smith H,
Panesso D, Ryan C, Smith DP, Guzman M, Zurita J, Sebra R, Deikus G, Nolan RL, Tenover
FC, Weinstock GM, Robinson DA, Arias CA. Parallel epidemics of community associated
methicillin-resistant Staphylococcus aureus USA300 infection in North and South America.
J Infect Dis. 2015;212:1874–82.
125. Podewils LJ, Liedtke LA, McDonald LC, Hageman JC, Strausbaugh LJ, Fischer TK, Jernigan
DB, Uyeki TM, Kuehnert MJ. A national survey of severe influenza-associated complications
among children and adults, 2003-2004. Clin Infect Dis. 2005;40:1693–6.
126. Popovich KJ, Hota B, Aroutcheva A, Kurien L, Patel J, Lyles-Banks R, Grasso AE, Spec
A, Beavis KG, Hayden MK, Weinstein RA. Community associated methicillin-resistant
Staphylococcus aureus colonization burden and HIV-infected patients. Clin Infect Dis.
2013;56:1067–74.
127. Popovich KJ, Snitkin ES, Hota B, Green SJ, Pirani A, Aroutcheva A, Weinstein RA. Genomic
and epidemiological evidence for community origins of hospital-onset methicillin-resistant
Staphylococcus aureus bloodstream infections. J Infect Dis. 2017;215:1640–7.
128. Raczniak GA, Gaines J, Bulkow LR, Kinzer MH, Hennessy TW, Klejka JA, Bruce MG. A
survey of knowledge, attitudes, and practices towards skin and soft-tissue infections in rural
Alaska. Int J Circumpolar Health. 2016;75:30603.
3 Emergence of MRSA in the Community 67

129. Ray GT, Suaya JA, Baxter R. Incidence, microbiology, and patient characteristics of skin and
soft-tissue infections in a US population: a retrospective population-based study. BMC Infect
Dis. 2013;13:252.
130. Rhee Y, Aroutcheva A, Hota B, Weinstein RA, Popovich KJ. Evolving epidemiology of
Staphylococcus aureus bacteremia. Infect Control Hosp Epidemiol. 2015;36:1417–22.
131. Rolo J, Miragaia M, Turlej-Rogacka A, Empel J, Bouchami O, Faria NA, Tavares A,
Hryniewicz W, Fluit AC, de Lencastre H, The CONCORD Working Group. High genetic
diversity among community associated Staphylococcus aureus in Europe: results from a mul-
ticenter study. PLoS One. 2012;7:e34768.
132. Ross S, Rodriguez W, Controni G, Khan W. Staphylococcal susceptibility to penicillin G. The
changing pattern among community strains. JAMA. 1974;229:1075–7.
133. Rossi F, Diaz L, Wollam A, Panesso D, Zhou Y, Rincon S, Narechania A, Xing G, Di Gioia
TS, Doi A, Tran TT, Reyes J, Munita JM, Carvajal LP, Hernandez-Roldan A, Brandão D, van
der Heijden IM, Murray BE, Planet PJ, Weinstock GM, Arias CA. Transferable vancomycin
resistance in a community associated MRSA lineage. N Engl J Med. 2014;370:1524–31.
134. Santos F, Mankarious LA, Eavey RD. Methicillin-resistant Staphylococcus aureus: pediatric
otitis. Arch Otolaryngol Head Neck Surg. 2000;126:1383–5.
135. Saravolatz LD, Markowitz N, Arking L, Pohlod D, Fisher E. Methicillin-resistant
Staphylococcus aureus. Epidemiologic observations during a community-acquired outbreak.
Ann lntern Med. 1982;96:11–6.
136. Sattler CA, Mason EO Jr, Kaplan SL. Prospective comparison of risk factors and demographic
and clinical characteristics of community-acquired, methicillin-resistant versus methicillin-­
susceptible Staphylococcus aureus infection in children. Pediatr lnfect Dis J. 2002;21:910–7.
137. Schinasi L, Wing S, Augustino KL, Ramsey KM, Nobles DL, Richardson DB, Price LB,
Aziz M, PD MD, Stewart JR. A case control study of environmental and occupational expo-
sures associated with methicillin resistant Staphylococcus aureus nasal carriage in patients
admitted to a rural tertiary care hospital in a high density swine region. Environ Health.
2014;13:54.
138. Schmidt CS, White CJ, Ibrahim AS, Filler SG, Yeaman MR, Edwards JE Jr, Hennessey JP Jr.
NDV-3, a recombinant alum-adjuvanted vaccine for Candida and Staphylococcus aureus, is
safe and immunogenic in healthy adults. Vaccine. 2012;30:7594–600.
139. Schmitz GR, Bruner D, Pitotti R, Olderog C, Livengood T, Williams J, Huebner K,
Lightfoot J, Ritz B, Bates C, Schmitz M, Mete M, Deye G. Randomized controlled trial
of trimethoprim-­sulfamethoxazole for uncomplicated skin abscesses in patients at risk for
community-­associated methicillin-resistant Staphylococcus aureus infection. Ann Emerg
Med. 2010;56:283–7.
140. See I, Albrecht V, Mu Y, Dumyati G, Koeck M, Lynfield R, Nadle J, Ray SM, Schaffner
W, Limbago B, Kallen A. Changes in incidence and strains of methicillin-resistant
Staphylococcus aureus bloodstream infections, 2005–2013. IDweek 2016 conference, New
Orleans, LA. Presented October 29, 2016. Abstract #1750; 2016.
141. See I, Wesson P, Gualandi N, Dumyati G, Harrison LH, Lesher L, Nadle J, Petit S, Reisenauer
C, Schaffner W, Tunali A, Mu Y, Ahern J. Socioeconomic factors explain racial disparities
in invasive community-associated methicillin-resistant Staphylococcus aureus disease rates.
Clin Infect Dis. 2017;64:597–604.
142. Sehulster L, Chinn RY. Guidelines for environmental infection control in health-care facili-
ties. Recommendations of CDC and the healthcare infection control practices advisory com-
mittee (HICPAC). MMWR Recomm Rep. 2003;52:1–42.
143. Self WH, Wunderink RG, Williams DJ, Zhu Y, Anderson EJ, Balk RA, Fakhran SS, Chappell
JD, Casimir G, Courtney DM, Trabue C, Waterer GW, Bramley A, Magill S, Jain S, Edwards
KM, Grijalva CG. Staphylococcus aureus Community-acquired pneumonia: prevalence,
clinical characteristics, and outcomes. Clin Infect Dis. 2016;63(3):300–9.
144. Shallcross LJ, Fragaszy E, Johnson AM, Hayward AC. The role of the Panton-Valentine leu-
cocidin toxin in staphylococcal disease: a systematic review and meta-analysis. Lancet Infec
Dis. 2013;13:43–54.
68 L. P. Gleason et al.

145. Siberry GK, Tekle T, Carroll K, Dick J. Failure of clindamycin treatment of methicillin-­
resistant Staphylococcus aureus expressing inducible clindamycin resistance in vitro. Clin
Infect Dis. 2003;37:1257–60.
146. Siegel JD, Rhinehart E, Jackson M, Chiarello L, Healthcare Infection Control Practices
Advisory Committee. Management of multidrug-resistant organisms in healthcare settings,
2006. Available at: https://www.cdc.gov/infectioncontrol/guidelines/mdro/. 2006.
147. Siegel JD, Rhinehart E, Jackson M, Chiarello L, Healthcare Infection Control Practices
Advisory Committee. Guideline for isolation precautions: preventing transmission of infec-
tious agents in healthcare settings 2007. Available at http://www.cdc.gov/ncidod/dhqp/pdf/
guidelines/Isolation2007.pdf. 2007.
148. Singer AJ, Talan DA. Management of Skin Abscesses in the era of methicillin-resistant
Staphylococcus aureus. N Engl J Med. 2014;370:1039–47.
149. Stacey AR, Endersby KE, Chan PC, Marples RR. An outbreak of methicillin resistant
Staphylococcus aureus infection in a rugby football team. Br J Sports Med. 1998;32:153–4.
150. Stevens DL, Bisno AL, Chambers HF, Everett ED, Dellinger P, Goldstein EJ, Gorbach SL,
Hirschmann JV, Kaplan EL, Montoya JG, Wade JC. Practice guidelines for the diagnosis and
management of skin and soft-tissue infections. Clin Infect Dis. 2005;41:1373–406.
151. Talan DA, Krishnadasan A, Gorwitz RJ, Fosheim GE, Limbago B, Albrecht V, Moran GJ,
EMERGEncy ID Net Study Group. Comparison of Staphylococcus aureus from skin and
soft-tissue infections in US emergency department patients, 2004 and 2008. Clin Infect Dis.
2011;15:144–9.
152. Talan DA, Mower WR, Krishnadasan A, Abrahamian FM, Lovecchio F, Karras DJ, Steele
MT, Rothman RE, Hoagland R, Moran GJ. Trimethoprim–Sulfamethoxazole versus placebo
for uncomplicated skin abscess. N Engl J Med. 2016;374:823–32.
153. Toleman MS, Reuter S, Coll F, Harrison EM, Blane B, Brown NM, Török ME, Parkhill
J, Peacock SJ. Systematic surveillance detects multiple silent introductions and household
transmission of methicillin-resistant Staphylococcus aureus USA300 in the east of England.
J Infect Dis. 2016;214:447–53.
154. Tong SYC, Nelson J, Paterson DL, Fowler VG, Howden BP, Cheng AC, Chatfield M, Lipman
J, Van Hal S, O’Sullivan M, Robinson JO, Yahav D, Lye D, Davis JS, CAMERA2 study
group and the Australasian Society for Infectious Diseases Clinical Research Network.
CAMERA2 – combination antibiotic therapy for methicillin-resistant Staphylococcus aureus
infection: study protocol for a randomized controlled trial. Trials. 2016;17:170.
155. Uhlemann AC, Dordel J, Knox JR, Raven KE, Parkhill J, Holden MTG, Peacock SJ, Lowy
FD. Molecular tracing of the emergence, diversification, and the transmission of S. aureus
sequence type 8 in a New York community. Proc Natl Acad Sci. 2014;111:6739–43.
156. Uhlemann AC, Otto M, Lowy FD, DeLeo FR. Evolution of community- and healthcare-­
associated methicillin-resistant Staphylococcus aureus. Infect Genet Evol. 2014;21:563–74.
157. van Cleef BA, Verkade EJ, Wulf MW, Buiting AG, Voss A, Huijsdens XW, van Pelt W,
Mulders MN, Kluytmans JA. Prevalence of livestock-associated MRSA in communities with
high pig-densities in the Netherlands. PLoS One. 2010;5:e9385.
158. VandenBergh MF, Yzerman EP, Van Belkum A, Boelens HA, Sijmons M, Verbrugh
HA. Follow-up of Staphylococcus aureus nasal carriage after 8 years: redefining the persis-
tent carrier state. J Clin Microbiol. 1999;37:3133–40.
159. Vayalumkal JV, Suh KN, Toye B, Ramotar K, Saginur R, Roth VR. Skin and soft tissue infec-
tions caused by methicillin-resistant Staphylococcus aureus (MRSA): an affliction of the
underclass. CJEM. 2012;14:335–43.
160. Vetter RS, Bush SP. Reports of presumptive brown recluse spider bites reinforce improb-
able diagnosis in regions of North America where the spider is not endemic. Clin Infect Dis.
2002;35:442–5.
161. von Eiff C, Becker K, Machka K, Stammer H, Peters G. Nasal carriage as a source of
Staphylococcus aureus bacteremia. Study Group. N Engl J Med. 2001;344:11–6.
3 Emergence of MRSA in the Community 69

162. Walters M, Eggers P, Albrecht V, Travis T, Lonsway D, Hovan G, Taylor D, Rasheed K,

Limbago B, Kallen A. Notes from the field: Vancomycin-resistant Staphylococcus aureus —
Delaware, 2015. MMWR. 2015;64:1056.
163. Wang J, Barth S, Richardson M, Corson K, Mader J. An outbreak of methicillin resistant
Staphylococcus aureus cutaneous infection in a saturation diving facility. Undersea Hyperb
Med. 2003;30:277–84.
164. Wardyn SE, Forshey BM, Farina SA, Kates AE, Nair R, Quick MK, Wu JY, Hanson BM,
O’Malley SM, Shows HW, Heywood EM, Beane-Freeman LE, Lynch CF, Carrel M, Smith
TC. Swine farming is a risk factor for infection with and high prevalence of carriage of
multidrug-resistant Staphylococcus aureus. Clin Infect Dis. 2015;61:59–66.
165. Weese JS, Archambault M, Willey BM, Hearn P, Kreiswirth BN, Said-Salim B, McGeer A,
Likhoshvay Y, Prescott JF, Low DE. Methicillin-resistant Staphylococcus aureus in horses
and horse personnel, 2000-2002. Emerg Infect Dis. 2005;11:430–5.
166. Weigel LM, Clewell DB, Gill SR, Clark NC, McDougal LK, Flannagan SE, Kolonay JF,
Shetty J, Killgore GE, Tenover FC. Genetic analysis of a high-level vancomycin-resistant
isolate of Staphylococcus aureus. Science. 2003;302:1569–71.
167. Wejnert C, Hess KL, Hall HI, Van Handel MV, Hayes D, Fulton P, An Q, Koenig LJ, Prejean J,
Valleroy LA. Vital signs: trends in HIV diagnoses, risk behaviors, and prevention among per-
sons who inject drugs — United States. MMWR Morb Mortal Wkly Rep. 2016;65:1336–42.
168. Williams REO. Healthy carriage of Staphylococcus aureus: its prevalence and importance.
Bacteriol Rev. 1963;27:56–71.
169. Young DM, Harris HW, Charlebois ED, Chambers H, Campbell A, Perdreau-Remington
F, Lee C, MankanI M, Mackersie R, Schecter WP. An epidemic of methicillin-resistant
Staphylococcus aureus soft tissue infections among medically underserved patients. Arch
Surg. 2004;139:947–51.
170. Zinderman CE, Conner B, Malakooti MA, Lamar JE, Armstrong A, Bohnker BK. Community-­
acquired methicillin-resistant Staphylococcus aureus among military recruits. Emerg Infect
Dis. 2004;10:941–4.
Chapter 4
Resistance of Gram-negative Bacilli
to Antimicrobials

Charles R. Dean, Gianfranco De Pascale, and Bret Benton

4.1  he Expanding Problem of Multidrug-Resistant (MDR)

T
Gram-negative Bacilli

Much has transpired in the realm of antibiotic resistance in the 10 years since the first
edition of this text. This chapter began in 2007 with the line, “At the beginning of the
twenty first century, we now find ourselves experiencing a taste of what life was like
prior to the advent of the antibiotic age…,” and this reality is continuing to sink in. So
much so, in fact, that antibiotic resistance is now routinely broached in the popular
media and has the attention of government agencies and philanthropic groups and to
some extent may be prompting a return to antibiotic discovery within the pharmaceuti-
cal industry. In 2009, the Infectious Diseases Society of America (IDSA) released the
updated call to action for a coordinated effort to bring antibiotic development to the
forefront, specifically regarding the “ESKAPE” pathogens, Enterococcus faecium,
Staphylococcus aureus, Klebsiella pneumoniae, Acinetobacter baumannii,
Pseudomonas aeruginosa, and Enterobacter spp. [1]. These pathogens cause the major-
ity of US hospital infections, and resistance is a major issue. The Gram-negative bacilli
are well represented in this group and pose a very significant emerging problem, particu-
larly in the case of pan-antibiotic-­resistant A. baumannii, multidrug-resistant (MDR) P.
aeruginosa, and carbapenem-­resistant Enterobacteriaceae (CRE). More recently the
IDSA has begun “the 10x20 Initiative” (http://www.idsociety.org/10x20/). In February
of 2017, the World Health Organization established its priority list for drug-resistant
pathogens, and in the “critical” category are carbapenem-resistant A. baumannii and P.
aeruginosa and carbapenem-resistant, extended-spectrum β-lactamase (ESBL)-
producing Enterobacteriaceae (http://www.who.int/mediacentre/news/releases/2017/
bacteria-antibiotics-needed/en/). The notion of tackling antimicrobial resistance was

C. R. Dean (*) · G. De Pascale · B. Benton

Infectious Diseases, Novartis Institutes for BioMedical Research, Emeryville, CA, USA
e-mail: [email protected]

© Springer International Publishing AG, part of Springer Nature 2018 71

I. W. Fong et al. (eds.), Antimicrobial Resistance in the 21st Century, Emerging Infectious
Diseases of the 21st Century, https://doi.org/10.1007/978-3-319-78538-7_4
72 C. R. Dean et al.

also addressed by the economist Jim O’Neill, who articulated the human costs and eco-
nomic and security threats associated with a failure to act (https://amr-review.org/sites/
default/files/160518_Final%20paper_with%20cover.pdf). Discussions about how to
incentivize antibiotic discovery have followed and the establishment of research funding
through agencies such as the Biomedical Advanced Research and Development
Authority (BARDA, https://www.phe.gov/about/BARDA/Pages/default.aspx) and
Combating Antibiotic-Resistant Bacteria Biopharmaceutical Accelerator (CARB-X,
http://www.carb-x.org); Wellcome Trust and Pew Charitable Trust have also come
online to promote antibiotic discovery. These activities, although encouraging, still
underscore the challenge upon us. Therefore, an understanding of resistance in Gram-
negative pathogens is informative and will be discussed in the following sections.
This chapter addresses intrinsic resistance and mutationally or horizontally
acquired resistance mechanisms. Intrinsic (or innate) resistance varies widely
among different pathogens and is determined by the general makeup of a cell, where
the overall complement of genes and their expression levels establish a baseline
susceptibility to an antibacterial. Here we focus on two broad elements important
for intrinsic resistance, the impermeability of the Gram-negative cell envelope,
which impedes a compound’s entry into the cell to exert its effect, and energy-­
dependent active efflux which extrudes a compound back out of the cell before it
can engage its target. We begin there, since (i) these can be important hurdles to
overcome in efforts to discover new antibiotics for Gram-negative pathogens and
(ii) they can facilitate/exacerbate the emergence of mutationally or horizontally
acquired resistance. The organism-specific genetic blueprint for intrinsic resistance
provides the background within which mutations can be selected that further
decrease susceptibility to antibacterial compounds. As well, the horizontal acquisi-
tion of new genetic material is an important route of acquired resistance. The pro-
gression to resistance is often multifactorial, and several acquired mechanisms can
accumulate over time to cause clinically significant resistance and multidrug resis-
tance (MDR). In that regard, the meaning of “resistance” is context-specific. In
clinical antimicrobial susceptibility testing, resistance is based on a specific mini-
mal inhibitory concentration (MIC) of an antibiotic tested under standardized con-
ditions, and clinical resistance occurs if the MIC of the antibiotic is above an
established clinical resistance “breakpoint” [2]. Here, we use the term more gener-
ally to convey the idea that the mechanisms discussed will alter (increase) the level
of resistance (or decrease susceptibility), but not all resistance mechanisms will
cause the specifically defined clinical resistance (shift over the breakpoint). The first
edition of this chapter pertained mainly to antibiotics that had been in clinical use
for some time (e.g., fluoroquinolones). These sections are updated here, but two
additional aspects are now included. In 2007, tigecycline was just entering the
clinic, and we update on what has happened in the approximately 12 years it has
been in widespread clinical use (Sect. 4.2.4). Second, polymyxins were reintro-
duced into the clinic as a last line of defense against MDR Gram-negative p­ athogens.
In a relatively short time, resistance has emerged and has begun to erode the clinical
utility of these compounds, and this is discussed in Sect. 4.2.6.
4 Resistance of Gram-negative Bacilli to Antimicrobials 73

4.2 Resistance in Gram-negative Bacilli

4.2.1  he Gram-negative Cell Envelope: Efflux and Outer

T
Membrane Impermeability
4.2.1.1 Active Efflux

Bacteria have a broad range of efflux pumps that can actively extrude molecules
from the cell. Efflux pumps can serve natural physiological roles such as extrusion
of metabolites but also function in efflux of toxic molecules that enter the cells.
Efflux of toxic molecules serves to lower their intracellular accumulation to reduce
access to the intracellular target(s), thereby protecting the bacteria. The five broad
efflux pump superfamilies most important in bacteria are the ATP-binding cassette
(ABC) family, the major facilitator superfamily (MFS), the small multidrug resis-
tance (SMR) family, the multidrug and toxic compound extrusion (MATE) family,
and the resistance-nodulation-cell division (RND) family (reviewed in [3] (Fig. 4.1)).
The ABC family differs from the other families in that they derive energy to drive
active efflux from hydrolysis of ATP, whereas the other families derive energy from
the proton gradient maintained at the bacterial cytoplasmic membrane. The pump
proteins that mediate compound recognition and energy-dependent extrusion for all
families are situated in the bacterial cytoplasmic membrane. Members of all pump
families except RND pumps are found in both Gram-positive and Gram-negative
bacteria. The RND family pumps are unique to Gram-negative bacteria and have
additional components and an overall architecture necessary for efflux across the
Gram-negative outer membrane (OM) (Fig. 4.1). Depending on the context, all of
the families can contribute to resistance in Gram-negative pathogens, but non-RND
family pumps can only efflux compounds into the periplasmic space between the
cytoplasmic and OM but not to the outside of the cell. Furthermore, RND pumps are

Fig. 4.1 General architecture of efflux pump families and placement in the Gram-negative cell
envelope. The Gram-negative envelope has two membranes (the inner membrane, shown here as a
symmetrical bilayer in blue, and an outer membrane, which is asymmetrical) and has phospholipid
(blue) at the inner leaflet and lipopolysaccharide (gold) at the outer leaflet. RND family pumps
have an architecture that spans both membranes. Some compounds can enter the cells via water-­
filled porins (yellow)
74 C. R. Dean et al.

notable for their large amorphous compound-binding pockets [4, 5] which confer
the ability to recognize and extrude a very broad range of structurally unrelated
molecules. For these reasons, RND family pumps are regarded as the most signifi-
cant efflux pumps overall in Gram-negative bacteria in terms of antibacterial resis-
tance. However, it is also clear that there is cooperation between networks of pumps
of different families when their substrates overlap [6]. In those cases, the single-­
component pumps may efflux a substrate into the periplasm, and the RND pump
may then expel the compound from the periplasm to the outside of the cell. RND
pumps are tripartite structures, comprised of the inner membrane-located RND
pump component, an OM channel component (outer membrane factor (OMF)), and
a periplasmic membrane fusion protein (MFP) that links these components
(Fig. 4.1). This architecture spans the double membrane of the Gram-negative cell
to allow compound extrusion across the OM through the OMF, driven by the proton-­
motive force (PMF) at the inner membrane. RND pumps are typically named in the
order MFP-pump-OMF, and the best studied RND pumps are AcrA-AcrB-TolC
(shortened to AcrAB-TolC) of E. coli and MexAB-OprM of P. aeruginosa. RND
family pumps have been found in all Gram-negative bacteria so far studied, and
most RND pumps have a broad substrate range, allowing them overall to accom-
modate most classes of antibiotics, biocides, dyes, organic solvents, detergents, bile
salts, β-lactamase inhibitors, and other molecules [3]. Moreover, some bacteria pos-
sess several different RND pumps with partially overlapping substrate specificities,
increasing their ability to deal with toxic compounds (Table 4.1). The complement
of efflux pumps in a particular Gram-negative species likely reflects the variability
of its environment. For example, the ubiquitous environmental organism P. aerugi-
nosa has a large and highly regulated genome that encodes 12 different putative
RND family efflux pumps [7, 8], presumably enhancing survival in the presence of
toxic molecules, including natural product antibacterials encountered in the envi-
ronment. In contrast, Haemophilus influenzae, which is adapted mainly to the
human respiratory tract, has only one RND pump.

4.2.1.2 Mechanism of Efflux by RND Family Pumps

Significant advancements have been made in the understanding of RND pump

assembly and function in recent years. The pump proteins AcrB in E. coli and MexB
in P. aeruginosa organize as a trimeric structure in the cytoplasmic membrane with
each protein having an extension into the periplasm made up of a porter and funnel
domain. MFP components are anchored in the inner membrane by a palmitate acyl
chain and have four domains: membrane proximal, β-barrel, lipoyl, and α-helical.
The MFS protein AcrA was shown to organize as a hexamer. Finally the OMF is
organized as a trimer in the OM with large domains extending into the periplasm.
Interaction between AcrA and AcrB and AcrA and TolC has been demonstrated
in vitro, consistent with AcrA acting as a linker between the AcrB pump and the
TolC OMF. TolC assumes a closed shape when not partnered with AcrA, and the
interaction of TolC with the α-helical hairpins of AcrA is thought to mediate the
4 Resistance of Gram-negative Bacilli to Antimicrobials 75

Table 4.1 Example RND efflux pumps in Gram-negative pathogens and range of antibiotics
accommodated by each pump
Pump component
Organism MFP RND OMF Antibiotics pumped
A. baumannii AdeA AdeB AdeC AG, CM, FQ, TC (MC), TG
AdeIa AdeJa AdeKa BL, CM, EM, FQ, TC (MC), TG
AdeF AdeG AdeH FQ, TG
B. cepacia CeoA CeoB OpcM CM, FQ, TM
E. coli AcrAa AcrBa TolCa BL, CM, FQ, ML, NO, RF
AcrA AcrD Tolc AG, FU, NO
acrE Acrf TolC FQ
H. influenzae AcrAa AcrBa TolCa EM, NO
K. pneumoniae AcrA AcrB TolC BL, CM, EM, FQ, TG
OqxA OqxB TG
KpgA KpgB KpgC TG
P. aeruginosa MexAa MexBa OprMa AG, BL, CM, ML, NO, TC, TG, TM,CM,
CP, FQ, TC
MexC MexD OprJ CM, FQ
MexE MexF OprN EM, TC
MexJ MexK OprM/ CM, EM, FQ, TC
OprH
MexV MexW OprM AG, ML, TC, TG
MexM MexN OprM BL
MexX MexY OprM AG, ML, TC, TG
S. enterica serovar AcrAa AcrBa TolCa BL, CM, EM, FQ, NO, RF, TC
Typhimurium
S. maltophilia SmeA SmeB SmeC AG, BL, FQ
SmeDa SmeEa SmeFa EM, FQ, TC (MC), TG
Table 4.1 summarizes data extracted from Li et al. [3]. For additional pumps and details regarding
substrate ranges and pump regulation, consult this very comprehensive review
MFP membrane fusion protein, RND resistance-nodulation-division pump component, OMF outer
membrane factor
Antibiotics: AG aminoglycosides, BL β-lactams, CM chloramphenicol, EM erythromycin, FQ
fluoroquinolones, FU fusidic acid, ML macrolides, NO novobiocin, RF rifampicin, TC tetracy-
clines, MC minocycline, TG tigecycline, TM trimethoprim
a
Denotes a pump that is expressed constitutively (housekeeping pump), but regulatory mutations
can further upregulate expression

switch to an open state of TolC [9]. A direct interaction between AcrB and TolC has
also been shown in vitro and also in cells using chemical cross-linking [10], but
other models suggest an alternative mechanism of assembly where AcrB and TolC
do not interact [11]. A very recent study showing in vitro reconstitution of AcrAB-­
TolC and MexAB-OprM using nanodisc technology and characterization by single-­
particle electron microscopy revealed a structure whereby the pump and OMF were
linked by the MFP but did not directly interact [12]. Whether that structure repre-
sents the final functional pump assembly in a cellular context or if direct interaction
76 C. R. Dean et al.

Fig. 4.2 Rotating functional mechanism of efflux by RND family pumps (represented by AcrAB-­
TolC); compounds enter at AcrB access conformation; AcrB undergoes a conformational change
to the binding mode and then to the extrusion mode where the compound is released into the outer
membrane channel. Side view of assembled pump with access and extrusion depicted (top); top
view cross section of functional AcrB rotamers (bottom)

between the pump and OMF is required for function is not currently resolved, but
these observations support the notion that the MFP component itself likely forms
part of the exit duct between the RND pump component and the OMF. Structural
studies with AcrB done by independent groups [13–15] and later simulation studies
[16, 17] revealed that drug efflux occurs by a functional rotation mechanism
(Fig. 4.2). Each of the three protomers of the assembled pump component (e.g.,
AcrB) can exist in one of three states referred to as “access” (or “loose”), “binding”
(or “tight”), and “extrusion” (or “open”) (Fig. 4.2). A complete functional cycle
occurs as follows: a compound enters the access conformation of AcrB from the
periplasm (likely from the outer leaflet of the cytoplasmic membrane) via a trans-
membrane domain called the vestibule. AcrB then changes conformation to the
binding conformation, which opens the large compound-binding pocket to accom-
modate the entry of the compound, and finally AcrB rotates to the open (extrusion)
conformation which releases the compound from the binding pocket into the funnel
region toward the OMF (TolC). As mentioned above, the interaction of the MFP
component with TolC keeps TolC in an open formation allowing compounds to be
expelled outside the cell. Energy for this process is derived from transport of pro-
tons from the periplasm to the cytoplasm, and it is suggested that a proton is released
to the cytoplasm when AcrB transitions from the tight to the open conformation
[14]. Consistent with RND pumps requiring proton-motive force to function, energy
4 Resistance of Gram-negative Bacilli to Antimicrobials 77

decouplers like CCCP inhibit efflux. The location of the vestibule in pump proteins
like AcrAB is such that compounds enter from the periplasmic leaflet of the cyto-
plasmic membrane, thereby suggesting generally that RND pumps recognize com-
pounds as they are entering the cell rather than after they ultimately reach the
cytosol. This is consistent with early observations that certain RND family pumps
reduced susceptibility to β-lactam antibiotics or β-lactamase inhibitors that target
penicillin-binding proteins or β-lactamase enzymes, respectively, which are located
in the periplasm [18–20], and with the reported importance of amino acid residues
in periplasmic loops of the inner membrane pump components in determining sub-
strate recognition [21–23]. As mentioned above, single-component pumps from
other families may also play a possibly underappreciated role in acting coopera-
tively with RND pumps when they have overlapping substrate specificities and the
cellular antibacterial target of a compound is cytosolic [6]. This has been fairly well
established in the case of tetracycline-­specific MFS (TetA/C) pumps which specifi-
cally efflux tetracycline into the periplasm where broader specificity RND pumps
that recognize tetracycline, such as MexAB-OprM, can extrude the compound from
the periplasm [24]. It remains to be determined in detail where these multi-pump
interactions are important in terms of clinical resistance. Without the contribution of
an RND pump in this sequential efflux, the compound may accumulate in the peri-
plasm where it may readily diffuse back in across the cytoplasmic membrane. When
effluxed out of the periplasm by the RND pump, it can diffuse away or alternatively
must reenter the cell by again traversing the OM. This raises the concept of com-
pound influx and the role of the Gram-negative OM permeability barrier as it relates
to RND-mediated efflux, which is discussed in the next section.

4.2.1.3  he Gram-negative Outer Membrane (OM) Permeability Barrier

T
and Its Interrelationship with Efflux

The OM of Gram-negative bacteria differs from the cytoplasmic membrane phos-

pholipid bilayer in that it is asymmetrical, having an inner leaflet of phospholipid
and an outer leaflet of lipopolysaccharide (LPS) (Fig. 4.3). The basic structure of
LPS is comprised of lipid A, which forms the outer leaflet of the membrane bilayer,
to which is attached the core oligosaccharide that extends out from the cell surface
[25]. Lipid A core is often decorated with a highly variable polysaccharide repeat-
ing unit (O-antigen). Each lipid A molecule contains several acyl chains, and lipid
A is packed together by Mg2+ cross-links between phosphates on the lipid
A. Additional cross-linking between phosphates on the core oligosaccharide can
also be important in some bacteria [26]. Because of this, the Gram-negative OM
bilayer can provide a formidable permeability barrier to a wide variety of mole-
cules, since it has a net negative charge combined with the hydrophobic layer pro-
vided by the lipid portion of the bilayer. Differences in lipid A structures and
variation in lipid A cross-linking among Gram-negative bacteria can cause differ-
ences in the permeability barrier of the OM bilayer.
78 C. R. Dean et al.

Fig. 4.3 Chemical structure of lipopolysaccharide from E. coli 0157:H7. The lipid A forms the
outer leaflet of the asymmetrical outer membrane. Acyl chain number and lengths and level of lipid
A and core phosphorylation vary among different Gram-negative bacteria

The requirement for nutrient uptake across the OM bilayer is generally met by
water-filled protein β-barrel channels that span the bilayer, known as porins (see
Fig. 4.1). Porins allow for the passage of small hydrophilic molecules across the
OM, essentially establishing the overall OM as a molecular sieve. These channels
are also thought to allow influx of certain hydrophilic antibiotic molecules that are
small enough to traverse the porin channels [27]. Gram-negatives such as E. coli
have several relatively nonspecific large porins such as OmpC and OmpF with
molecular weight cutoffs of approximately 600 Da [28]. In contrast P. aeruginosa
harbors a number of more specialized or restrictive (smaller) porins to allow influx
of nutrients. This organism also has the general porin OprF, but this exists only
occasionally in the conformation that allows the channel to be open [29]. This
highlights that, along with variability in the lipid bilayer characteristics among
different Gram-negative bacteria, the number and characteristics of the OM porin
channels can also vary, causing large differences in the effectiveness of the OM
permeability barrier. Reflecting this, the OM of P. aeruginosa was estimated to be
more than tenfold less permeable than that of E. coli [30]. The best studied exam-
ples of antibiotic permeation via porins center on various β-lactams. For example,
carbapenems can enter E. coli via OmpC porins, and some carbapenems such as
imipenem enter P. aeruginosa via OprD. In the latter case, the natural function of
4 Resistance of Gram-negative Bacilli to Antimicrobials 79

OprD is transport of basic amino acids, which bear some structural resemblance
to certain carbapenems. Imipenem was also recently shown to enter P. aeruginosa
via OpdP [31], suggesting that some antibiotics may access cells via multiple
porins. Hydrophobic and/or larger compounds not able to enter by porins can
enter the bacterium by diffusion across the membrane bilayer, although these pro-
cesses are likely slower and are not well understood. As well, there are a limited
number of specialized energy-dependent active transporters in the OM that import
scarce nutrients such as iron (as siderophore or protein-bound complexes) [32] or
cobalamin [33].
Overall, the OM of Gram-negative bacteria is highly evolved to provide a strong
protective permeability barrier, but it still allows influx of important nutrients, either
by passive diffusion in the case of porins or active transport in other cases. For most
antibiotics though, the OM slows their influx considerably, and the combination of
reduced antibiotic influx due to the OM permeability barrier and active efflux
together typically determines the levels of susceptibility [34, 35]. RND family
efflux pumps do not always exhibit a high velocity of antibiotic efflux, so beyond
whether the pump recognizes a specific antibiotic, its effectiveness depends to a
large extent on how slowly a given substrate antibiotic is entering the cell. If the
influx rate is too fast, the pump may not “keep up,” and even if the compound is a
substrate, the pump may not confer meaningful resistance. In contrast, if the mem-
brane barrier is slowing influx, efflux pumps can then become a very significant
resistance factor. This was elegantly shown for oxacillin compared to ampicillin in
E. coli. These antibiotics were shown to be very similar as substrates of the AcrB
efflux pump, but oxacillin traversed porins more slowly. Deletion of the acrB gene
had a much larger impact on susceptibility to oxacillin than to ampicillin [36, 37].
Furthermore, disruption of either efflux or the OM permeability barrier in P. aeru-
ginosa strongly increased antibiotic susceptibility, with an even greater increase in
susceptibility when both were disrupted simultaneously, showing the interplay
between these two factors [38]. More recently it was reported that expression of an
OM iron-siderophore transporter that had been engineered to create very large
porin-like channels in the bacterial OM strongly increased susceptibility to a range
of antibacterial compounds, further showing the important role of the OM permea-
bility barrier and its interrelationship with active efflux in several Gram-negative
pathogens [39, 585]. Correspondingly, mutations that impact either the OM perme-
ability barrier or efflux in Gram-negative bacteria can decrease susceptibility to
antibiotics, and this is discussed in the next section.

4.2.1.4  he Role of Efflux and the OM Permeability Barrier

T
(Cell Permeability) in Decreasing Antibiotic Susceptibility

The combination of the OM permeability barrier and efflux is important for dictat-
ing the spectrum of Gram-negative pathogens that a clinically used or novel antibi-
otic under development will be sufficiently active. In those cases where useful
antibacterial activity does occur (e.g., with currently used antibiotics), mutations
80 C. R. Dean et al.

that increase the expression of efflux pumps, alter substrate recognition, or impact
the OM permeability barrier (decreased compound influx) erode this activity over
time and limit a compound’s therapeutic longevity. Since many RND efflux pumps
have broad substrate ranges, selection of pump upregulation with one compound
will usually affect susceptibility to multiple antibiotics, contributing to multidrug
resistance. These factors are particularly problematic since they contribute to the
therapeutic demise of current antibiotics while also being the major impediment to
the discovery of novel replacement antibiotics. As an example illustrating both of
these points, tigecycline, a glycyl derivative of minocycline that evades the classic
tetracycline-specific resistance mechanisms of ribosomal protection and efflux by
tetracycline-specific single-component efflux pumps (e.g., TetA) (discussed in
detail in Sect. 4.2.4), was still subject to intrinsic RND-mediated efflux in P. aeru-
ginosa [40], and therefore its spectrum does not include this pathogen. Although
tigecycline achieves useful antibacterial activity against other Gram-negatives and
has been successfully implemented clinically, mutations leading to RND pump
upregulation can erode this activity in organisms such as Proteus mirabilis [41], K.
pneumoniae [42], E. coli [43], and A. baumannii [44].
The impact of cell impermeability on the discovery of new anti-Gram-negative
antibiotics is difficult to overstate, as the vast majority of compounds are subject to
some level of efflux and/or limited influx. In one study, the majority of antimicrobial
compounds identified from direct antibacterial screening in E. coli were AcrAB-
TolC pump substrates [45]. In a broader discussion of overall screening efforts con-
ducted at AstraZeneca, the inability of compounds to accumulate in Gram-negative
bacteria was cited as a significant impediment to novel antibiotic discovery using
corporate compound collections [46]. Additional examples of novel compounds
that inhibit specific bacterial targets but are subject to efflux include the peptide
deformylase inhibitor LBM415, which is subject to AcrAB-TolC-mediated efflux in
H. influenza [47], and CHIR-090 an LpxC inhibitor that has potent intrinsic activity
against P. aeruginosa but selects in vitro for mutations that upregulate expression of
several RND efflux pumps [48]. Similarly, standard antibiotics with good intrinsic
activity against Gram-negative pathogens, such as fluoroquinolones, many
β-lactams, and aminoglycosides, also select for mutations causing increased pump
expression which can decrease susceptibility substantially [3].
The selection of mutations leading to pump upregulation underscores the idea
that although RND pump expression is generally subject to intricate regulation,
pump expression is not typically induced by antibiotics of clinical importance. An
exception is the strong induction of the MexXY efflux pump of P. aeruginosa by
compounds that perturb protein synthesis. Even in that case, however, induction
occurs in response to a range of structurally and mechanistically unrelated protein
synthesis inhibitors including aminoglycosides and tetracyclines [40], novel ribo-
some inhibitors such as argyrin B [49] or to mutations that impair ribosome function
[50, 51]. Therefore, MexXY expression is responsive to ribosome impairment [52,
53] rather than to the specific antibacterial compounds, and in some cases such as
argyrin B, the inducing compound may not be a pump substrate. Novobiocin was
also shown to directly bind to the NalD repressor of the MexAB-OprM efflux pump
and induce pump expression [54].
4 Resistance of Gram-negative Bacilli to Antimicrobials 81

In general, pump upregulation leading to decreased susceptibility to antibiotics

generally occurs by selection of stable mutations in the pump gene promoter region
or more often in regulatory genes. The circuits controlling efflux pump expression
can be highly complex, and overall this topic is beyond the scope of this section, but
in general, efflux pump regulation often involves repressor proteins that bind opera-
tors upstream of pump genes to reduce pump expression unless relieved by either an
intracellular signal or interaction with other modulatory elements (e.g., MexR and
NalD, which control expression of MexAB-OprM in P. aeruginosa [55–57]), posi-
tive activators that bind and induce pump gene expression in response to intracel-
lular signals (e.g., MexT which activates expression of MexEF-OprN in P.
aeruginosa [58, 59]), or in some cases two-component histidine kinase sensor
response regulator pairs (e.g., BaeRS and CpxRA [60–63]). In most Gram-negative
bacteria, there is usually a housekeeping pump (e.g., MexAB-OprM of P. aerugi-
nosa or AcrAB-TolC of E. coli) that is expressed constitutively, with additional
pumps that are not appreciably expressed, at least under laboratory conditions (e.g.,
MexCD-OprJ of P. aeruginosa and AcrEF-TolC of E. coli). Expression of pumps
such as MexAB-OprM and AcrAB-TolC can be increased by mutations in genes
encoding their cognate repressors (e.g., mexR [64] and nalD [65] or other regulators
(nalC [66]) for MexAB-OprM or acrR [67] for AcrAB-TolC). Mutations in genes
controlling typically silent pumps, such as MexCD-OprJ and MexEF-OprN, can
turn on expression to generally high levels with corresponding increases in resis-
tance to their substrate antibiotics [59, 68]. Upregulated pumps are routinely found
among clinical isolates [3]. Since pump upregulation can result from simple loss-of-­
function mutations in repressor genes, these mutants can be selected at high fre-
quencies under antibiotic exposure levels within which the efflux pump can
accommodate. Furthermore, since RND or other pumps can extrude common bio-
cides, such as chlorhexidine, pump upregulation is likely selected in the environ-
ment by biocide-containing cleaning solutions [69].
The complex regulatory circuits controlling expression of some efflux pumps can
also control the expression of OM porins through which some antibiotics cross the
OM. For example, the highly complex MAR (multiple antibiotic resistance
(reviewed in [3, 7, 70]) regulatory circuit controls AcrAB-TolC expression and
porin expression in E. coli. Therefore mutants having both reduced antibiotic influx
and increased efflux via AcrAB-TolC can emerge. Similarly, P. aeruginosa mutants
that overexpress MexEF-OprN are also downregulated for expression of the porin
OprD, the main entry route of carbapenems into the cell [59]. Reduced susceptibil-
ity to carbapenems in nfxC mutants is thought to be mediated mainly by reduced
influx rather than efflux. Mutations affecting porin expression or function, including
mutations within porin genes, have been described in several bacteria and in particu-
lar are associated with carbapenem resistance in Enterobacteriaceae and P. aerugi-
nosa clinical isolates [71–74]. Mutations in genes encoding the two-­component
regulator ParRS in P. aeruginosa were shown to cause inducible or constitutive
resistance to four classes of antibiotic (polymyxins, aminoglycosides, fluoroquino-
lones, and β-lactams) via a combination of increased efflux (MexXY/OprM), porin
downregulation, and aminoarabinose modification of lipopolysaccharide (LPS)
82 C. R. Dean et al.

[75]. The latter affects the ability of polymyxin (cationic peptides) to interact with
LPS which is required for their entry into cells (discussed in Sect. 4.2.6). These
examples serve to illustrate the extensive ability of many Gram-­negative bacteria to
survive exposure to toxic molecules by preventing their accumulation in the cell.
The ability of mechanisms such as efflux and porin loss to enhance survival under
exposure to antibiotics also supports the emergence of other resistance mechanisms
such as specific target mutations, ultimately facilitating the emergence of very high
levels of resistance.

4.2.1.5  fforts to Address Compound Accumulation in Gram-negative

E
Bacteria

Lack of sufficient compound accumulation in cells is arguably the single most

important specific factor hindering the discovery and development of novel antibiot-
ics for Gram-negative pathogens. The search for novel antibacterials with new
mechanisms of action is predicated on avoidance of cross-resistance with existing
mechanisms selected by currently used antibiotics. The presence of RND efflux
pumps (and selection of pump over-expressors clinically), as a general broad resis-
tance mechanism, often serves to defeat this strategy in cases where the novel com-
pound is a pump substrate and has an intracellular target. It is reasonable to speculate
that the efflux-/OM-mediated permeability barrier has coevolved with many intra-
cellular targets that are essential for growth or viability in order to exclude most
molecules with the characteristics required to optimally bind and inhibit intracellu-
lar essential targets. Therefore, strong interest has developed within the antibiotic
discovery field in understanding the Gram-negative OM permeability barrier and
efflux with a view toward two general goals: interfering with cell impermeability as
a way of potentiating antibiotic activity by increasing the cellular accumulation of a
partner antibiotic (combination therapy) and the understanding of the design of
inhibitors that are less impacted by efflux and can penetrate cells effectively to reach
their target. See the Innovative Medicines Initiative (IMI) translocation effort
(https://www.imi.europa.eu/content/translocation) for more details.

Potentiation of the Cellular Activity of Antibiotics

One strategy to potentiate partner antibiotics that has garnered extensive interest
over the years is the design of efflux pump inhibitors (EPIs). In theory, a potent EPI
could improve the spectrum and potency of a range of currently used antibiotics
whose usefulness is compromised by efflux. EPIs could also serve to enhance and
extend the clinical usefulness of novel agents that are or may become affected by
efflux. Several EPIs have been described over the last two decades.
The first EPI described in detail as an inhibitor of multiple RND family pumps,
MC207,110 (phe-arg-β-napthylamide, PAβN) was originally identified by screen-
ing for compounds that potentiated the activity of pump substrate fluoroquinolone
4 Resistance of Gram-negative Bacilli to Antimicrobials 83

antibiotics in P. aeruginosa [76]. Inhibitors such as MC 207,110 are pump sub-

strates and are thought to act through competitive binding and interference with
substrate antibiotic recognition [77]. MC207,110 is a lipophilic amine and as such
falls into a chemical property space known for target promiscuity and associated
challenges in achieving an acceptable safety profile [78]. Perhaps reflecting this,
MC207, 110 was also shown to have some bacterial membrane-disrupting activity
[76, 79]. In contrast to MC207,110, the EPI D13-9001 is more specific to the
MexAB-OprM efflux pump within P. aeruginosa [80]. Compounds like D13-9001
would therefore need to be partnered with an antibiotic effluxed primarily by
MexAB-OprM. The mechanism of pump inhibition involves binding of D13-9001
into a hydrophobic “trap” extending off the substrate translocation channel within
MexB, ultimately compromising the pump’s functional rotation [81, 82]. This
mechanism is more likely than that of a substrate competition mechanisms, such as
that of MC207,110, to block efflux of multiple antibiotics and to enable better inhib-
itor potency. Furthermore, D13-9001 is zwitterionic which places it in a potentially
less promiscuous chemical property space. Newer pyranopyridine EPI molecules
(e.g., MBX2319 and analogs [83–85] that take advantage of the hydrophobic trap
mechanism have achieved substantial increases in potency. Recently described EPIs
(NSC 60339 and analogs) were shown to have a very novel mechanism of inhibition
of AcrAB-TolC, by binding the membrane fusion protein (AcrA), inducing struc-
tural changes, and possibly interfering with assembly of the functional pump [86].
Advances in the structural understanding of pump assembly and function, as well as
the binding of several EPI molecules and the diversity of potential pump inhibitory
mechanisms, may increase the ability to design new EPI molecules in the future.
Selection of a suitable partner antibiotic is potentially a complex issue, espe-
cially with clinically used antibiotics. This is because non-efflux-based resistance
mechanisms (e.g., target or modifying enzyme-based mechanisms) affecting many
standard antibiotics may have become widespread, and those mutants may be resis-
tant even if efflux is fully inhibited. Secondly, many antibiotics are effluxed by
several pumps within a given Gram-negative pathogen or across different bacteria,
thereby requiring a broad spectrum of EPI activity to cover multiple pumps. To date
no Gram-negative EPI has reached clinical use. MC207,110-based analogs were
lipophilic cations, and ultimately unfavorable toxicity profiles could not be over-
come [77]. It remains to be seen if this approach will be successful with other novel
inhibitors. Finally, the intriguing finding that a significant percentage of P. aerugi-
nosa clinical isolates recovered from cystic fibrosis patients have mutationally lost
MexAB-OprM function and become susceptible to ticarcillin has prompted sugges-
tions that this antibiotic may find use in treating this subpopulation [87].
An alternative approach to efflux inhibition may be disruption of the bacterial
membrane, thereby improving the ability of a partner antibiotic to gain access to
the cell. It is well established that mutations affecting the synthesis or assembly
of the Gram-negative OM cause hypersusceptibility to a range of antibiotics,
especially those that are more hydrophobic in nature or are large molecular
weight (i.e., too large to pass through porins and whose exclusion from cells is
mediated mainly by the permeability barrier of the OM bilayer). Targets impor-
84 C. R. Dean et al.

tant for OM biosynthesis/assembly that may affect the OM permeability barrier

if inhibited include the lipid A biosynthetic genes lpxA, lpxB, and lpxC [88, 89]
and the LPS transport/assembly genes lptD and lptE in E. coli [90–92] and P.
aeruginosa [93, 94]. Chemical ­inhibitors of targets such as these could induce
disruption of the OM permeability barrier and strongly potentiate the activity of
many antibiotics used in combination, although precisely where this may occur
at a clinically useful level remains to be determined. Alternatively upon target
inhibition, a corresponding progressive disruption of the permeability barrier
may generate a cycle of increased uptake of the inhibitor itself, improving cel-
lular potency, although again this remains to be shown for specific examples.
Since LPS synthesis and assembly per se are essential in many Gram-negative
bacteria, enzymes such as LpxC (first committed step in lipid A biosynthesis),
LpxA/LpxD, and LptD have generated interest as targets for the design of novel
inhibitors. For example, extensive medicinal chemistry efforts have resulted in
potent small molecule inhibitors of LpxC with antibacterial activity [95–108], at
least one of which reached early phase clinical evaluation [109]. P. aeruginosa-­
specific peptidomimetics targeting LptD (POL7001, POL7080) have very potent
antipseudomonal activity [108, 110], suggesting the potential of this approach.
Intriguingly, there are a small number of Gram-negative bacteria that can survive
in the absence of lipid A biosynthesis, including the important pathogens
Neisseria meningitides [111, 586] and some A. baumannii strains [112–114].
Loss of lipid A in the latter has been shown to result from mutations in lipid A
biosynthetic genes lpxA, lpxC, or lpxD [113], and lpxC and lptD can be geneti-
cally deleted in some A. baumannii [115]. Therefore, inhibitors of these targets
would not be expected to have good, or any, antibacterial activity against such
strains. However, reflecting the potential of inhibitors of these targets to potenti-
ate antibiotics, strains harboring these genetic mutations or wild-type strains
exposed to LpxC inhibitors became susceptible to several antibiotics [113, 115,
116]. Importantly, an intact OM permeability barrier is also important for sur-
vival of pathogens (virulence) during infection, by providing protection from
serum complement and other host immune factors.
Molecules such as LPS are also strong activators of toll-like receptors (e.g.,
TLR4). Consistent with this, an LpxC inhibitor with no appreciable in vitro antibac-
terial activity against A. baumannii was highly protective in a mouse infection
model [117]. This was attributed to increased opsonophagocytic killing and lower
levels of released LPS, leading to reduced inflammation. This example illustrates
the additional potential of inhibitors of targets such as LpxC, used alone or in com-
bination with other antibiotics, resulting from effects on the OM. A. baumannii
represents an extreme example to show this, since some A. baumannii can tolerate
a large or total loss of lipid A synthesis, providing the widest possible window to see
this effect. It remains to be understood how broadly this might translate across dif-
ferent targets/inhibitors and different Gram-negative pathogens in clinically rele-
vant scenarios. As with any novel antibacterial, in particular ones that must still
reach an intracellular target such as LpxC, a variety of resistance mechanism are
likely to be able to impact their cellular activity. In vitro, mechanisms such as target
4 Resistance of Gram-negative Bacilli to Antimicrobials 85

mutations, target overexpression, partial bypass by mutations in fabG, and upregu-

lated efflux can all reduce the activity of the LpxC inhibitor CHIR-090 in P. aerugi-
nosa [48], and mutations in fabZ reduce susceptibility of E. coli to LpxC inhibitors
[118]. In contrast to targets like LpxC which are intracellular, the OM itself is also
a target, and compounds that could interact with and disrupt the OM might potenti-
ate antibiotics without having to reach an intracellular target. Classic examples of
this are cationic molecules, such as polymyxins, which interact with LPS via phos-
phates attached to lipid A and disrupt the OM permeability barrier causing sensiti-
zation to antibiotics [119–121]. Such an approach to potentiate antibiotics against
Gram-negative pathogens (compound SPR741, an analog of polymyxin B nonapep-
tide) is currently being pursued [122].

Understanding Compound Penetration to Improve Access to Targets

While factors limiting a compound’s OM and inner membrane (IM) permeability

are fairly well understood, structure activity relationships for efflux remain incom-
plete. Strategies to improve compound access to intracellular targets must consider
the specific compartment in the cell where the target resides. A target might reside
anywhere from the cell surface to the periplasmic space between the inner and OM
through to the cytosol of the cell. Different issues come into play for each of these.
One way to reduce the complexity of cell penetration is to pursue targets located
near or on the cell surface, circumventing the need for significant cell penetration.
As mentioned above, polymyxin-based antibiotic potentiators such as SPR741
directly target the LPS on the cell surface. Similarly, the cationic peptidomimetic
POL7001 targets LptD, an essential protein localized in the OM [110] at or near the
cell surface. The number of OM protein targets that are essential for growth is low
(i.e., LptD and BamA), and these targets exist as components of complex machinery
which could be more difficult to disrupt or inhibit by small molecules, but this
remains to be fully understood. To date, only larger peptidomimetic inhibitors of
these targets have been described [110, 123], providing some possible insights into
the types of molecules expected to be active in this context. On a related note, elimi-
nating the need for cell penetration is a salient feature of the monoclonal antibody
approach, which functions specifically by exploiting surface-­exposed antigens. An
example is a recently described bispecific antibody (MEDI13902) targeting PcrV
(type III secretion) and the exopolysaccharide Psl [124, 125] which is currently
undergoing clinical trials for prevention of nosocomial pneumonia. Progressing fur-
ther into the cell, some targets reside in the periplasmic space between the IM and
OM. The number of essential targets here is also limited but includes the clinically
validated penicillin-binding proteins (PBPs) that are inhibited by the β-lactam class
of antibiotics. Inhibitors of periplasmic targets need only to cross the OM, which
lends itself to utilization of water-filled porins for compound access, with corre-
sponding optimization for this route of entry. Entry through porins relies to a large
extent on a size small enough to traverse the porin (porin cutoff is approximately
600 Da ), compound polarity (hydrophilicity), and appropriate charge distribution.
86 C. R. Dean et al.

The mechanisms by which translocation of compounds occurs via porins have been
the subject of extensive investigation [126–128], and assays to evaluate porin trans-
location for use in drug design are being pursued [129, 130]. A recent study sug-
gested that porin traversal may be optimal for small polar compounds with charged
groups and a dipole moment having a component aligned perpendicular to its main
axis [128]. Since β-lactams have periplasmic targets, they may represent the class of
antibiotic that can best be optimized specifically for permeation through porins. It is
likely that this contributes in some cases to reducing the impact of efflux since rapid
influx can overwhelm the capacity of RND family efflux pumps even if the com-
pound is a pump substrate [36, 37]. The majority of novel antibacterial targets how-
ever are located in the cytosol, and understanding compound penetration (and
evasion of efflux) becomes more complex in that case than for compounds such as
β-lactams. This stems from the necessity to traverse the two distinctly different
membranes. The OM severely limits influx of larger or more hydrophobic com-
pounds, which must diffuse across the asymmetrical OM bilayer which has low
fluidity and presents a high barrier for lipophilic compounds, since they are excluded
from entry through porins [3]. Smaller more hydrophilic molecules can traverse the
OM through porins, but extensive optimization for polarity to maximize this can
hinder entry across the symmetrical phospholipid inner membrane bilayer, which
favors diffusion of hydrophobic molecules. Antibiotics directed at cytosolic targets
may therefore need some element of amphiphilicity to cross both membranes,
which also may increase recognition by RND efflux pumps [131]. A much better
understanding of the chemical property space required for the design of cell active
inhibitors of intracellular targets, and the representation of this chemical property
space typically found in corporate screening libraries, is likely required to overcome
the ongoing inability to deliver novel antibacterials in this area. Initial efforts toward
this understanding were described by O’Shea and Moser [132], who examined the
properties of a wide range of antibacterial compounds and, consistent with our
understanding of the cell envelope, correlated properties such as molecular weight
and polarity with Gram-negative antibacterial activity. A more recent look at data
from a wide range of screening efforts at AstraZeneca also suggested that polarity
and small size correlated with reduced efflux and cellular activity and increasing
compound hydrophobicity could drive biochemical target inhibition but possibly at
the expense of cellular activity [46]. However, increased polarity itself was not suf-
ficient to ensure antibacterial activity [46], again suggesting a fine balance of prop-
erties is likely necessary. This is consistent with the notion that the cell envelope and
efflux likely coevolved with intracellular targets to exclude molecules with the
properties to strongly bind and inhibit essential targets. These properties may also
be compound scaffold specific. Modulation of physicochemical properties (pKa and
logD) improved the antibacterial activity of novel bacterial type II topoisomerase
inhibitors [133]. A very recent study directly measured accumulation of compounds
in E. coli using mass spectrometry, and computational analysis indicated that rigid,
amphiphilic compounds with low globularity and containing an amine moiety accu-
mulated better. These rules were applied to convert a compound that was active only
against Gram-positive bacteria into one with E. coli activity [134]. This suggests it
4 Resistance of Gram-negative Bacilli to Antimicrobials 87

Fig. 4.4 Representative structures for each class of β-lactam. The core structure is depicted in
blue; the specific side chains are depicted in black

may be possible to derive some general rules to engineer compounds with Gram-
negative accumulation and cellular activity, but more research will be necessary to
validate this concept. Furthermore, compound accumulation must be achieved
together with low toxicity in order for resulting compounds to be therapeutically
useful. Efforts are also underway to explore new methods complementary to mass
spectrometry to measure cellular accumulation of compounds, which could further
assist in defining rules for cell penetration [135, 136]. Finally, the Trojan-horse
concept has also been applied to antibacterial design. In this scenario, an antibiotic
is linked to a compound that is actively transported into cells, thereby exploiting the
active uptake mechanism to drive intracellular accumulation of the chimeric antibi-
otic molecule. Among others, a recent example of this is compound cefiderocol
(S-649266), a catechol cephalosporin that is proposed to utilize energy-dependent
siderophore-iron uptake systems in Gram-­negative bacteria for improved cellular
access [137].

4.2.2 β-Lactams and β-Lactamase Inhibitors

The identification of the β-lactam benzylpenicillin in the 1920s essentially started

the antibiotic era [138]. Initially used to treat soldiers in World War II, the lifesaving
potential of these compounds was quickly realized, leading to the design of novel
β-lactams that continues to this day. β-lactams fall into four classes: penicillins,
cephalosporins, carbapenems, and monobactams (monocyclic β-lactams) (Fig. 4.4).
Together these comprise by far the most widely used class of antibiotics worldwide.
The history and details of the development of the multitude of β-lactam antibiotics
are beyond the scope of this chapter and have been recently reviewed in [139].
Examples of β-lactams with a broader spectrum that can include serious Gram-­
negative pathogens such as E. coli, K. pneumoniae, and P. aeruginosa are the peni-
cillins, such as ampicillin, amoxicillin, carbenicillin, and piperacillin; the
cephalosporins such as ceftazidime and cefepime; the carbapenems such as imipe-
nem, meropenem, and doripenem; and the monobactams, such as aztreonam.
β-lactams are bactericidal and target penicillin-binding proteins (PBPs) [140].
Gram-negative bacteria possess multiple PBPs which are important for cross-­
linking of the peptidoglycan that makes up the rigid bacterial cell wall [141].
88 C. R. Dean et al.

β-lactams resemble segments of the growing peptidoglycan (e.g., D-Ala-D-Ala)

and, after the formation of a low-affinity complex, covalently bind (acylate) the PBP
at its active-site serine residue. Bacteria possess multiple PBPs, broadly classified
into high-molecular-weight (HMW) and low-molecular-weight (LMW) categories
[141]. In general, the LMW PBPs are monofunctional D-Ala-D-Ala carboxypepti-
dases, whereas the HMW PBPs are either bifunctional (class A, transpeptidase and
transglycosylase) or monofunctional (class B, transpeptidase) [141]. Not all PBPs
are essential for growth, but certain ones such as PBP3 (main target of aztreonam)
are essential [141, 142]. Furthermore, many β-lactams can acylate the active-site
serine of several PBPs which can contribute meaningfully to increased antibacterial
activity. Inhibition of PBPs like PBP3 or multiple PBPs like PBP3 with PBP1a/
PBP1b can ultimately lead to cell lysis [140, 142]. Recent studies have now eluci-
dated that inhibition of PBPs also triggers a lethal malfunctioning of the cell wall
synthetic machinery [143].
The high potency of many β-lactams against Gram-­negative pathogens relies to
a large extent on two factors, compound access and the nature of the targets them-
selves. PBPs are located in the Gram-negative periplasmic space. This means that
β-lactams only need to cross the OM to access their targets. Therefore β-lactams can
be polar molecules able to traverse water-filled porins, which have the fortuitous
benefit of improving their safety. The second factor is that these relatively “exposed”
PBPs are particularly good targets in conjunction with this class of inhibitor, since
the compound-target interaction is covalent (essentially irreversible), and PBP inhi-
bition causes severe impairment of a fundamental cellular process, leading ulti-
mately to lethality. However, the PBP targets also form the basis of the overarching
issue with resistance to β-lactams, which is the expression of β-lactamase enzymes.
These enzymes are expressed in the periplasm and appear to have evolved from
PBPs to attack and efficiently hydrolyze the β-lactam, mediating resistance [144].
Non-β-lactamase mechanisms that affect susceptibility to β-lactams in Gram-
negative pathogens include efflux, loss of uptake porins, and amino acid substitu-
tions in the target PBPs. These topics are addressed in the following sections.

4.2.2.1 β-Lactamases

The discovery of β-lactamases predated the clinical use of benzylpenicillin, but the
widespread use of these agents in the clinic has, over time, led to the emergence of
an astonishing number of β-lactamase variants [144, 145], which as a group can
degrade most or all β-lactam antibiotics. Indeed the development of new β-lactam
antibiotics is to some extent a continuing story of addressing the emergence of new
β-lactamases [146, 147], as is the ongoing development of β-lactamase inhibitors
(BLIs) for use in combination with β-lactams to restore their activity against
β-lactamase-expressing strains (see Sect. 4.2.2.3). β-lactamases hydrolyze the
β-lactam ring of all classes of β-lactam antibiotics by one of the two major mecha-
nisms. The first is mediated by an active-site serine (Ser), via a covalent enzyme
intermediate that is rapidly hydrolyzed causing inactivation of the antibiotic.
β-lactamases that operate by this mechanism are therefore referred to as serine
4 Resistance of Gram-negative Bacilli to Antimicrobials 89

Table 4.2 Classification of clinically relevant β-lactamases

Representative
Molecular Functional Representative enzyme in
class group Description Substrates families clinical isolates
A 2be Extended-­ Penicillins, TEM, SHV, TEM-3, SHV-2,
spectrum cephalosporins, CTX-M, PER, PER-1, VEB-1,
β-lactamases monobactams VEB CTX-M-15,
(ESBLs) CTX-M-9,
CTX-M-14,
CTX-M-3
A 2br Inhibitor-­ Penicillins, TEM, SHV TEM-30,
resistant narrow-spectrum SHV-10
β-lactamases cephalosporins
A 2f Serine Carbapenems, KPC, IBC, KPC-1, KPC-2,
carbapenemases cephalosporin, IMI, NMC, KPC-3, SME-1
cephamycins SME, GES,
SFC,
B 3a Metallo-­ Carbapenems, IMP, VIM, VIM-1 VIM-2
carbapenemases penicillins, NDM, SPM, IMP-1
cephalosporins, GIM, SIM, NDM-1
cephamycins AIM, DIM,
FIM, POM
C 1 AmpC Cephamycins, CMY, FOX, CMY-1, CMY-2
β-lactamases cephalosporins, ACC, LAT, ACT-1, DHA-1,
narrow-spectrum ACT, MOX, DHA-2,
monobactams, DHA, MIR, CMY-13,
and penicillins CMY-4
D 2de Extended-­ Cephalosporins, OXA OXA-10,
spectrum oxacillins OXA-13,
β-lactamases OXA15,
(ESBLs) OXA-18,
OXA-45
D 2df Carbapenemases Carbapenems, OXA OXA-48,
oxacillins OXA-23
OXA-40,
OXA-51,
OXA-58

β-lactamases. This mechanism is reminiscent of that for PBP inactivation by

β-lactam antibiotics, as β-lactamases share an active-site Ser-XX-Lys motif with
PBPs. A main difference in these processes is a comparatively very low rate of
hydrolysis of the covalent adduct in the case of PBPs. The second mechanism is
metal-mediated, whereby one or two bivalent metal ions activate a water molecule
that attacks the β-lactam ring [148]. These β-lactamases are correspondingly
referred to as metallo-β-lactamases. The large number of serine and metallo-β-­
lactamases is categorized via two different classification systems (Table 4.2). The
Ambler classification is based on protein sequence homology that divides
β-lactamases into four classes (A, B, C, and D). Classes A, C, and D are all serine
β-lactamases, whereas class B is the metallo-β-lactamases. The second classifica-
tion scheme in use for β-lactamases, defined by Bush-Jacoby, is based on enzymatic
90 C. R. Dean et al.

functionality and divides β-lactamases into three major groups: group 1 cephalospo-
rinases (class C), group 2 serine β-lactamases (classes A and D), and group 3
metallo-β-lactamases. Each major group is then divided into several subgroups
based on specific attributes [145]. The first β-lactamase, TEM-1, identified in a clin-
ical isolate was reported in the early 1960s in an E. coli isolate from a patient in
Greece [149]. Since then, the number of β-lactamases identified has constantly
grown. A recent report estimated that over 2000 unique β-lactamases sequences
have been identified [150]. The major players in the clinic for infections caused by
Gram-negative pathogens are the extended-spectrum β-lactamases (ESBLs), the
AmpC cephalosporinases, and the serine and metallo-carbapenemases.

Extended-Spectrum β-Lactamases (ESBLs)

Extended-spectrum β-lactamases (ESBLs) confer resistance to nearly all β-lactam

antibiotics except carbapenems and cephamycins. ESBLs were first identified in the
mid-1980s in K. pneumoniae and Serratia marcescens [151]. The occurrence of
ESBLs in clinical isolates has been constantly increasing in the past two decades. A
recent Centers for Disease Control and Prevention (CDC) report estimated nearly
26,000 healthcare-associated Enterobacteriaceae infections are caused by ESBL-­
producing Enterobacteriaceae (19% of isolates) causing 1700 deaths each year
(https://www.cdc.gov/drugresistance/threat-report-2013/index.html). The fast
spread of ESBL-producing strains is due to the presence of these genes on mobile
genetic elements, such as plasmids, usually carrying other antibiotic resistance
genes [145, 152, 153]. Early ESBLs evolved from the TEM and SHV enzymes to be
able to hydrolyze oxyimino-cephalosporins, and these are molecular class A, func-
tional group 2be. Subsequently, the ESBL category expanded to include enzymes
such as the CTX-M family, mainly present in E. coli and K. pneumoniae; the PER
family identified in Pseudomonas, Acinetobacter, and Salmonella species; and the
VEB family reported in A. baumannii. These β-lactamases are not genetically
related to TEM or SHV β-lactamases but have similar hydrolytic profiles and are
part of the functional group 2be [153–155]. The most recent ESBLs are the OXA
family, originally reported in P. aeruginosa, isolated in Turkey and France. The
OXA family, in contrast to the other ESBLs, belongs to molecular class D and func-
tional group 2de [156].
ESBLs are prevalent in the clinic and present serious problems in hospital-
acquired infections, leading to increased mortality worldwide. ESBL prevalence
varies across different geographic regions. In a recent report on Enterobacteriaceae
isolates collected in 63 US hospitals from 2012 to 2014, 13.7% of these isolates had
an ESBL profile. Different trends were observed among different species of
Enterobacteriaceae; in E. coli the ESBLs occurrence increased from 12.7% (2012)
to 15.1% (2014), whereas in K. pneumoniae, rates decreased from 18.9% (2012) to
15.5% (2014). Also, a statistically significant variation was observed across differ-
ent regions in the United States. In the South Atlantic Region, the ESBL rates
decreased from 20.8% (2012) to 9.2% (2014). Conversely in the Pacific region, the
4 Resistance of Gram-negative Bacilli to Antimicrobials 91

ESBL rates increased from 11.4% (2012) to 16.9% (2014). The predominant ESBLs
identified in this study were CTX-M-15 (59% of ESBLs) followed by SHV (19% of
ESBLs), both mainly in K. pneumoniae isolates, and CTX-­M-­14 (18% of ESBLs)
[157]. In Europe the ESBL rates vary considerably by country. The prevalence of
ESBLs in E. coli in the 2014 European surveillance varies from 3.3% in Iceland to
40.4% in Bulgaria. Even more alarming is the prevalence of ESBLs in K. pneu-
moniae with rates over 70% in Greece, Bulgaria, and Romania. On a positive note,
the rates of ESBLs did not increase from 2009 to 2014, attributed to the increased
use of carbapenems [158]. Also across Europe, the most prevalent ESBL types iden-
tified in clinical isolates were the CTX-M family β-lactamases, but the specific type
varies considerably among countries, with CTX-­M-­9 and CTX-M-14 enzymes
dominant in Spain and CTX-M-3 and CTX-M-15 dominant elsewhere [159].
Another growing family of ESBLs is the OXA-type enzymes that confer resis-
tance to ampicillin and cephalothin, are characterized by their high hydrolytic activ-
ity against oxacillin and cloxacillin, and are very poorly inhibited by clavulanic
acid. The OXA-type enzyme genes differ genetically from all other ESBLs. To date,
over 500 different OXA-type variants have been reported, but not all are ESBLs.
The OXA-type enzymes with activity against oxyimino-­cephalosporins are OXA-
10 and its variants (OXA-11, OXA-14, OXA-16, and OXA-17), OXA-13 and its
variants (OXA-19 and OXA-32), and some other OXA enzymes (OXA-15, OXA-
18, and OXA-45). These enzymes have been identified mainly in P. aeruginosa
isolates [155, 160].
Even though ESBL incidence rates have not been increasing in the past few
years, they are still very high in some parts of the world and are a major health con-
cern. Further, ESBLs are often present on mobile genetic elements with other anti-
biotic-resistant determinants, including those for aminoglycosides and
fluoroquinolones. The use of carbapenems to treat infections caused by ESBL-
producing pathogens is increasing the emergence of carbapenem-resistant strains,
starting the debate on how to better treat those pathogens. Using a β-lactamase
inhibitor (see Sect. 4.2.2.3) with a β-lactam is in principle a targeted and effective
approach. A detailed analysis on the benefit of β-lactam/β-lactamase inhibitor com-
binations for the treatment of ESBL-producing pathogens can be found in a recent
review by Viale et al. [161].

Class C β-Lactamase (AmpC)

AmpC β-lactamases belong to class C and functional group 1. They confer resis-
tance to cephamycins, such as cefoxitin and cefotetan, and cephalosporins, includ-
ing oxyimino-cephalosporins such as ceftazidime, cefotaxime, and ceftriaxone.
They are also able to hydrolyze to a lesser extent penicillins and aztreonam [162].
The majority of AmpC β-lactamases are not or are only weakly inhibited by inhibi-
tors of class A enzymes such as clavulanic acid, sulbactam, and tazobactam. Some
AmpC variants have been reported to be inhibited by tazobactam or sulbactam [162,
163]. Several AmpC β-lactamases are chromosomally encoded enzymes, found in
92 C. R. Dean et al.

Acinetobacter spp., C. freundii, Enterobacter spp., E. coli, Hafnia alvei, Morganella

morganii, P. aeruginosa, and Yersinia enterocolitica. AmpC basal expression is
generally low but can be induced to high levels in some bacteria (e.g., M. morganii
and P. aeruginosa) upon exposure to some β-lactams. The regulation of AmpC
expression varies among different organisms. In bacteria where AmpC is inducible,
the ampC gene is accompanied by ampR, encoding a member of the LysR transcrip-
tional regulator family. Disruption of peptidoglycan synthesis by β-lactams causes
accumulation of peptidoglycan fragments that dislodge oligopeptides of UDP-N-­
acetylmuramic acid normally bound to AmpR, causing a conformational change
where AmpR then positively activates transcription of ampC. The activity of AmpR
is controlled indirectly by the activities of AmpD, a N-acetyl-muramyl-L-alanine
amidase, and an inner membrane permease, AmpG, which are both involved in
recycling of peptidoglycan intermediates. Therefore, although ampC expression is
inducible in these cases, constitutive upregulation can occur via mutations in the
genes encoding these regulatory factors. In clinical isolates, the most common cause
of AmpC hyperexpression is mutation in ampD. Mutations in ampR causing AmpC
hyperexpression have been reported but are not as common. Mutations in ampG
only result in constitutive low-level expression and are the least common [162]. In
P. aeruginosa PAO1, AmpC expression is very tightly regulated by the presence of
three AmpD genes with different affinities for their substrates. These AmpD genes
are also reported to be involved in P. aeruginosa PAO1 virulence [164]. Other
organisms like E. coli and Shigella lack AmpR, and regulation occurs via a weak
promoter and a strong attenuator. In these cases, AmpC expression is not inducible
by β-lactams, and hyperexpression of AmpC leading to resistance results from
mutation in the ampC promoter or attenuator [165–167]. AmpC β-lactamases can
also be expressed from plasmids. The first plasmid-encoded AmpC variant, CMY-1,
was identified in 1989 from K. pneumoniae isolated from a wound infection in
South Korea. The high degree of resistance to cefoxitin was due to the high-level
constitutive expression of CMY-1 [162, 168]. Since the identification of CMY-1,
several families of plasmid-encoded AmpC variants have been reported in clinical
isolates, especially in K. pneumoniae and E. coli.
Based on the source of the ampC gene, several plasmid-encoded AmpC families
have been reported: the two CMY families (CMY-1 and CMY-2), the FOX family,
the ACC family, the LAT family, the MIR family, the ACT family, the MOX family,
and the DHA family [162, 169, 170]. Plasmids encoding ACT-1, DHA-1, DHA-2,
and CMY-13 typically contain an ampR gene, and as such expression of these
β-lactamases is inducible, whereas the other plasmid-encoded AmpC variants lack
ampR and are not inducible. The high level of expression for the non-inducible
plasmid-encoded AmpC variants is mainly due to strong promoters and high-gene
copy number. As with several other plasmid-­borne antibiotic resistance genes, plas-
mids harboring AmpC β-lactamase genes often carry resistance determinants for
fluoroquinolones, sulfonamides, tetracyclines, aminoglycosides, chloramphenicol,
trimethoprim, and other β-lactamases (ESBLs and metallo-β-lactamases) [162, 168,
171, 172]. Clinical isolates of K. pneumoniae and Salmonella enterica carrying
plasmid-encoded AmpC have been reported to be resistant to cephalosporins and
4 Resistance of Gram-negative Bacilli to Antimicrobials 93

cephamycins as well as carbapenems. Detailed analysis of these strains showed that

the resistance to cephalosporins and cephamycins was due to the plasmid-mediated
CMY-4 β-lactamase for Salmonella enterica and DHA-1 for K. pneumoniae,
whereas resistance to carbapenems also involved the lack of outer membrane porin
proteins [173–177]. Combination of plasmid-encoded AmpC with deletions of
porin genes (and possibly increased efflux) results in high-level resistance to most
if not all β-lactams and leaves clinicians with few treatment options.

Carbapenemases

Carbapenems are considered the most effective β-lactams for the treatment of seri-
ous infections caused by Gram-negative bacteria and present a broad spectrum of
antibacterial activity. Furthermore, carbapenems are relatively stable to most ESBLs
and class C enzymes and are deemed to be safer to use than any other last-resort
antibiotic. Therefore, the increasing number of reports of β-lactamases able to
hydrolyze carbapenems over the last few years is of major concern. Carbapenemases
are a heterogeneous group of β-lactamases, including members from classes A, B,
and D [178, 179]. Most carbapenemases are able to hydrolyze a very broad spec-
trum of β-lactams. Several class A enzymes, functional group 2f, with carbapenem-­
inactivating activity, have been reported over the years. The first, SME-1, was
reported in 1990 in a Serratia marcescens isolate from the United Kingdom [180].
Subsequently, GES-1 in K. pneumoniae, SFC-1 in Serratia fonticola, and IBC-1/
IMI-1/NMC-A in Enterobacter cloacae have been reported [181]. One of the most
recent and widespread class A carbapenemases is the K. pneumoniae carbapene-
mases (KPCs). KPC-1 was the first carbapenemase identified for this family of
enzymes, reported in 1996 in North Carolina [182, 183]. KPCs, in contrast to the
other class A carbapenemases that are chromosomally encoded, are usually on
mobile genetic elements and since their discovery have spread to many other organ-
isms including most species of Enterobacteriaceae (including Enterobacter spp.,
Serratia marcescens, and Salmonella spp.), P. aeruginosa, and several other genera.
A recent study that analyzed 147 cases of infections due to carbapenem-resistant K.
pneumoniae from 2013 to 2014, in one hospital in Northern Italy, showed that the
major resistance determinant was KPC-3 (83.8%). The death rate was an alarming
24.0% in 2013 and 37.5% in 2014 [184]. This is of great concern especially in
Southern European countries where CREs expressing KPC are spreading rapidly.
Several enzymes of the class D family (OXA type) are able to degrade car-
bapenems [185]. The first identified enzyme in this class able to hydrolyze imi-
penem was OXA-23. It was isolated in the United Kingdom in 1985 from an A.
baumannii isolate and was originally characterized as ESBL [186]. Since this
initial report, several other OXA enzymes have been described in
Enterobacteriaceae, such as OXA-23-like, OXA-40-like, OXA-51-like, OXA-
58-like, and OXA-48-like enzymes [185, 187]. The most widespread enzymes in
this class are the OXA-48-­like enzymes. OXA-48 was first isolated in a patient
with UTI in Turkey in 2001 from a strain of K. pneumoniae and is now widely
94 C. R. Dean et al.

spread across Enterobacteriaceae and other Gram-negative species [188]. In the

recent past, OXA-48-like carbapenemases have been responsible for several out-
breaks among carbapenem-resistant K. pneumoniae in Spain [189] and Greece
[190], and currently no broadly active inhibitors of class D enzymes are on the
market, in part due to high structural diversity within this class of enzymes [185].
Metallo-β-lactamases (MBLs) belong to Ambler class B, functional group 3, and all
can inactivate carbapenems [191]. MBL hydrolysis of β-lactams is mediated by zinc
and inhibited by metal chelators such as ethylenediaminetetraacetic acid (EDTA) but
not clavulanic acid or other clinically used β-lactamase inhibitors. MBLs can hydro-
lyze to varying degrees members of all β-lactam classes except monobactams. Since
the early 1990s, when the first MBL, IMP-1, was detected, the number of transmissible
genes encoding MBLs has been constantly growing. The major MBLs currently pres-
ent in the clinic are IMP-like, VIM-like, and NDM-like enzymes. Other MBLs such us
the SPM, GIM, SIM, AIM, DIM, FIM, and POM have been reported but are not widely
distributed [181]. More than 30 derivatives of IMP-like enzymes have been reported
and are commonly found in Japan, China, and Australia causing sporadic outbreaks
[187]. In contrast, the VIM-like carbapenemases have been reported from hospitals
worldwide. The first VIM-like carbapenemase, VIM-1, was identified in Italy from a P.
aeruginosa isolate in 1997 [192]. Currently, over 30 VIM-like carbapenemase have
been reported around the world with VIM-2 being the most widespread MBL in P.
aeruginosa [191]. VIM-like enzymes are often harbored in gene cassettes and are also
associated with integrons [193]. The prevalence of VIM-like enzymes among MBL-
producing Enterobacteriaceae isolates in Europe is very high, especially in countries
like Italy and Greece. A multicenter European survey showed the presence of VIM-1-
like enzymes in 98.9% of Enterobacteriaceae isolates producing MBLs [187].
The New Delhi metallo-β-lactamase (NDM), first reported in 2009, is the latest
carbapenemase described that is threatening the usefulness of β-lactams globally. So
far, 13 different variants of NDM have been reported; in several cases, the mutation in
the blandm gene seems to predict rates of β-lactam hydrolysis [194]. NDM producers
have been isolated across the globe but predominately in the Asian continent and
mainly in India. In the United Kingdom and the Middle East, outbreaks of NDM-­
producing strains have been reported in the recent past. Several reports have shown that
NDM-1-producing pathogens are resistant to many other antibiotics, thus limiting
options for treating these infections to a small number of agents such as polymyxins,
fosfomycin, and aminoglycosides which are rarely used due to efficacy and/or safety
concerns [194]. As well, many NDM-1-producing strains also possess ribomethylase,
leading to aminoglycoside resistance (see Sect. 4.2.5.2). In 2010 in Canada, a NDM-1-
producing K. pneumoniae outbreak was reported, which included a patient with no
prior history of traveling to Asian countries that are high risk for these infections. In
vitro susceptibility tests showed that the strain was resistant to the majority of available
antibiotics except colistin and tigecycline [195]. Several studies summarized in a recent
review by Zmarlicka et al. have suggested however that in vitro resistance to several
β-lactams by NDM-1-producing organisms does not always translate to clinical out-
comes, suggesting that some carbapenem/β-­lactamase inhibitor combinations may still
work in the clinic [194]. More data and in-depth analysis would be needed to fully
understand this.
4 Resistance of Gram-negative Bacilli to Antimicrobials 95

4.2.2.2 Non-β-Lactamase-Mediated Resistance

Since β-lactam antibiotics cross the Gram-negative OM via porins, mutations

causing loss of porins or affecting their structure or expression level can reduce
susceptibility by reducing influx. Well-characterized examples of this are loss of
OmpK35 and/or OpmK36 in K. pneumoniae [196, 197] and loss of OprD in P.
aeruginosa [198, 199]. Loss or downregulation of these may have a significant
impact on susceptibility to carbapenems and other classes. Mutations leading to
upregulation of RND family efflux pumps can also reduce susceptibility to
β-lactams, and in some cases, efflux upregulation occurs concomitant with porin
downregulation (see Sect. 4.2.2). These mechanisms cause modest shifts in sus-
ceptibility generally but become significant in isolates where β-lactamases are
also expressed [174, 200]. Alterations in PBPs (target mutations) had not gar-
nered a lot of attention in Gram-negative pathogens, although it stands to reason
that the widespread use of β-lactams/β-­lactamase inhibitors is applying selective
pressure for the emergence of altered patterns of PBP expression and/or muta-
tions in PBPs of these organisms. Consistent with this, altered expression pat-
terns of PBPs have been reported in pan-β-lactam-­resistant clinical isolates of P.
aeruginosa, although no changes in the amino acid sequences were found [199].
Amino acid substitutions in PBP2 have been found in E. coli clinical isolates that
affect susceptibility to carbapenems [201]. Recently amino acid insertions in
PBP3 were identified in clinical E. coli isolates that affect susceptibility to a
range of β-lactams including the monobactam aztreonam [202, 203]. The mecha-
nisms of porin loss, efflux, and PBP changes in isolation only shift β-lactam
susceptibility modestly, but cumulatively they can have a large impact, especially
when combined with the expression of β-lactamases. That PBP3 insertions mod-
estly shift susceptibility to aztreonam is concerning since monobactams are the
only class of β-lactams that are intrinsically stable to NDM metallo-β-­lactamases,
for which no inhibitors are currently available, and PBP3 insertions have been
reported in NDM-1 expressing E. coli isolates in certain geographic areas [202].
These strains often express various serine β-lactamases as well. β-lactamase
inhibitors such as avibactam can address these β-lactamases, but underlying
mutations altering PBPs combined with porin loss and efflux are likely to
erode the effectiveness not only of β-lactams but also currently available
β-lactam/β-lactamase inhibitor combinations.

4.2.2.3 β-Lactamase Inhibitors

Dissemination of β-lactamases prompted efforts to identify β-lactamase inhibitors

(BLIs) to restore effectiveness of partner β-lactams used in combination. Currently,
there are four BLIs in clinical use: clavulanic acid, sulbactam, tazobactam, and the
newly approved avibactam (Table 4.3).
96 C. R. Dean et al.

Table 4.3 β-Lactamase inhibitors on the market or in clinical development

Name Chemical class Combination with Current status
Clavulanic acid β-Lactam Amoxicillin Marketed (also generic)
Sulbactama β-Lactam Ampicillin Marketed (also generic)
Tazobactam β-Lactam Piperacillin or Marketed
ceftolozane
Avibactam Diazabicyclooctane Ceftazidime Marketed
Avibactam Diazabicyclooctane Ceftaroline, Phase III, Pfizer
aztreonam Phase II, Pfizer
Relebactam Diazabicyclooctane Imipenem, Phase III, Merck
cilastatin
Nacubactam: Diazabicyclooctane Meropenem Phase I, Roche
RG6080, OP0505
Vaborbactam Boronate Meropenem New drug application
(RPX7009) (NDA) (Carbavance)
ETX2514 Diazabicyclooctane Sulbactama Phase I, Entasis [204]
AAI101 β-Lactam (tazobactam Cefepime or Phase I, Allecra
analog) piperacillin
Zidebactam Diazabicyclooctane Cefepime Phase I, Wockhardt
a
Sulbactam has antibacterial activity against A. baumannii

The first of these to be identified and brought to the clinic was clavulanic acid, a
natural product isolated from Streptomyces clavuligerus [205], followed by the
semisynthetic penicillanic acid sulfone class of inhibitors (sulbactam and tazobac-
tam) [206, 207]. These BLIs all possess the basic core structure of a β-lactam which
allows for recognition and binding to β-lactamase. However, key structural differ-
ences from β-lactams eliminate most or all intrinsic antibacterial activity against
many bacteria and render them mechanism-based “suicide inhibitors” of sensitive
β-lactamases [208]. The mechanism of β-lactamase inactivation by these inhibitors
is complex, but in general, the active-site serine of the β-lactamase attacks the car-
bonyl group in the β-lactam ring of clavulanic acid leading to acylation of the
β-lactamase. This is then followed by a series of secondary reactions in the enzyme
active site that irreversibly inactivate the enzyme [209–211]. A main difference
between the BLI molecules and β-lactams that facilitates this mechanism is that
BLIs possess good leaving groups at the C-1 position of their five-membered rings.
This allows for secondary ring opening and subsequent β-lactamase enzyme modi-
fication. Important factors for BLI efficacy include high acylation and low deacyla-
tion rates, which localize them for a longer time period in the enzyme active site and
a low number of hydrolytic events (inhibitor molecules hydrolyzed per unit time)
necessary for complete enzyme inactivation (termed turnover number or tn).
Differences exist in these factors between clavulanic acid, sulbactam, and tazobac-
tam, and these are affected by differences in the active sites among β-lactamases.
Clavulanic acid and tazobactam cover most class A β-lactamases including ESBLs.
Sulbactam also covers these but is less potent against some enzymes. Tazobactam
and sulbactam are better inhibitors of class C carbapenemases than clavulanic acid
4 Resistance of Gram-negative Bacilli to Antimicrobials 97

and notably differ from clavulanic acid in that they do not induce expression of
AmpC in bacteria where this enzyme is inducible [212]. However, none of these can
cover strains producing metallo-β-lactamases such as NDM-1. Clavulanic acid is
partnered in the clinic with amoxicillin or ticarcillin, sulbactam with ampicillin, and
tazobactam with either piperacillin or ceftolozane.
Avibactam, the first non-β-lactam BLI approved for clinical use, is a broad inhib-
itor of class A (including KPCs), class C, and some class D β-lactamases. Avibactam
is a member of the diazabicyclooctane (DBO) chemical class [213], and as such it
has a different mechanism of inhibition from previous BLIs. This mechanism is not
yet fully understood, but it appears that avibactam functions as a slowly reversible
covalent inhibitor with release of intact avibactam for most class A and C
β-lactamases [214]. The enzymes appear to be slowly acylated and slowly deacyl-
ated, with no or only low-level hydrolysis of the inhibitor molecule. An exception to
this was inhibition of KPC-2 which was rapidly acylated but slowly deacylated with
hydrolysis of avibactam, so differences do exist. The release of intact avibactam in
most cases however is thought to allow for recycling of the inhibitor by β-lactamases
in the cell, leading to better inhibitory efficiency than BLIs like clavulanic acid or
tazobactam that are hydrolyzed. Avibactam is a substantially more effective inhibi-
tor of key β-lactamases like TEM-1, KPC-2, and AmpC from P. aeruginosa than
clavulanic acid, sulbactam, or tazobactam but has limited coverage of class D
enzymes, although it does cover OXA-48. Like previous BLIs, avibactam does not
cover metallo-β-lactamases like NDM-1. Avibactam was introduced into the clinic
very recently (2015), partnered with ceftazidime, and is in clinical trials for combi-
nation use with ceftaroline or aztreonam. The latter partnering with the monobac-
tam aztreonam is meant to capitalize on the idea that monobactams are inherently
stable to metallo-β-lactamases, and the combination should therefore cover strains
expressing metallo-β-lactamases and/or serine β-lactamases. Although ceftazidime/
avibactam has only been in clinical use a short while, resistance to this combination
due to porin loss and upregulation of β-lactamase expression has been reported
[215]. It has recently been suggested that ceftazidime/avibactam could be used in
combination with aztreonam for coverage of metallo-β-lactamase/serine-β-­
lactamase-producing clinical isolates [216], but appropriate dosing would need to
be established.
The success of avibactam has inspired efforts to identify next-generation
DBO β-lactamase inhibitors. These include relebactam (MK7655), directed at
some class A, including KPCs, and class C β-lactamases (currently in Phase III
trials, in combination with imipenem [217]). Relebactam possesses a narrower
spectrum than avibactam since it does not include class D β-lactamases such as
OXA-48. Nacubactam (RG6080, OP0595), directed at class A and C β-lactamases,
is currently in Phase I trials and is intended for combination with meropenem
(Table 4.3). Some DBOs also possess intrinsic antibacterial activity and this war-
rants some discussion. They inhibit PBP2 in some Gram-negative pathogens,
similar to the β-lactam mecillinam [218]. PBP2 inhibition can be synergistic
with inhibition of other PBPs (i.e., with other β-lactams), as has been reported
for nacubactam [218, 219], and this has been referred to as an “enhancer effect”
98 C. R. Dean et al.

independent of BLI activity. Compounds that inhibit PBP2, such as mecillinam

(amdinocillin) or nacubactam, select for mutants with reduced susceptibility
in vitro at high frequency [219, 220]. There is a multiplicity of mutations that
engender tolerance of PBP2 inhibition, generally related to the stringent or enve-
lope stress responses [220], as well as stringent response-independent mecha-
nisms [221]. Such mutations do not affect inhibition of PBP2 specifically;
therefore the enhancer effect of DBOs is retained. Some intrinsically active
DBOs may be considered as potent stand-alone antibacterials against Gram-
negative pathogens, including P. aeruginosa, and this potential may be somewhat
overlooked [222]. Interestingly, the gene encoding PBP2 was reported as nones-
sential in P. aeruginosa based on genetic deletion [223], suggesting that chemi-
cal inhibition of PBP2 is distinct from genetic deletion, possibly due to induction
of a futile cell wall pathway cycle as described in [144], but this is not fully
understood. At least one DBO with potent activity against P. aeruginosa demon-
strated a high frequency of in vitro resistance selection [224]. It is not known if
the intrinsic antibacterial activity of some PBP2-specific DBOs and their poten-
tial in vitro resistance profile will affect the clinical outcome when used in com-
bination with a β-lactam. It has been proposed that mutants selected in vitro may
not be fit enough to survive in the host [220], and it is possible that PBP2 inhibi-
tion would still provide efficacy against these mutants in vivo. Studies with
potent antibacterial DBO molecules in relevant animal models to examine resis-
tance potential in that context are needed to resolve this ongoing discussion.
Another DBO has recently been described (ETX2514 [204], Entasis) with
broader inhibitory activity than other DBOs since it includes class D enzymes. Like
nacubactam, ETX2514 has intrinsic antibacterial activity against some Gram-­
negative pathogens. ETX2514 is being paired with sulbactam (which has intrinsic
antibacterial activity against A. baumannii) [217] for treatment of A. baumannii
infections.
Another non-β-lactam BLI class is the boronic acid chemical scaffold. Boronic
acid compounds were originally shown to inhibit serine proteases, and this observa-
tion was then extended to the serine β-lactamases. These inhibitors form a covalent
reversible adduct between the boronate moiety and the catalytic serine of the
β-lactamase. The most advanced of these is the cyclic boronate compound RPX7009
(Vaborbactam, The Medicines Company) which was the first of this class for which
in vivo efficacy was demonstrated [218, 219]. Vaborbactam is active against class A
carbapenemases (including KPCs), as well as other class A and class C β-lactamases
[220, 221], but does not inhibit metallo-β-lactamases like NDM-1. A new drug
application (NDA) has been filed for the combination of vaborbactam with merope-
nem for treatment of complicated urinary tract infections (cUTIs). Additional details
on the inhibitors described above can be found in recent reviews [139, 210, 222].
The category of β-lactamases that has proven most challenging for the design of
broad-spectrum inhibitors is the class B metallo-β-lactamases. To date, no inhibitor
of these enzymes has reached the market. An alternative strategy to the design of a
metallo-β-lactamase inhibitor is to exploit the fact that the monobactam aztreonam
is intrinsically stable to metallo-β-lactamases and can therefore be partnered with
4 Resistance of Gram-negative Bacilli to Antimicrobials 99

avibactam; this combination has the potential to cover strains expressing both
metallo-β-lactamase and serine β-lactamases. This combination is currently under-
going Phase III clinical trials. More recently, an innovative approach was under-
taken to design novel next-generation monobactams that are not significantly
impacted by most serine β-lactamases while retaining their intrinsic stability to the
metallo-β-lactamases. One of these, LYS228 [223], demonstrated excellent potency
against MDR Enterobacteriaceae, including CRE [224, 225], and has entered Phase
II clinical trials (Novartis). Significant effort has also been devoted to the discovery
of therapeutically useful inhibitors of the class B metalloenzymes. This has lagged
to some extent since class B enzymes have a different mechanism than serine
β-lactamases, and it appears that the design of inhibitors capable of covering mul-
tiple clinically important class B enzymes is technically challenging. Inhibitors of
class B enzymes would also need to be highly specific and avoid human metalloen-
zymes to avoid toxicity issues. While the prevalence of class B enzymes remained
relatively low in the past and their contribution to worldwide carbapenem resistance
was initially considered minimal, this viewpoint has changed in recent years with
increased spread of class B enzymes like NDM-1 and their linkage with other resis-
tance determinants. Recent efforts in the search for class B inhibitors include the
discovery of the natural product aspergillomarasmine, which is active against
NDM-1 and VIM-2 [226]. Novel bisthiazolidine (BTZ) inhibitors of class B
enzymes have also recently been described [227]. Of most interest are reports from
The Medicines Company on a new series of cyclic boronate compounds derived
from RPX7009 (Table 4.3) with broad-spectrum carbapenemase activity including
metallo-β-lactamases, in preclinical development [228].

4.2.2.4 Resistance to β-Lactamase Inhibitors

The implementation of β-lactamase inhibitors extended the clinical usefulness of

several β-lactam antibiotics for decades; however the emergence of variant
β-lactamases and other mechanisms has eroded their usefulness. One factor that
impacts the effectiveness of BLIs is the expression level of β-lactamases and/or the
number of β-lactamases being expressed in a given isolate. Even if a BLI is potent
inhibitor of serine β-lactamases, this can be overwhelmed by high-level expression
of one or multiple β-lactamase enzymes [229]. Overexpression can result from the
β-lactamase gene residing on multicopy plasmids or via mutations in the promoter
region causing high-level expression [230]. In particular, high-level expression of
TEM-1 was an early mechanism identified that reduced susceptibility to amoxicillin-­
clavulanate [231, 232]. This can also be related to the induction of β-lactamase by
certain BLIs. Clavulanic acid induces the expression of AmpC β-lactamase in bac-
teria where this enzyme is inducible. Since clavulanic acid is not a good inhibitor of
class C enzymes, this induction can be antagonistic toward the partner antibiotic
[233]. This is an issue in particular for the case of chromosomal inducible ampC (P.
aeruginosa) or plasmid-borne inducible ampC such as DHA-1 in K. pneumoniae. In
fact, antagonism by clavulanic acid is used as a diagnostic for the presence of
100 C. R. Dean et al.

inducible ampC [234, 235]. Sulbactam and tazobactam do not have this induction
effect and as such can be better options (e.g., tazobactam paired with piperacillin
against P. aeruginosa). Active efflux of BLIs [20, 236] or changes in influx, possibly
due to porin loss, may also serve to reduce their concentration relative to the
β-lactamases in the cells, decreasing their effectiveness, although influx of com-
pounds such as avibactam is not currently well understood [237]. Defects in porins
OmpK35 and/or OpmK36 were, however, associated with decreased effectiveness
of imipenem/relebactam and meropenem/RPX7009 in clinical isolates isolated
from hospitals in New York [219, 238]. Efflux was implicated as an important medi-
ator of resistance to ceftazidime/avibactam in P. aeruginosa, but this remains to be
further explored [239]. Porin mutations combined with upregulated expression of
plasmid-borne KPC-3 have been associated with clinical resistance to ceftazidime/
avibactam in K. pneumoniae [215], and porin mutations were associated with cef-
taroline/avibactam resistance in E. cloacae mutants selected in vitro [240].
Resistance to BLIs, mainly clavulanic acid, also resulted from the emergence of
new β-lactamases that are resistant to the BLI. Very soon after the introduction of
clavulanic acid into clinical use, such variants began to emerge, with the first
reported being variants of the class A TEM enzyme that were resistant to clavulanic
acid, found in E. coli clinical isolates [241, 242]. These TEM variants were altered
at their Arg244 residues to either Cys or Ser [243]. This position had been shown
earlier to be important for clavulanic acid inhibitory function, and so this clinical
outcome might have been expected. These variants were initially designated
inhibitor-­resistant TEM (IRT-1 and IRT-2), but since then, 37 clavulanic acid-­
resistant TEM variants have been identified (cataloged at http://www.lahey.org/
Studies/temtable.asp, functional group br), and the convention now is that these all
have TEM numerical designations. These variants are found mainly in E. coli iso-
lates but also occur in Klebsiella [244], Proteus [245], Shigella [246], and
Citrobacter [247]. Inhibitor resistance can also be combined with amino acid sub-
stitutions conferring β-lactamase activity against oxyimino-β-lactams (ESBL), and
these are referred to as complex mutant TEMs (CMT). Currently 11 of these vari-
ants have been described (http://www.lahey.org/Studies/temtable.asp, functional
group ber). As well, seven inhibitor-resistant variants of the class A enzyme SHV
have also been described (http://www.lahey.org/Studies/), with the most recent
being SHV-107 found in a K. pneumoniae clinical isolate [248]. It should be noted
that inhibitor-resistant β-lactamases generally refer to clavulanic acid, and these can
also be resistant to sulbactam, but generally they remain susceptible to tazobactam
[249–251]. Therefore these enzymes mainly affect amoxicillin/clavulanate, ticarcil-
lin/clavulanate, or ampicillin/clavulanate but not piperacillin/tazobactam. However,
in 2010, the emerging class A ESBL KPC-2 carbapenemase [252] was shown to
also be resistant to clavulanic acid, sulbactam, and tazobactam, raising serious con-
cerns [253].
Although ceftazidime/avibactam is active against KPC producers, providing
effective treatments in the near term, the emergence of β-lactamase variants resis-
tant to avibactam has already been reported [254]. In vitro selection studies demon-
strated that variants of KPC-3 [254] or certain inhibitor-resistant SHV enzymes
4 Resistance of Gram-negative Bacilli to Antimicrobials 101

(particularly the S130G variant) [255] were less susceptible to ceftazidime/avibac-

tam. In the former study, the most frequently isolated variant of KPC-3 was
Asp179Tyr, and the authors speculate this particular change may increase ceftazi-
dime specificity rather than mediating resistance to avibactam per se. Interestingly,
many of the alterations also appeared to impair the ability of the β-lactamase to
hydrolyze carbapenems (reversal of ESBL), thereby increasing susceptibility of the
bacteria to those agents. Consistent with these in vitro studies, plasmid-borne vari-
ants of KPC-3 have been described with reduced susceptibility to avibactam in K.
pneumoniae clinical isolates [256]. These emerged within 10–19 days of ceftazi-
dime/avibactam exposure. The KPC-3 variants found had either a D179Y/T243 M
double substitution, D179Y single substitution, or V240G single substitution.
Interestingly, these mutations also seemed to decrease the KPC-3 carbapenemase
activity enough in some isolates to render them susceptible, and it has been sug-
gested that agents like meropenem could be used to ameliorate to some extent the
impact of such mutations. A large-scale analysis of the binding pockets of class A
serine β-lactamases indicated that most would be susceptible to avibactam but some
outliers were identified. In particular, PER-4 was shown to be highly resistant to
avibactam [257], indicating the preexistence of class A β-lactamase variants in the
clinic that are resistant to avibactam.

4.2.3 Quinolones

Quinolones and the related fluoroquinolones (Fig. 4.5) were introduced into clinical
use in the 1960s and 1980s, respectively. First-generation quinolones (e.g., nalidixic
acid) were restricted generally to treating urinary tract infections, because of subop-
timal systemic distribution and somewhat limited activity. Second-generation fluo-
roquinolones (e.g., norfloxacin and ciprofloxacin) had improved tissue distribution
and a broadened antibacterial spectrum, allowing for expanded use and perhaps
overuse. Newer third- and fourth-generation fluoroquinolones (e.g., ofloxacin, lome-
floxacin, levofloxacin, trovafloxacin, gatifloxacin, moxifloxacin, sparfloxacin) were
focused mainly on improved Gram-positive and atypical (e.g., Mycoplasma,
Legionella) coverage. Quinolone antibiotics act by inhibiting DNA gyrase and
topoisomerase IV enzymes that control DNA topology and play essential roles in
DNA replication, transcription, and recombination [258]. The DNA gyrase holoen-
zyme tetramer consists of two subunits each of GyrA and GyrB, which act to intro-
duce negative superhelicity into DNA. This is required for initiation of replication,
replication fork movement, and transcription [259]. The domain responsible for
DNA strand passage resides on GyrA, whereas GyrB contains an ATPase domain.
The topoisomerase IV tetrameric holoenzyme similarly consists of two subunits,
each of ParC and ParE, and functions to relax both positive and negative supercoils
and to direct decatenation (unlinking) of replicated chromosome copies to allow for
chromosomal partitioning upon cell division. The DNA strand passing domain is
located on ParC, and the ATPase activity is mediated by ParE. Both holoenzymes are
102 C. R. Dean et al.

Fig. 4.5 Chemical structure of key quinolones. Nalidixic acid (first generation), ciprofloxacin
(second generation), levofloxacin (third generation), and trovafloxacin (fourth generation)

type II topoisomerases that introduce double-stranded breaks in DNA and pass DNA
strands/helices through each other via a transient “cleaved complex” where the
enzyme, covalently linked to the DNA, serves as a bridge between the DNA ends,
mediating strand breakage, strand passage, and resealing [259]. Although the exact
mechanisms by which different quinolones kill bacteria have not been fully unrav-
eled, their general mechanism involves forming reversible non-covalent complexes
with the topoisomerases bound to DNA. This forms a drug-enzyme-DNA complex
(ternary complex) that is trapped as the cleaved complex and ligation of the DNA
ends is prevented [258]. Subsequent destabilization of the complex without rejoin-
ing the ends introduces double-stranded DNA breakage, fragmenting the genome
and ultimately causing cell death [258, 260]. The trapped cleaved complexes also
interfere with progression of replication forks, blocking DNA synthesis [261] and
with transcription by blocking RNA polymerase [262] and disrupting the action of
DNA helicases [263]. Additional mechanisms may contribute to cell killing in cer-
tain Gram-negatives. For example, a recent report detailing the transcriptomic inter-
rogation of ciprofloxacin-treated P. aeruginosa implicated induction of a pyocin
system in cell-killing activity [264] (See also Chaps. 16 and 20).
Since there is significant amino acid sequence homology between the GyrA
and ParC and GyrB and ParE proteins, individual quinolone molecules can
inhibit the activities of both enzymes, and most quinolones will inhibit both tar-
gets to varying degrees. Either topoisomerase can constitute the primary or sec-
ondary target of quinolones in different bacteria. Given their broad-spectrum and
4 Resistance of Gram-negative Bacilli to Antimicrobials 103

excellent tissue penetration, fluoroquinolones are well suited to empiric therapy

and became one of the most broadly used classes of antibiotic. Quinolones are
also used fairly extensively in agriculture [265]. Given their widespread use, it is
not surprising that resistance has emerged at a significant rate around the world.
Resistance to fluoroquinolones in Gram-negative pathogens is mediated by sev-
eral mechanisms, the most common being chromosomal mutations that alter the
quinolone binding sites of the GyrA/B and ParC/E proteins. Additional mecha-
nisms include chromosomal mutations that upregulate RND-mediated efflux or
decrease compound penetration. Plasmid-based mechanisms also occur, includ-
ing efflux, target protection, and compound modification. Each of these is dis-
cussed below, and the epidemiology of fluoroquinolone resistance is discussed in
Chap. 10.

4.2.3.1 Target Mutations Conferring Quinolone Resistance

The most well-characterized mechanism conferring specific resistance to quino-

lones in both Gram-positive and Gram-negative bacteria is target alteration resulting
from chromosomal mutations in the gyrA and/or parC genes, with mutations in
gyrB and parE less frequently observed [266]. These changes occur in specific seg-
ments of the proteins referred to as their quinolone resistance determining regions
(QRDRs). The GyrA QRDR consists of amino acids 67–106 and for ParC encom-
passes residues 63–102 (E. coli numbering). Both of these regions comprise quino-
lone-binding domains and are located near amino-terminal active-site tyrosines that
interact covalently with transiently broken DNA [267–270]. Binding of quinolones
to GyrA or ParC occurs via water-metal ion bridges between the hydroxyl of con-
served serine or acidic amino acids within the QRDR and the oxygen of the quino-
lone amine group. Correspondingly the most frequently encountered resistance
alterations are QRDR substitutions at Ser83 of GyrA or ParC, with the next most
common being located at the Asp87 (acidic residue). Substitutions at Ser83 reduce
quinolone binding but do not substantially impact gyrase function [271], whereas
substitution at Asp87 decreases catalytic efficiency [272]. Although quinolones will
usually engage one or the other topoisomerase preferentially, they still impact the
secondary enzyme, and when mutations occur that reduce susceptibility of the pri-
mary target, alterations of the secondary target will usually also occur. For example,
quinolones preferentially target GyrA in E. coli, and therefore changes at the Ser83
of GyrA are most commonly found [272]. Single substitutions generally cause
modest changes in susceptibility to quinolones (Table 4.4), but over time additional
substitutions can occur in GyrA and/or ParC which ultimately lead to high-level
resistance [266, 272]. Alterations of the GyrB/ParE subunits are much less common
but do occur [266].
104 C. R. Dean et al.

Table 4.4 Summary of the impact of different resistance mechanisms on ciprofloxacin

susceptibility of E. coli and P. aeruginosa
Organism and resistance mechanism Fold change in ciprofloxacin MIC References
E. coli
  gyrA 32–64 [268, 273]
  gyrA + parC 128–2048 [273]
 Efflux upregulation 4–8 [274]
  qnr 32 [268]
  aac(6′)-lb-cr 8 [268]
P. aeruginosa
  gyrA 8–16 [275]
  gyrA + parC 256 [275]
 Efflux upregulation 2–16 [275]
  gyrA + parC + efflux upregulation 256–2048 [275]

4.2.3.2 Efflux and Reduced Compound Influx

The role of Gram-negative RND family efflux pumps in intrinsic and mutationally
acquired resistance to antibiotics is covered in detail in Sect. 4.2.1 and in recent
reviews [3]. The potent broad-spectrum activity of fluoroquinolones against even
intrinsically resistant Gram-negative bacteria such as P. aeruginosa indicates that
RND efflux does not mediate enough intrinsic resistance to some fluoroquinolones
to limit their spectrum. This may relate to some extent to the hydrophilicity [3] as
well as to the overall target potency and cidality of fluoroquinolones. However, fluo-
roquinolones are substrates of a wide range of RND family pumps (Table 4.1)
including the AcrAB-TolC pump of E. coli and Salmonella spp., the AcrEF pumps
of E. coli and Salmonella enterica, the CmeABC pump of Campylobacter jejuni,
and MexAB-OprM, MexXY-OprM, MexCD-OprJ, and MexEF-OprN of P. aerugi-
nosa [3]. Mutations in regulatory genes causing pump overexpression and decreased
susceptibility can be readily selected by in vitro exposure to quinolones. However,
in most cases overexpression of efflux pumps alone, in the absence of other mecha-
nisms, affords only modest reductions in fluoroquinolone susceptibility (Table 4.4).
Efflux pump overexpressing mutants are routinely found among clinical isolates
[276]. Upregulation of the AcrAB-TolC pump in fluoroquinolone-resistant E. coli
clinical isolates contributed to high-level fluoroquinolone resistance along with
QRDR mutations [67, 277]. Similarly, AcrAB-TolC upregulation played a role in
fluoroquinolone resistance in K. pneumoniae and K. oxytoca clinical isolates [278].
A P. aeruginosa clinical isolate with a mutation in gyrB and upregulated for MexAB-­
OprM emerged during ciprofloxacin monotherapy [279]. Since RND pumps have
broad substrate ranges, selection of pump overexpression during previous treatment
with various antibiotics will result in selection of pump upregulation which will
affect fluoroquinolones and vice versa. RND family efflux pumps function in con-
cert with the OM permeability barrier, and therefore any reduction in a compound’s
ability to cross the OM will have a corresponding enhancing effect on
4 Resistance of Gram-negative Bacilli to Antimicrobials 105

efflux-­
mediated resistance. Fluoroquinolones cross the outer membrane either
through water-filled porin channels or by diffusion through lipid domains in the
outer membrane depending on the hydrophobicity of the quinolone [280]. Reduced
porin levels have also been associated with fluoroquinolone resistance in E. coli
[277, 281] and S. enterica [282] clinical isolates. Reduced porin levels often occur
concomitantly with upregulation of efflux pumps [283] potentially linking reduced
influx and increased efflux via a single mutation.

4.2.3.3  lasmid-Mediated Quinolone Resistance (PMQR): Topoisomerase

P
Protection, Quinolone Modification, and Efflux

Three different plasmid-borne quinolone resistance mechanisms have been

described. These are topoisomerase protection (Qnr), quinolone modification, and
efflux. Each is discussed below. The plasmid-borne quinolone resistance determi-
nant qnr (now termed qnrA1) was first identified from a quinolone-resistant K.
pneumoniae clinical isolate in 1998 [284]. Other qnr determinants were subse-
quently identified including qnrB [285] and qnrS [286], and over the years, the
number has expanded to where there are currently seven families of Qnr proteins,
identified from a range of organisms: QnrA, QnrB, QnrC, QnrD, QnrS, and QnrVC
(cataloged at http://www.lahey.org/qnrStudies). The Qnr proteins are typified by
having tandem repeats of a pentapeptide consensus sequence and as such are
referred to as pentapeptide repeat proteins [287]. The mechanism of Qnr proteins is
referred to as topoisomerase protection and involves binding of the Qnr protein to
gyrase and topoisomerase subunits and the holoenzymes [288–290]. Binding is not
dependent on DNA or ATP and likely occurs prior to establishment of the ternary
complex, reducing quinolone interaction with the topoisomerases. More recent
structural information indicated that Qnr can assume a rodlike structure resembling
B-form DNA, suggesting it may compete with quinolones by binding in the gyrase
QRDR or DNA-gate region [291]. Although Qnr proteins can bind to a number of
subunits in vitro, they appear to mediate resistance to quinolones or other agents
that bind the QRDR region of GyrA, but not to agents that target the ATPase func-
tion (e.g., GyrB) [292]. Qnr proteins, specifically, only cause a marginal shift in
quinolone susceptibility similar to that of single-target mutations [284] (Table 4.4).
There are wide dissemination of qnr plasmids in Enterobacteriaceae clinical iso-
lates around the world [293–296] and significant diversity of plasmids that carry
these genes (reviewed in [266]). In contrast they seem to be rare among non-­
fermenters such as P. aeruginosa and A. baumannii. Interestingly, qnr genes cer-
tainly predate the introduction of the synthetic quinolones into clinical use and are
also found on the chromosomes of several bacteria [292]. It has been suggested that
mobilization from the genome to small transmissible plasmids may have originated
in Proteeae [297]. A final concern is that qnr genes typically reside on a range of
plasmids that also encode other resistance markers, in particular extended-spectrum
β-lactamases, such as SHVs and CTX-Ms, and AmpC-like enzymes such as DHA-1
(reviewed in [296]).
106 C. R. Dean et al.

The second identified plasmid-borne resistance determinant was a bifunctional

variant of the aminoglycoside-modifying enzyme aac(6′)-1b (see Sect. 4.2.5.1 on
aminoglycoside-modifying enzymes) [298]. This variant, designated aac(6′)-1b-cr,
differs from aac(6′)-1b by encoding two amino acid substitutions, Trp102Arg and
Asp179Tyr. These differences allow the enzyme to bind and acetylate fluoroquino-
lones that have an amino nitrogen on the piperazinyl ring (e.g., ciprofloxacin and
norfloxacin), thereby reducing their activity. Fluoroquinolones that have modifica-
tions on the piperazinyl structure (e.g., levofloxacin or moxifloxacin) are not
affected. Importantly, this variant enzyme retains its aminoglycoside-modifying
activity, thus creating a single protein that can affect two different classes of antibi-
otic. Like qnr, the aac(6′)-1b-cr gene is usually found in a cassette as part of an
integron in multiresistance plasmids that encode β-lactamases or qnr, is dissemi-
nated worldwide, and can also be found on the chromosome of some bacteria (sum-
marized in [298]).
The most recently identified class of plasmid-borne resistance determinants are
genes encoding fluoroquinolone efflux pumps. The first of these was oqxAB, identi-
fied in E. coli isolates of agricultural origin and which encodes an RND family
efflux pump with a broad substrate range [299–301]. This was later found in a range
of human, animal, and environmental isolates [302–305]. Intriguingly, the oqxAB
pump genes are found on the chromosome in K. pneumoniae, including drug-­
resistant human clinical isolates [306], and this appears to be the likely reservoir/
origin of the plasmid-borne version [307]. Typical of chromosomally encoded RND
family pumps, upregulation of OqxAB expression in K. pneumoniae requires muta-
tions in the oqxR regulatory gene [308]; however expression from plasmid-borne
oqxAB genes is constitutive, and therefore, this was the first report of a ­constitutively
expressed, mobile plasmid-borne efflux pump [307]. The second was qepA, identi-
fied in 2008 an E. coli clinical isolate and which encodes a member of the major
facilitator efflux pump superfamily [309]. Another variant qepA2 has also been
described [310]. Of concern, qepA genes often reside on mobile elements with
genes encoding ribosomal methyltransferases which mediate resistance to amino-
glycosides [311], again linking fluoroquinolone resistance with resistance to other
antibiotic classes.

4.2.3.4 Interplay of Resistance Mechanisms

Efflux and possibly lowered porin levels reduce susceptibility to fluoroquinolones

in Gram-negative clinical isolates. Efflux upregulation in isolation may however
cause only modest shifts in susceptibility. In cases of higher-level resistance, muta-
tions in the QRDR regions are also found along with upregulation of efflux. In many
Gram-negative pathogens, such as E. coli, target-based resistance progresses from
single mutations (e.g., encoding alteration at Ser83 of GyrA), which cause only
small shifts in susceptibility, to accumulation of multiple target mutations leading to
high-level resistance. The accumulation of mutations depends on stepwise enrich-
ment of mutants, in turn depending in part on the level of quinolone being within the
4 Resistance of Gram-negative Bacilli to Antimicrobials 107

mutant selection window, defined as being between the concentration required to

block the growth of 99% of bacteria in culture (MIC99) and the MIC of the least
susceptible next step mutant, (termed the mutant prevention concentration (MPC)
[312]. Since high-level resistance requires two or more target mutations, fluoroqui-
nolone levels higher than the MPC can only select the simultaneous double mutants
from a wild-type background at an extremely low frequency (approximately 10−12).
The relatively rapid emergence of target-based high-level resistance to fluoroquino-
lones in the clinic in organisms such as P. aeruginosa and E. coli and a general
association with efflux suggest that a key role for efflux may be in enhancing sur-
vival of first step target mutants which then rapidly accumulate additional mutations
conferring stable high-­level resistance. Factors such as suboptimal exposure to drug
can contribute to enrichment of earlier stage mutants, and in cases where certain
target mutations confer clinical resistance levels only in conjunction with efflux, the
presence of pumps would be very important to this process. A hollow fiber model
used to simulate human drug treatment with E. coli lends support to this notion in
that mutants with a two- to eightfold shift in susceptibility due to upregulation of
AcrAB-TolC emerged first, followed by emergence of target mutations (single or
double) [313]. The emergence of target mutations was also strongly delayed in a
strain lacking AcrAB-TolC function. Efflux upregulation regressed after the emer-
gence of target mutations suggesting that once target-based resistance was estab-
lished, efflux upregulation may no longer be required and may revert back to
wild-type expression.
In the case of P. aeruginosa, efflux has been shown to provide a significant con-
tribution to establishing the intrinsic susceptibility to fluoroquinolones, which cor-
respondingly enhances the ultimate levels of resistance caused by target QRDR
mutations [314]. Upregulation of various pumps including MexAB-OprM, MexCD-­
OprJ, or MexEF-OprN can also provide substantial shifts in fluoroquinolone sus-
ceptibility without target mutations [314, 315]. The combination of target mutations
and efflux in P. aeruginosa can mediate very high-level resistance [314]. Pump
upregulation can occur at very high frequencies at lower compound levels since in
many cases, this requires only loss-of-function mutations in regulatory genes like
mexR (MexAB-OprM) or nfxB (MexCD-OprJ). Selection of resistance in P. aerugi-
nosa in vitro at 4X MIC of levofloxacin occurred at 10−6–10−7, whereas the fre-
quency at 4X MIC for an efflux-defective strain was <10−11, indicating that selection
of resistance in the absence of efflux even at relatively modest multiples of the MIC
could be rare [314]. This suggests overall that efflux was even more of a factor in
the emergence of target-based resistance in P. aeruginosa, consistent with the very
rapid rise in fluoroquinolone resistance seen in P. aeruginosa in the United States
after widespread fluoroquinolone use began, and the association of this with resis-
tance to multiple antibiotics [316, 317]. The use of an efflux pump inhibitor to
assess the prevalence of pump-mediated fluoroquinolone and multidrug resistance
among P. aeruginosa clinical isolates also suggested a correlation between fluoro-
quinolone treatment and the co-emergence of target and pump-mediated multidrug
resistance [276]. More recently, associations were seen in clinical isolates between
target mutations and efflux pump upregulation; however the expression levels of
108 C. R. Dean et al.

several pumps could not be correlated with higher-level resistance seen for some
isolates that also harbored QRDR mutations, suggesting that other as yet unidenti-
fied factors may mediate higher-level resistance in some QRDR mutants [275].
Similar interplay between plasmid-borne resistance mechanisms and other mecha-
nisms is likely also occurring, both in terms of determining susceptibility and in
facilitating the emergence of high-level resistance. Qnr proteins only cause a mar-
ginal shift in quinolone susceptibility [284] (Table 4.4), but this is additive with
target-based or other mechanisms and will contribute to the emergence of higher-­
level clinically relevant resistance [295]. Like qnr, the level of resistance conferred
by aac(6′)-1b-cr alone was modest; however, more significant levels of resistance
were observed when aac(6′)-1b-cr and qnrA were found together (Table 4.4).
Furthermore, the presence of plasmid-borne aac(6′)-1b-cr in E. coli resulted in a
greater recovery of resistant mutants during selection experiments with ciprofloxa-
cin [298], essentially by widening the mutant selection window. This again high-
lights the interplay of determinants such as aac(6′)-1-cr and qnrA in the stepwise
acquisition of clinically significant resistance [318]. An additional factor that may
have contributed to the emergence of high-level fluoroquinolone resistance is that
DNA damage and interference with DNA replication caused by quinolones induce
the SOS response, leading to upregulation of error-prone DNA polymerases.
Evolution of quinolone resistance in E. coli in vitro and in an animal model of infec-
tion was curtailed in mutants lacking the SOS response [319].

4.2.3.5 New Strategies for Targeting Type II Topoisomerases

Delafloxacin (Melinta Therapeutics) [320–322] is a new structurally unique anionic

fluoroquinolone that was recently approved by FDA for the treatment of acute bac-
terial skin and skin structure infections (ABSSSI) and is in clinical trials for
community-­ acquired pneumonia and complicated urinary tract infection.
Delafloxacin is particularly potent against Gram-positive pathogens but is also
active against several Gram-negative pathogens including H. influenzae,
Enterobacteriaceae spp., and P. aeruginosa. Delafloxacin targets DNA gyrase and
topoisomerase IV equally, which may reduce the emergence of resistance. Other
efforts to discover novel agents that circumvent target-based or other fluoroquino-
lone resistance mechanisms in Gram-negative pathogens include the design of com-
pounds referred to as novel bacterial type II topoisomerase inhibitors (NBTIs) that
engage the GyrA/ParC targets via a mode of inhibition distinct from fluoroquino-
lones and that are not affected by QRDR mutations [133]. These compounds have
activity against Gram-negative pathogens including E. coli and P. aeruginosa. These
NBTIs also seem to benefit from balanced inhibition of both gyrase and topoisom-
erase IV targets, thereby requiring at least two target mutations in E. coli in order to
observe decreased susceptibility [323]. Efforts to design novel inhibitors of the
ATPase function of type II topoisomerases, in order to exploit DNA replication as a
target but avoid QRDR-mediated resistance, have resulted in potent antibacterial
compounds [324, 325]. Similarly, novel spiropyrimidinetrione agents with a mode
4 Resistance of Gram-negative Bacilli to Antimicrobials 109

of action distinct from fluoroquinolones that may involve targeting GyrB (AZD0914,
now ETX0914, Entasis Therapeutics) are in clinical trials and may find utility in
treating infections due to Gram-positive and/or fastidious Gram-negative patho-
gens, such as N. gonorrhoeae [326–328]. Another class of novel inhibitors of GyrB,
which bind to the TOPRIM domain and are not affected by fluoroquinolone resis-
tance mutations, has recently been described [329]. A detailed discussion of non-­
quinolone inhibitors of topoisomerases is presented in Chap. 19.

4.2.4 Tetracyclines

Tetracyclines are bacteriostatic and prevent bacterial growth by binding to the ribo-
some, thereby blocking protein synthesis. They bind the A-site of the ribosomal 30S
subunit which prevents the entrance of aminoacyl-tRNAs into the mRNA-ribosome
complex, ultimately preventing incorporation of amino acids into the newly emerg-
ing polypeptide [330–332]. The ribosomal target is relatively conserved in bacteria,
and tetracyclines can therefore have a broad spectrum of antibacterial activity, cov-
ering many Gram-positive, Gram-negative, anaerobic, and atypical pathogens. The
original tetracycline, chlortetracycline (also referred to as Aureomycin), is a natural
product produced by Streptomyces aureofaciens and was identified in the late 1940s
by Benjamin Duggar at Lederle Laboratories [333]. Over time other natural exam-
ples were discovered, and routes for making semisynthetic tetracyclines were devel-
oped. The latter allowed detailed exploration of this chemical scaffold, leading to
second-generation tetracyclines doxycycline and minocycline and culminating with
third-generation tetracyclines omadacycline and the glycylcycline tigecycline [334]
which has now been in clinical use for over 10 years. Tigecycline (Tygacil®, Pfizer
Inc.) is approved in the United States and Europe for the treatment of complicated
skin and intra-abdominal infections and in the United States for community-­
acquired bacterial pneumonia. More recently eravacycline (TP-434), a fully syn-
thetic fluorocycline of the tetracycline class, has completed a Phase II study in
complicated intra-abdominal infection (cIAI) and is currently undergoing Phase III
studies in both cIAI and complicated urinary tract infection (cUTI) (www.clinical-
trials.gov) [335]. The latter two compounds are of particular interest in that they
have a broader spectrum of antibacterial activity and largely evade the tetracycline-­
specific, acquired resistance mechanisms of MFS efflux and ribosomal protection
[336], described below in Sects. 4.2.4.1 and 4.2.4.2. Examples of tetracycline chem-
ical structures are shown in Fig. 4.6.
Tetracyclines have now been in use for several decades in human and veterinary
medicine as well as in agriculture. Correspondingly, resistance to earlier-generation
tetracyclines became fairly widespread some time ago [337–339]. There are two
main tetracycline-specific mechanisms of resistance in Gram-negative pathogens:
tetracycline-specific active efflux and ribosomal protection. Additional mechanisms
are active-site rRNA mutations and tetracycline-­modifying enzymes. Efflux by
broad specificity RND family pumps (described in Sect. 4.2.1) also affects suscep-
110 C. R. Dean et al.

Fig. 4.6 Example of tetracyclines (left side) and glycylcyclines (right side). The key modification
at the 9 position (tert-butyl-glycylamido) differentiating the glycylcycline scaffold is depicted in
blue for tigecycline

tibility and contributes to defining intrinsic susceptibility to tetracyclines in differ-

ent Gram-negative pathogens. Examples of tetracycline-­ specific resistance
determinants are listed in Table 4.5 and are discussed below. To a large extent, the
impacts of tetracycline-specific efflux and ribosomal protection have been circum-
vented by the third-generation compound tigecycline, currently in clinical use, and
the fluorocycline eravacycline, and so these mechanisms have a comparatively more
important effect on earlier-generation tetracyclines such as minocycline.

4.2.4.1 Efflux

Tetracycline-Specific MFS Family Efflux Pumps

There are many tetracycline-specific efflux pumps described for both Gram-positive
and Gram-negative bacteria [340], and they function by actively extruding tetracy-
cline from the cell and preventing accumulation to a level sufficient to fully inhibit
the ribosome. Examples of pumps that are commonly found in Gram-negative
pathogens are listed in Table 4.5, and an updated table of the distribution of these
4 Resistance of Gram-negative Bacilli to Antimicrobials 111

Table 4.5 Examples of tetracycline-specific resistance determinants found in selected Gram-­

negative pathogens
Organism Genetic determinant Mechanism
Acinetobacter tet(A) tet(B) tet(G) tet(H), tet(L), tet(39), Efflux
tet(Y)
tet(M) tet (O) tet(W) Ribosomal protection
Klebsiella tet(A-E) tet(L) Efflux
tet(M) tet (S) tet(W) Ribosomal protection
tet(X) Enzymatic modification
Enterobacter tet(A-D), tet(G), tet(L), tet(39) Efflux
tet(M) Ribosomal protection
tet(X) Enzymatic modification
Escherichia tet(A-E), tet(G) tet(J) tet(L), tet(Y), Efflux
tet(M) tet(W) Ribosomal protection
tet(X) Enzymatic modification
Haemophilus tet(B)tet(K) Efflux
tet(M) Ribosomal protection
Data extracted from https://faculty.washington.edu/marilynr/tetweb1.pdf and http://faculty.wash-
ington.edu/marilynr/tetweb2.pdf. These sites maintain a comprehensive list of the mechanisms
and their distribution among Gram-negative bacteria, and the reader is directed there for additional
details

genes among Gram-negatives is maintained at https://faculty.washington.edu/mari-

lynr/tetweb1.pdf and http://faculty.washington.edu/marilynr/tetweb2.pdf. The TetA
efflux pump is perhaps the most broadly distributed among important Gram-­negative
pathogens, and indeed plasmid-encoded TetA was the first bacterial antibiotic efflux
pump identified in 1980 [341, 342]. Generally, tetracycline-specific efflux pumps
genes reside on mobile genetic elements and are thus horizontally acquired resis-
tance mechanisms. The genes encoding TetA and TetB efflux pumps were later
identified in natural oxytetracycline-producing Streptomyces as mechanisms pro-
tecting the producer organisms from the effects of the tetracyclines they were pro-
ducing [343] suggesting this is the likely original source of this resistance
mechanism. Genes such as tetA are commonly used as antibiotic selection markers
for genetic engineering in bacteria, highlighting the effectiveness with which they
can confer resistance to first-generation tetracyclines. The genes encoding
tetracycline-­specific pumps are usually accompanied by the tetR gene, which
encodes a tetracycline-responsive repressor that controls expression of the TetA
efflux pump [344–347]. Unbound TetR functions as a repressor of tetA expression
by binding to tandem operator sequences upstream of tetA as a homodimer and
blocking expression [348]. Upon binding tetracycline, TetR dissociates from the
DNA, allowing transcription to occur.
The tetracycline-specific transporters, typified by TetA, are located in the bacte-
rial inner (cytoplasmic) membrane, and those found in Gram-negative bacteria are
about 46 kDa in size and have 12 transmembrane spanning regions [341]. They
belong to the major facilitator superfamily (MFS) and are tetracycline-H+ antiport-
112 C. R. Dean et al.

ers that operate through the exchange of a proton for the tetracycline molecule
which drives transport of the tetracycline against a chemical concentration gradient,
in this case from the cytoplasm across the inner membrane into the periplasmic
space between the inner membrane and OM. Single-component pumps like TetA,
located in the inner membrane, are generally thought to be more effective at extrud-
ing compounds from the cytoplasm than are pumps of the more broadly active RND
family. The latter are generally thought to recognize compounds in the periplasm or
when diffusing into the inner membrane. This is important, since in Gram-negative
bacteria, the single-component pumps cannot extrude antibiotics completely out of
the cell into the surrounding milieu but will deposit the compound into the periplas-
mic space between the inner and outer membranes. This can concentrate the tetra-
cycline in the periplasm, from which it could diffuse back across the inner membrane
into the cell in the absence of additional efflux across the OM. Consistent with this,
higher levels of resistance mediated by pumps like TetA often require interplay with
efflux across the OM by RND family pumps such as MexAB-OprM in P. aerugi-
nosa or AcrAB-TolC in E. coli [6, 24]. These RND family pumps have a broad
substrate specificity which includes tetracyclines, and the combined effect of spe-
cific single pump efflux from the cytosol and subsequent RND-mediated efflux
from the periplasm to the outside of the bacterium can lead to high levels of resis-
tance in some Gram-negative pathogens.
Although the TetA MFS family efflux pumps are currently widespread among
clinical isolates, presumably driven by the extensive use of earlier-generation tetra-
cyclines, there have been significant advancements in circumventing the impact of
these pumps with each subsequent generation of tetracyclines. Understanding the
emergence of clinical resistance to early generations of tetracyclines, along with
improved understanding of tetracycline mechanism of action, was a key driving
force for renewed interest in developing new tetracyclines that were not subject to
this mechanism. Correspondingly, efforts leading to the identification of tigecycline
were specifically directed toward achieving cellular activity against tetracycline-­
resistant bacteria [334, 349–351], including those expressing tetracycline efflux
pumps. That this was achieved with tigecycline is shown by the potent antibacterial
activity in broad susceptibility testing with resistant clinical isolates that supported
clinical development as an “expanded-spectrum” antibiotic for treatment of
multidrug-­resistant Gram-negative infections (excluding P. aeruginosa) [349, 352–
359]. Specifically showing that tigecycline circumvents resistance mediated by
these pumps, expression of Tet(A), Tet(B), or Tet(X) in a susceptible E. coli strain
background conferred very high levels of resistance to tetracyclines (MIC ≥ 128 μg/
mL) but had a much smaller or no impact on the third-generation tigecycline (or
eravacycline (TP-434)), depending on the pump [360]. Furthermore, no correlation
was seen between the presence of tetracycline-specific efflux genes tet(A) to tet(E)
and insusceptibility to tigecycline in strains of Enterobacteriaceae [353]. It should
be noted though that some variation in amino acid residues important for recogni-
tion of tetracyclines has been reported for TetA proteins expressed from tet(A) genes
residing on different genetic elements and this can have a modest effect on how
much susceptibility is shifted when the pump is expressed [360, 361]. Whether this
4 Resistance of Gram-negative Bacilli to Antimicrobials 113

portends the selection of mutations in tetracycline pump genes over time that
increase recognition of tigecycline remains to be seen.

Efflux Mediated by RND Family Efflux Pumps

Tetracyclines are substrates of several RND family pumps [3] including AcrAB-­
TolC in E. coli and K. pneumoniae and the MexAB-OprM, MexXY-OprM, and
MexCD-OprJ pumps in P. aeruginosa. RND efflux pumps are therefore important
for determining the Gram-negative spectrum of these compounds. Second-­
generation compounds such as doxycycline and minocycline possess a broader anti-
bacterial spectrum than tetracycline, most notably against Acinetobacter,
Burkholderia, and Stenotrophomonas, but their activities against P. aeruginosa and
most species of Enterobacteriaceae are still limited, in large part due to RND-­
mediated efflux [362]. Tigecycline has an expanded spectrum, covering a range of
Gram-negative pathogens including Enterobacteriaceae and non-fermenters such
as Acinetobacter, Stenotrophomonas, and Burkholderia [363]. Therefore basal-­
level RND-mediated efflux alone does not exclude these organisms from the spec-
trum of tigecycline. However, RND-mediated efflux is a factor excluding P.
aeruginosa (MexAB-OprM, MexXY-OprM) [40] and Proteus mirabilis (AcrAB)
[41] from the spectrum of tigecycline. Since tigecycline is a substrate of AcrAB-­
TolC present in many Gram-negative pathogens within the spectrum of tigecycline,
RND-mediated efflux also posed a threat as a resistance mechanism, either via
mutational upregulation of pump expression or indirectly by exacerbating other as
yet unknown mechanisms. Supporting this, RND-mediated efflux has been impli-
cated as a determinant of resistance in laboratory and clinical isolates of Morganella
morganii [364], K. pneumoniae [42, 365, 366], E. coli [43], Enterobacter cloacae
[367]/E. aerogenes [368], and Salmonella enterica [369]. As well, another RND
family pump, OqxAB, may play a role in decreasing susceptibility to tigecycline in
K. pneumoniae [366], and the AdeABC [61, 370, 371], AdeFGH [372, 373], and
AdeIJK [374] efflux pumps have been associated with decreased susceptibility to
tigecycline in A. baumannii.

4.2.4.2  ibosomal Protection, Target Mutations, and Tetracycline-­

R
Modifying Enzymes

Resistance to some tetracyclines (first and second generation) can be caused by the
action of ribosomal protection proteins (RPPs). Several of these proteins have been
identified (e.g., Tet(B), Tet(O), Tet(M), Tet (S), Tet(Q), Tet(W)) [375] (Table 4.5),
and a comprehensive update on the distribution of these determinants in Gram-­
negative bacteria is maintained at http://faculty.washington.edu/marilynr/tetweb2.
pdf. The best studied of these proteins are Tet(O) and Tet(M) [375]. Plasmid-borne
tet(O) was first identified in a Campylobacter jejuni clinical isolate [376] and later
in Campylobacter coli. A similar gene, designated otr(A), was identified in the
114 C. R. Dean et al.

tetracycline-­producing organism Streptomyces rimosus [377], suggesting that, like

tetracycline-specific efflux, ribosomal protection likely originated as a mechanism
to protect the tetracycline-producing organisms and has spread on mobile genetic
elements. Genes encoding Tet(M) and Tet(Q) also occur on mobile elements [338].
Ribosomal protection proteins are generally conserved GTPases that resemble the
elongation factor EF-G (and to a lesser extent EF-Tu) [378, 379]. Early studies
based on chemical probing and cryoelectron microscopy indicated that RPPs bind a
similar site on the 50S ribosomal subunit as EF-G and subsequently cause the tetra-
cycline to be released from the ribosome [375, 380–382]. It is thought that hydroly-
sis of GTP is not strictly necessary for causing the release of tetracycline from the
ribosome but is required for RPP dissociation from the ribosome. Although the
RPP-binding site is removed from the tetracycline-binding site on the 30S ribo-
somal mRNA, it was originally hypothesized that RPP binding caused an overall
conformational shift in the ribosome sufficient to dislodge bound tetracycline [380]
and stimulate binding of tRNA to the A-site which also reduced rebinding of tetra-
cycline [379]. More recent cryoelectron microscopy and modeling of Tet(O) and
Tet(M) bound to the 70S ribosome suggest that bound RPPs may also intrude
directly into the binding site of tetracyclines located around residue C1054 of the
16SrRNA [383, 384]. Ribosomal protection confers resistance primarily to tetracy-
cline, doxycycline, and minocycline, but as is the case for tetracycline-specific
efflux, ribosomal protection has been circumvented by the third-generation tetracy-
clines including tigecycline, omadacycline, and eravacycline [360, 385, 386].
Target-based (active-site) resistance to tetracyclines is relatively rare but does occur.
Mutations in the rRNA target were initially reported in the Gram-positive
Propionibacterium acnes (G1058C) [387] and later in Helicobacter pylori [388,
389]. Resistance to tetracycline in Neisseria gonorrhoeae can be mediated by a
mutation in rpsJ, encoding Val57Met substitution in the 30S ribosomal protein S10
[390]. This mechanism was later found in K. pneumoniae KPC-2-producing clinical
isolates and associated with reduced susceptibility to tigecycline [391, 392].
The first identified tetracycline-modifying enzyme, TetX, was encoded on trans-
posons isolated from Bacteroides fragilis [393], and several more have been identi-
fied since then (see Table 4.5). These enzymes are monooxygenases that act by
hydroxylating the tetracycline, interfering with the tetracycline magnesium-chelat-
ing properties which are needed for ribosome binding [393, 394]. The hydroxylated
tetracycline is also less stable and can then decompose. These enzymes interact with
the central core of the tetracycline molecule, explaining why they can act on all
tetracyclines including third-generation compounds [395, 396]. Nonetheless they
appear to be less effective in conferring resistance to third-generation tetracyclines
[360]. These enzymes have not emerged or spread as a major source of resistance
yet, particularly to tigecycline, but should be monitored in the clinic.
4 Resistance of Gram-negative Bacilli to Antimicrobials 115

4.2.4.3  vasion of Tetracycline-Specific Resistance by Third-Generation

E
Glycylcyclines (Tigecycline)

Third-generation tetracyclines were developed in direct response to the emergence

and spread of resistance [350], and this effort led to the synthesis of a novel class of
glycyl-substituted C9-aminotetracyclines that are referred to as glycylcyclines
[334]. One of these, GAR-936, now tigecycline, bears a t-butyl amine substitution
and is very potent against a broader range of Gram-positive and Gram-negative
bacteria than tetracycline. Moreover, it evades the two main categories of acquired
resistance to tetracycline, tetracycline-specific efflux and ribosomal protection.
This is partly because tigecycline binds the ribosome with a much higher affinity
(10–100-fold higher) than does tetracycline, and this is reflected in more potent
inhibition of translation as measured using in vitro translation assays [397–399].
The basis for the improved affinity of tigecycline is an additional stacking interac-
tion between the 9-t-butylglycylamido portion of tigecycline (C-9 moiety) and the
C1054 nucleobase of the 16S rRNA [398]. There is also additional steric clash
between tigecycline and the anticodon stem loop of the A-site tRNA compared to
tetracycline, making it more effective in preventing tRNA entry into the A-site. This
is likely important for the overall improved potency against a broader range of
Gram-negatives than tetracycline (i.e., overcoming intrinsic resistance). The C-9
moiety also may enhance the target on-rate of tigecycline and sterically clash with
important residues of the RPP TetM (within loop 3 of domain 4 of TetM) that inter-
act with C1054, thus preventing TetM from dislodging tigecycline from the ribo-
some. Therefore the C-9 moiety itself is likely preventing TetM and other RPPs
from conferring resistance rather than this being a function solely of higher tigecy-
cline binding affinity [398]. As mentioned above, tigecycline also appears to escape
recognition by tetracycline-specific efflux pumps, shown using TetB containing
vesicles [400]. Tigecycline may also be less effective as an inducer of tetracycline
efflux pump expression [401].

4.2.4.4 Mechanisms of Tigecycline Resistance Emerging in the Clinic

Resistance to early-generation tetracyclines in the clinic emerged rapidly after their

introduction in the late 1940s, and the epidemiology of tetracycline resistance is
described in Chap. 10. As described above, third-generation tetracyclines, exempli-
fied by tigecycline, are able to largely overcome established resistance by circum-
venting tetracycline efflux and/or ribosomal protection. Tigecycline entered the
clinic in 2005 (often used as a last line of defense in treating MDR isolates), and
since it was refractory to the main resistance mechanisms, it was not clear what
mechanisms of resistance would emerge in clinical use, although it seemed likely
that RND efflux would play a role. The Tigecycline Evaluation and Surveillance
Trial (TEST) is an ongoing global study to monitor in vitro susceptibility to tigecy-
cline and other antibiotics in MDR isolates. The most recent report [130] examined
isolates collected worldwide between 2004 and 2014 and found that rates of MDR
116 C. R. Dean et al.

E. coli ranged from 4% in North America to 18% in Latin America. Approximately

94% of MDR E. coli were resistant to minocycline, but only 0.2% were resistant to
tigecycline, the lowest rate for all antibiotics tested. Tigecycline-resistant E. coli
isolates did not appear in this study until 2008; however eight resistant isolates have
been identified between 2009 and 2014 across a wide geographic range that included
one isolate from North America. For K. pneumoniae, rates of MDR were approxi-
mately 12%. Among those, the rate of tigecycline resistance was higher than seen
for E. coli, at approximately 6%, although this was still the lowest of all the antibiot-
ics tested. A. baumannii had the highest frequency of MDR, at 44% with some
geographic areas having >50% MDR. The lowest rate of resistance among the MDR
isolates was reported for minocycline (13%). Tigecycline resistance breakpoints are
not established for A. baumannii, but tigecycline had a lower MIC90 than minocy-
cline. The most recent report [130] concluded that tigecycline has remained active
against most MDR isolates (excluding P. aeruginosa which has high intrinsic resis-
tance), although this varies by geographical region. The acquisition of tigecycline
resistance in small numbers of E. coli isolates over the course of the collection of
these strains was observed and should be further monitored.
So far, decreased susceptibility to tigecycline in the clinic has been attributed to
upregulation of RND family efflux pumps. For example, resistance has been corre-
lated with upregulation of AcrAB-TolC expression in clinical isolates of E. coli [43,
402], K. pneumoniae [42, 365, 403], and E. cloacae [367] and of AdeABC in A.
baumannii [44, 370]. Interestingly the tetX gene encoding enzymatic modification
was detected in a clinical isolate of A. baumannii from China [44]. Decreased sus-
ceptibility in Salmonella enterica was attributed to the combined activity of a
plasmid-­borne tet(A) gene and mutation in ramR, presumably leading to RND
efflux pump upregulation. There is not always a direct correlation between efflux
and susceptibility however, and mechanisms of resistance to tigecycline may ulti-
mately prove to be more complex as tigecycline is used longer in the clinic. For
example, a recent study of tigecycline-resistant A. baumannii found involvement of
AdeABC efflux but also uncovered a potential role for mutational disruption in the
trm methyltransferase gene in resistance in clinical isolates [404]. Finally, in vitro
tigecycline selection studies using strains harboring the well-characterized tetracy-
cline resistance genes tet(A), tet(K), tet(M), and tet(X) selected for mutations in
these genes that increased the ability of the encoded proteins to act on tigecycline
[405]. Since these genes are widespread in clinical isolates, it will be of interest to
see if this occurs in the clinic going forward.

4.2.4.5  ovel Agents and New Approaches: Circumvention

N
of Tetracycline-Specific Resistance Mechanisms in Third-­
Generation Tetracyclines (Glycylcyclines)

Current efforts in the search for next-generation tetracyclines are largely being done
by Tetraphase Pharmaceuticals, specifically centered on the fully synthetic fluoro-
cyclines eravacycline, TP-271, and TP-6076. Eravacycline (currently in Phase III)
4 Resistance of Gram-negative Bacilli to Antimicrobials 117

Fig. 4.7 Chemical structure of key aminoglycosides used in the clinic. The positions of covalent
chemical modification by aminoglycoside-modifying enzymes are shown in blue

is generally more potent overall and has a slightly better spectrum than tigecycline,
but also does not cover P. aeruginosa. TP-271 has a much more limited spectrum
and is directed at the bacterial pathogens responsible for community-acquired pneu-
monia [406]. TP-6076 has potent activity against a range of pathogens including A.
baumannii and carbapenem-resistant Enterobacteriaceae and has entered Phase I
trials as of this writing. It was also selected for funding support from CARB-X
(www.carb-x.org).

4.2.5 Aminoglycosides

Aminoglycosides are one of the major classes of antibiotics used in the clinic to
treat Gram-negative bacillary infections. The most widely used aminoglycosides
are tobramycin, gentamicin, and amikacin, mainly for the treatment of P. aerugi-
nosa meningitis and pneumonia. Tobramycin is also used in two inhaled formula-
tions (TOBI ® Podhaler ®, Novartis) for the treatment of chronic P. aeruginosa
infections in cystic fibrosis patients (Fig. 4.7). Streptomycin, neomycin, and kana-
mycin are used for the treatment of infections caused by E. coli, Proteus species,
Enterobacter aerogenes, K. pneumoniae, Serratia marcescens, and Acinetobacter
species [407–410]. Aminoglycosides act by binding to bacterial ribosomes and
therefore blocking bacterial protein synthesis. They mainly bind to the aminoacyl-­
tRNA recognition site (A-site) of the 16S ribosomal RNA of the 30S ribosome [331,
411–414]. This binding causes codon misreading and the corresponding introduc-
tion of incorrect amino acids in the growing polypeptide. This amino acid mis-­
incorporation causes rapid cell death [415]. All aminoglycosides are bactericidal
and have a prolonged postantibiotic effect due to the extended time needed to
118 C. R. Dean et al.

recover from protein synthesis inhibition [416–419]. Aminoglycosides have also

been shown to act synergistically in vitro with other classes of antibacterials, in
particular β-lactams [420–423]. These findings have encouraged the use of amino-
glycoside in combination with β-lactams for the treatment of several infections in
the clinic, especially for nosocomial infections caused by P. aeruginosa [410, 424–
427]. In the past few years, retrospective studies looking at mortality outcomes of
patients with carbapenem-resistant Enterobacteriaceae (CRE) have shown an
improved outcome when aminoglycosides were used in combination therapy with
β-lactams [428, 429] or tigecycline [430]. The clinical utility of aminoglycosides is
imperiled by high rates of resistance often in conjunction with resistance determi-
nants to other drugs used to treat Gram-negative infections [409, 431–436].
Aminoglycoside resistance in the clinic occurs via a number of different mecha-
nisms: mutations altering the target (rRNA or ribosomal proteins), transport defects,
efflux and, importantly, modifying enzymes that inactivate the drug [148, 409, 437].
Chromosomal mutations of the target are very rare in Gram-negative bacilli mainly
due to the high number of copies of the 16S rRNA [409, 437]. Reports of target-site
mutations in clinical isolates have been limited to Mycobacterium spp. [438, 439]
and Borrelia burgdorferi [440], and so will not be addressed further. Each of the
remaining mechanisms is discussed below, and the epidemiology of aminoglyco-
side resistance is discussed in Chap. 10.

4.2.5.1 Aminoglycoside-Modifying Enzymes

Aminoglycoside-modifying enzymes (AMEs) inactivate the drug by covalent

chemical modification, reducing the binding affinity of aminoglycosides to their
target. AMEs are the major resistance determinant in the clinic for this class of anti-
biotic and often encoded on plasmids harboring multiple resistant elements to mul-
tiple antibiotic classes [148, 409, 437, 441]. This presence on mobile genetic
elements has enhanced the number of isozymes circulating in pathogenic and non-
pathogenic bacteria. To date, well over 100 AMEs have been described and charac-
terized from clinical isolates as well as from soil-dwelling bacteria that produce
aminoglycosides [148, 442]. AMEs are divided into three major categories based on
the specific chemical modification: N-acetylation (AACs), O-phosphorylation
(AHPs), and O-adenylation (ANTs). These categories are further subdivided into
classes based on their specific site of modification of the aminoglycoside. Variants
of these are further subdivided using roman numerals, and in some cases, a letter is
added when the same position is modified [442, 443]. Table 4.6 shows some of the
major AMEs occurring in the clinic.
The aminoglycoside acetyltransferases (AACs) are the major class among
these modifying enzymes. These enzymes are acetyl CoA-dependent, and they
acetylate various amino groups found on the aminoglycoside structure [442, 443,
454, 455]. The most common AACs in Gram-negative bacteria are AAC(6′)-I,
AAC(3)-IIa, and AAC(3)-I. Recently, a broadening in activity spectra for some
of these enzymes has also been observed. AAC(6′)-Ib-cr has acquired the ability
4 Resistance of Gram-negative Bacilli to Antimicrobials 119

Table 4.6 Major aminoglycoside-modifying enzymes present in clinical isolates and their
resistance profile
Type Enzymes Resistance conferred References
Aminoglycoside AAC(6′)-I Tobramycin, amikacin, [441, 444]
acetyltransferases (AACs) netilmicin, dibekacin, sisomicin,
kanamycin, isepamicin
AAC(3)-IIa Tobramycin, gentamicin, [441, 445]
netilmicin, dibekacin, sisomicin
AAC(3)-I Gentamicin, sisomicin, [441, 446]
fortimicin
AAC(6′)-Ib-cr Kanamycin, amikacin and [447, 448]
tobramycin, ciprofloxacin, and
norfloxacin
Aminoglycoside APH (3′)-Ia Kanamycin, neomycin, [409, 431–436,
phosphotransferases (APHs) streptomycin, lividomycin, 441, 449, 450]
paromomycin, and ribostamycin
APH (3″)-III Kanamycin, neomycin, [441, 451]
lividomycin, paromomycin,
butirosin, and ribostamycin
Aminoglycoside ANT(2″) Tobramycin, gentamicin, [441, 452]
nucleotidyltransferases dibekacin, sisomicin, kanamycin
(ANTs) ANT(4′) Tobramycin, amikacin, [441, 453]
dibekacin, kanamycin,
isepamicin

to modify fluoroquinolones by acylation of the secondary amine of the pipera-

zine ring of the antibiotic present on ciprofloxacin and norfloxacin but not levo-
floxacin [447, 448]. The aminoglycoside phosphotransferases (AHPs) and
nucleotidyltransferases (ANTs) are both ATP-dependent enzymes. APHs phos-
phorylate the hydroxyl groups on the aminoglycoside similar to ATP-dependent
kinases, sharing high similarity with serine-threonine eukaryotic kinases [442,
443]. ANTs utilize ATP as AMP donor that is added on the aminoglycoside
hydroxyl groups. The major representatives of this class present in the clinic are
ANT(2″) and ANT(4′) described in Table 4.6 [409]. Bifunctional enzymes with
a broader spectrum of activity have been reported. ANT(3″)-Ii/AAC(6′)-IId that
confers resistance to streptomycin, spectinomycin, and gentamicin has been iso-
lated from Serratia marcescens, a human enteropathogen [456, 457]. Among the
aminoglycoside-modifying enzyme genes, aac(6′)-Ib was the most prevalent
(37.5% of isolates were positive), in a study looking at 200 Gram-negative bacilli
resistant to aminoglycosides [409]. In another study from Spain of 330 amino-
glycoside resistant Enterobacteriaceae isolates, the predominant resistance
determinant was Aph(3″)-Ib (65.4% of isolates were positive) in accordance with
the observed streptomycin resistance phenotype [409, 432].
120 C. R. Dean et al.

4.2.5.2 Ribosomal Protection

Posttranscriptional methylation of 16S rRNA by aminoglycoside rRNA methyl-

transferases (RMTs) is an emerging resistance mechanism for this class of antibiot-
ics. RMTs modify specific nucleotide residues (N7 position of nucleotide G1405 or
N1 position of nucleotide A1408) of the 16S rRNA, thereby preventing aminogly-
cosides from binding to their target [408, 458]. This mechanism was originally
identified and characterized in antibiotic-producing organisms as a self-protection
mechanism [459] but has now been emerging in several important Gram-negative
nosocomial pathogens [460]. In 2003, aminoglycoside rRNA methyltransferases
were reported in K. pneumoniae (encoded by the armA gene) and P. aeruginosa
(encoded by the rmtA gene) both conferring high-level resistance to 4,6-­disubstituted
deoxystreptamines [461, 462]. After the first identification of these genes, a series
of plasmid-encoded RMTs have been identified (encoded by rmtB1, rmtB2, rmtC,
rmtD, rmtD2, rmtE, rmtF, rmtG, and rmtH) in several clinical isolates [408]. In
2007 an aminoglycoside rRNA methyltransferase (encoded by the npmA gene) was
reported from E. coli isolated in 2003 from the urine of an inpatient in a general
hospital in Japan, which conferred resistance to 4,6- and 4,5-disubstituted
2-­deoxystreptamines [463]. Further information on the class of enzyme including
their origin and the impact on the use of aminoglycoside can be found in a review
published in 2016 by Doi et al. [408]. Even though a low prevalence of this class of
enzyme has been reported in clinical isolates, their ability to confer high-level pan
aminoglycoside resistance in conjunction with their presence on mobile elements
threatens the future use of aminoglycosides.

4.2.5.3 Decreased Permeability and Efflux

As described in Sect. 4.2.1, a major issue in Gram-negative bacilli is the inability of

many drugs to penetrate the cell membrane. However, aminoglycosides are cationic
and therefore can interact with negatively charged LPS to facilitate “self-promoted
uptake” across the OM. This is followed by energy-dependent (electron transport-­
mediated) uptake across the inner membrane. Correspondingly, alterations in LPS
or reductions in uptake across the inner membrane were proposed to play a role in
reducing susceptibility. In the case of P. aeruginosa, this may involve aminoarabi-
nose modification of the lipid A moiety of LPS, controlled by the PhoP-PhoQ two-­
component regulator pair. This system is a well-characterized determinant of
resistance to the polymyxin class of antibiotics (see Sect. 4.2.6), but its involvement
in aminoglycoside resistance is less well understood [464, 465]. Reduced expres-
sion of some oligopeptide transporters, such as OppA, may also reduce entry of
aminoglycosides [437].
Efflux by RND family pumps has been shown to play a significant role in
aminoglycoside extrusion and therefore resistance in several pathogens. In E.
coli, the AcrAD pump has been shown to be able to efflux aminoglycosides [466,
467]. Other pumps involved in aminoglycoside efflux are AmrAB-OprA and
4 Resistance of Gram-negative Bacilli to Antimicrobials 121

BpeAB-­OprB in Burkholderia pseudomallei [466, 468], AdeABC in A. bauman-

nii [76], and MexXY-OprM of P. aeruginosa [469, 470]. The MexXY-OprM
efflux pump is unique among the complement of P. aeruginosa pumps in its abil-
ity to extrude aminoglycosides. It is also induced by agents inhibiting protein
synthesis, contributing to both impermeability and adaptive aminoglycoside
resistance [470, 471]. The latter refers to the induction of reversible resistance by
exposure to aminoglycosides, which is now known to result largely from induc-
tion of MexXY and possibly by a concomitant upregulation of anaerobic respira-
tion genes which may compromise aminoglycoside uptake across the inner
membrane [472]. The regulation (inducibility) of MexXY expression involves
the MexZ repressor and PA5471, a protein of unknown function, which sense
disruptions of protein synthesis (translation) [52, 473]. More recent work showed
the induction of MexXY by aminoglycosides also depends on the two-compo-
nent system AmgRS [474]. This appears to be related to the role of AmgRS as a
cell envelope stress response regulator and, in the case of aminoglycosides, in
responding to incorporation of misfolded proteins in the inner membrane that
results from aminoglycoside action on the ribosome. Mutations in amgS can also
cause constitutive activation of MexXY expression. Mutations in genes encoding
another two-component system ParRS, involved in resistance to polymyxins,
also cause upregulation of MexXY and aminoglycoside resistance [75, 475].
It should be noted that although aminoglycoside-modifying enzymes are gener-
ally the most important resistance mechanism in Gram-negative bacteria, this
does not appear to be the case in P. aeruginosa isolates recovered from CF
patients. Since these patients tend to be colonized by strains common in the natu-
ral environment, there is less chance for these enzymes to accumulate, and there-
fore only a small percentage of CF isolates harbor aminoglycoside-modifying
enzymes [476]. Therefore, efflux by MexXY, likely in conjunction with other
mechanisms, is comparatively more important in this instance.

4.2.5.4 Biofilms

Growth in the biofilm mode is another barrier for the entry of aminoglycosides in
bacteria, contributing to intrinsic and adaptive resistance to this class of antibiotics.
Biofilms are defined as an intertwined community of bacteria adhering on a surface
and surrounded by a self-produced matrix composed of extracellular DNA, pro-
teins, and polysaccharides. Biofilms play a key role in chronic P. aeruginosa infec-
tions and have been associated with pulmonary infections in patients with CF where
aminoglycosides and, in particular, tobramycin are routinely used [477, 478].
Therefore understanding the role of biofilms in relation to aminoglycoside resis-
tance is of high importance. Subinhibitory concentrations of aminoglycosides,
especially tobramycin, have been shown to induce biofilm formation in P. aerugi-
nosa by the induction of the aminoglycoside response regulator (arr) gene. This
gene is postulated to be involved in the regulation of cell surface adhesiveness and
therefore contributes to biofilm-specific aminoglycoside resistance. Some studies
122 C. R. Dean et al.

Fig. 4.8 Structure of novel aminoglycosides currently in clinical development (plazomicin and
arbekacin) and preclinical characterization (TS3112)

have hypothesized the ability of the biofilm matrix to bind aminoglycosides and
therefore drastically reduce their antibacterial activity [479, 480]. Several in vitro
studies have shown co-dosing of an aminoglycoside with a cationic steroid antibi-
otic, CSA-13, or a cationic peptide, DJK-5, is effective in overcoming biofilm-­
mediated resistance [481, 482]. More study to understand the molecular mechanisms
of biofilm induction by aminoglycosides as well as device strategies to inhibit bio-
film formation are needed.

4.2.5.5 Novel Approaches and Treatment Strategies

Better understanding of aminoglycoside resistance, molecular mechanisms and

dosing regimens to minimize toxicity, together with the increase of resistance in the
clinic, has created interest in discovering new approaches or novel aminoglycosides
aimed at overcoming resistance. One approach that has been pursued to extend the
useful lifespan of aminoglycosides is the pursuit of aminoglycoside-modifying
enzyme (AME) inhibitors that would prevent inactivation of aminoglycosides, simi-
lar to the successful combinations of β-lactamase inhibitors with β-lactams. Despite
several three-dimensional structures of all representative members of the three
classes of AMEs [483–492] and better understanding of their molecular mechanism
of action, screening efforts and structure-based drug designs have not yet led to
viable clinical candidates. Labby and Garneau-Tsodikova have reviewed several
efforts aimed at identifying suitable AME inhibitors in a recent review [493].
The goal for new aminoglycosides has been mainly to improve their toxicity
profile, avoid modification by key AMEs, and not be impacted by RMTs widely
spread in clinical isolates. Arbekacin from Meiji Seika Pharma Co. and plazomicin
from Achaogen are the front-runners currently under development (Fig. 4.8).
Arbekacin is a broad-spectrum aminoglycoside active against Gram-positive bacte-
ria, including methicillin-resistant S. aureus, and Gram-negative bacteria, such as P.
aeruginosa, KPC-expressing K. pneumoniae, and ESBL-producing E. coli [494,
495]. Arbekacin is approved in Japan for the treatment of sepsis and pneumonia
caused by MRSA. It is currently in Phase I clinical trials as an inhalation solution
for the treatment of hospital-associated and ventilator-associated bacterial pneumo-
nia (HABP/VABP) and under evaluation for treatment of patients with infections
caused by multidrug-resistant organisms when treatment with other antibiotics can-
4 Resistance of Gram-negative Bacilli to Antimicrobials 123

not be used [494, 495]. Arbekacin is stable to some of the most common APHs,
ANTs, and AACs present in clinical isolates. Plazomicin, a sisomicin analog with
potent activity against Enterobacteriaceae (MIC90 ≤ 2 μg/mL), has recently com-
pleted Phase III trials in cUTI and in patients with serious bacterial infections due
to CRE (http://www.achaogen.com/plazomicin/). Plazomicin, like its parent com-
pound sisomicin, is resistant to several AMEs such as APH(3′)-III, APH(3′)-VI, and
APH(3′)-VII and ANT(4′)). The addition of hydroxyl-aminobutyric acid substitu-
ent at the N-1 position and hydroxyethyl substituent at the 6′ position has rendered
plazomicin resistant to AAC [3], ANT(2″), APH(2″), and AAC(6′) enzymes [409].
Plazomicin activity is abrogated by RMTs that are frequently present on mobile
genetic elements that also carry β-lactamases like NDM-1 in Enterobacteriaceae
[431, 434, 496]. This may turn out to be a liability for the clinical longevity of
plazomicin against Enterobacteriaceae. Meiji Seika Pharma Co. recently reported a
semisynthetic apramycin, named TS3112 (Fig. 4.8), which is active against Gram-­
positive and Gram-negative bacteria producing both AMEs and RMTs. TS3112
showed potent bactericidal activity in a murine thigh model of K. pneumoniae
expressing RTMs [497]. TS3112 is currently in early-stage characterization, show-
ing encouraging in vitro results.
New delivery strategies for aminoglycosides have been adopted in the past few
years that provide higher local concentration at the infection site with a lower total
amount of drug delivered, which reduces systemic exposure and safety liabilities of
this class of drug. The best example of new delivery method for inhaled aminogly-
coside is tobramycin inhalation powder (TOBI® Podhaler®), delivered via the
T-326 inhaler (Novartis). Long-term safety studies in patients with CF have shown
that it is well tolerated, there was no evidence of serum tobramycin accumulation
with successive cycles, and no unexpected adverse events were observed. Further,
the new powder delivery method improved compliance due to shorter administra-
tion time, convenience, and ease of use [498, 499]. Bayer Healthcare, in collabora-
tion with Nektar, is currently developing BAY41-6551, a drug-device combination
of a specially formulated amikacin. BAY41-6551 has recently completed Phase III
as an adjunctive treatment for intubated and mechanically ventilated patients with
Gram-negative pneumonia and showed bactericidal activity against most isolates
tested with amikacin MICs ≤ 256 μg/mL [500–502].
Aminoglycosides are a key class of antibiotic used by physicians to treat serious
infections caused by MDR Gram-negative and Gram-positive pathogens. Continued
characterization of aminoglycoside resistance mechanisms may enable the design
of resistance determinant inhibitors and/or new aminoglycosides capable of circum-
venting these mechanisms. Additionally, efforts aimed at optimization of dosing
regimens and discovery of new delivery strategies should help to maintain and
extend the clinical utility of this important antibiotic class.
124 C. R. Dean et al.

4.2.6 Polymyxins

Polymyxins are an older class of cationic cyclic lipopeptide antibiotics that were
introduced into clinical use in the 1950s (polymyxin B and colistin, also known
as polymyxin E). However, the use of polymyxins declined sharply around the
early 1970s due to concerns of toxicity [503, 504] and the availability of safer
antibiotics. The mechanism by which polymyxins kill bacteria is not fully under-
stood. One mechanistic step that is well established is an initial interaction of the
cationic peptide with negative charges on the lipopolysaccharide (LPS) that
forms the outer leaflet of the Gram-negative OM [505, 506]. This interaction
occurs mainly via negatively charged phosphate residues located on lipid A and
is required for the “self-promoted uptake” of polymyxins into the bacteria.
Binding of polymyxin is thought to displace divalent cations (Mg2+, Ca2+) that
cross-link adjacent LPS molecules, and this can disrupt to some extent the per-
meability barrier of the OM, which also increases uptake of the polymyxin. This
is unlikely to be responsible entirely for cell killing. As discussed in more detail
in Sect. 4.2.6.4 below, derivatives of polymyxin (e.g., polymyxin B nonapeptide)
exist with dramatically reduced antibacterial activity that retain the OM disrup-
tion activity. Polymyxins may ultimately kill via mechanisms that include lysis
of the inner membrane [505] generation of toxic hydroxyl radicals [507] and
inhibition of respiration via targets such as type II NADH-quinone oxidoreduc-
tases (NDH-2) [508]. The interaction of polymyxins with the Gram-negative OM
LPS has two main implications: first, that susceptibility to this class of com-
pound can be decreased by restructuring LPS to reduce its negative charge (dis-
cussed below) and, second, that it limits the spectrum of polymyxins to some but
not all Gram-negative pathogens. This includes the important Gram-negative
ESKAPE pathogens E. coli, K. pneumoniae, P. aeruginosa, and A. baumannii.
Polymyxins were reintroduced into the clinic in the early 2000s as a last-line
therapeutic option to address the emergence of MDR and XDR in these patho-
gens. Not long after this reintroduction into clinical practice, emergence of colis-
tin-resistant strains began to increase, prompting concern about their ongoing
therapeutic usefulness, especially considering that no other options may exist in
scenarios where colistin is being used. The main mechanism by which Gram-
negative pathogens that are susceptible to polymyxins develop resistance is via
alterations in their lipopolysaccharide that reduce its net negative charge, thereby
reducing uptake of polymyxins. This can occur through selection of chromo-
somal mutations in genes involved in regulating LPS remodeling or by horizontal
acquisition of plasmids harboring genes mediating this process. Some Gram-
negative pathogens (e.g., Burkholderia cepacia, Proteus mirabilis, Serratia
marcescens) are not susceptible to polymyxins (intrinsic resistance), because
their LPS always possesses such modifications. These mechanisms are discussed
in the sections below.
4 Resistance of Gram-negative Bacilli to Antimicrobials 125

4.2.6.1  utationally Acquired Resistance Mediated by Reduction

M
of Negative Charge Status of LPS

Gram-negative bacteria have a broad ability to remodel their OM. The best under-
stood and most widespread mechanism that decreases susceptibility to polymyxins
utilizes this capacity by modification of LPS to reduce its negative charge and, con-
sequently, the initial binding step of cationic polymyxin to the cell surface. This is
highly complex and varies among different strains but in general occurs in
Enterobacteriaceae and P. aeruginosa via masking the negatively charged 4′-phos-
phates on lipid A, and to a lesser extent the 1 position phosphate or the 3-deoxy-­d-
manno-octulosonic acid (KDO), by addition of 4-amino-4-deoxy-l-arabinose
(l-Ara4N) [509]. Synthesis and transfer of l-Ara4N are mediated by the products
of the arn locus (e.g., arnBCADTEFpmrE in P. aeruginosa, also known as pmrHFI-
JKLME). This is perhaps the most common mechanism of reducing the negative
charge of LPS described in these organisms. Enterobacteriaceae can also add phos-
phoethanolamine (pEtN) (mainly to the 1 position but also to other locations such
as 4′-lipid A, KDO, or core oligosaccharide) via transferases such as PmrC. The
contribution of pEtN to resistance appears to be smaller than that of l-Ara4N, but
both clearly play a role, and both decorations can occur together. Recently, P. aeru-
ginosa has also been shown to be able to modify its LPS with pEtN when zinc is
present, under the regulation of the ColRS two-component regulatory system [510].
A. baumannii lacks an arn locus and therefore cannot carry out L-Ara4N modifica-
tion but can undergo pEtN modification [526] or, as recently described, galactos-
amine modification [511]. Regulatory control of these modifications is highly
complex and is often mediated by interrelated networks of two-component regula-
tory systems (TCSs). The PmrAB regulator pair controls l-Ara4 and/or pEtN modi-
fication and is widespread in Gram-negative pathogens. Similarly the PhoPQ
system, also present in several organisms such as S. enterica, E. coli, K. pneu-
moniae, and P. aeruginosa (but absent in A. baumannii), is important for control of
LPS modification and can be interconnected with the PmrAB system. For example,
these two regulatory systems are interconnected in S. enterica and E. coli via the
PmrD protein [512]. These systems can upregulate LPS modification (e.g., upregu-
late expression of the arn locus) in response to certain conditions such as magne-
sium limitation or exposure to polymyxin or other cationic peptides [513–515].
Although there is likely some adaptive change in susceptibility to polymyxins medi-
ated by these systems upon drug exposure, or by exposure to cationic peptides in the
host, resistance is generally mutationally acquired, via selection of mutations in the
genes encoding these regulators, which leads to strong constitutive upregulation of
LPS modification. Some mutations in these regulatory genes may also lead to stron-
ger inducibility of the systems by the polymyxin [516]. Regulatory mutations can
be selected in vitro and have also been associated with resistance in the clinical set-
ting, including mutations found in pmrA/pmrB in K. pneumoniae [517–520], P.
aeruginosa [516, 521, 522], and A. baumannii [523–527] and phoP-phoQ in K.
pneumoniae [528–530] and P. aeruginosa [464, 522, 531]. A. baumannii lacks both
phoPQ and an arn locus, so pmrB mutations are frequently found in this pathogen,
126 C. R. Dean et al.

and these mutants will have pEtN modification through activation of the pmrC
transferase gene located in the pmrABC locus [526]. However, mutations in pmrAB
are not always found in colistin-resistant A. baumannii clinical isolates suggesting
that mutations elsewhere on the chromosome can upregulate pmrABC [532].
The importance of the PmrAB and PhoPQ systems in controlling OM remodel-
ing and resistance to polymyxins is well established; however a full understanding
of these phenomena is still forthcoming. Recently in the case of K. pneumoniae,
mutation of mgrB, which encodes a negative feedback regulator of the PhoPQ two-­
component system, was revealed as an important mediator of LPS modification and
colistin resistance [533, 534]. Loss of MgrB function leads to constitutive activation
of PhoPQ and LPS modification. This mechanism appears to be relatively wide-
spread in K. pneumoniae clinical isolates [529, 535, 536] and can be mediated by
insertion of genetic elements that may also carry other resistance genes such as
β-lactamases [537]. The fact that colistin resistance can arise from any loss-of-­
function mutation of mgrB likely explains its relatively high prevelance among
colistin-­resistant K. pneumoniae isolates. Several additional TCSs involved in poly-
myxin resistance have been characterized more recently including the CrrAB (colis-
tin resistance regulon) in K. pneumoniae [530]. Changes in CrrB result in
upregulation of the PmrAB system via CrrR, which then upregulates arn genes and
pmrC, leading to LPS modification and colistin resistance [538]. However, not all
K. pneumoniae harbor the crrAB genes. In P. aeruginosa, three additional TCSs also
known to be involved in modulating susceptibility to polymyxins have been
described: ParRS [539], ColRS [540], and CprRS [541]. ParRS and CprRS partici-
pate in adaptive resistance to polymyxins by upregulating expression from the arn
locus upon sensing polymyxins or other cationic peptides. ColRS and CprRS are
required for high-level polymyxin resistance resulting from mutations in phoPQ
[540]. These interactions appear complex, and mutational analysis also suggested
that additional factors beyond l-Ara4N modification of lipid A could be involved in
resistance in P. aeruginosa, but this remains to be fully elucidated [540]. Mutations
in the parRS genes were subsequently shown to reduce susceptibility to multiple
classes of antibiotic due to coordinately upregulating expression from the mexXY
efflux pump genes and arn and downregulating expression of the oprD porin gene
[75]. For additional details on polymyxin resistance mechanisms, see Jeannot et al.
[542]. Another aspect of resistance relevant to colistin is the phenomenon of hetero-
resistance, which refers to the presence of a substantial stable resistant subpopula-
tion in cultures of isolates that may score as susceptible to an antibiotic by standard
susceptibility testing. Colistin heteroresistance has been described mainly in K.
pneumoniae [528, 543, 544] and A. baumannii [545], and resistant subpopulations
can harbor a range of resistance mutations [543]. Heteroresistant isolates can be
recovered from patients with no prior treatment with polymyxins, and it is expected
that the use of colistin could rapidly enrich for the resistant subpopulation leading
to clinical failures. Heteroresistance can be missed by standard susceptibility tests,
suggesting that the rates of, and potential for, selecting colistin resistance in the
clinic may be underestimated. For more information on heteroresistance, see Chap.
9 in this volume. Finally, as mentioned above, a number of Gram-negative bacteria
4 Resistance of Gram-negative Bacilli to Antimicrobials 127

are intrinsically resistant to polymyxins. These include Burkholderia cepacia com-

plex, Proteus, Serratia, Providentia, and others. These organisms differ from sus-
ceptible strains in that their LPS is always constitutively modified with l-Ara4N or
has other alterations affecting polymyxin binding.

4.2.6.2 Mutations Causing Loss or Reduction of LPS

Synthesis of LPS (in particular the lipid A portion) and assembly of the LPS-­
containing OM are essential for the growth and/or viability of most Gram-negative
pathogens, but there are a few exceptions to this. Neisseria meningitidis, Moraxella
catarrhalis [586, 587] and a subset of A. baumannii have been shown to tolerate
loss of LPS biosynthesis. Indeed, this was uncovered in the case of A. baumannii
during in vitro studies of colistin resistance, where mutations in genes involved in
lipid A biosynthesis were directly selected from A. baumannii strain ATCC 19606
or other strains on polymyxin-containing medium [112, 546]. Point mutations were
initially identified in the lpxA, lpxC, or lpxD genes [112] that encode enzymes cata-
lyzing the first three steps of lipid A biosynthesis [25]. A follow-up experiment
identified mutants where lpxA or lpxC were inactivated by insertion sequence
ISAb11 [546], and subsequently an engineered mutant deleted for lpxC was reported
(described in [547]), confirming that lpxA and lpxC are dispensable in A. baumannii
ATCC 19606, at least under laboratory growth conditions. These mutants lack lipid
A, the target of the initial interaction with polymyxins, and as such are highly resis-
tant to polymyxins but are also highly susceptible to a range of other antibiotics due
to loss of the protective lipid A-containing OM [112]. To date, it appears that this
mechanism may be confined to mutations in genes encoding enzymes occurring
early in the lipid A biosynthesis pathway since inactivation of steps occurring later
in the lipid A biosynthetic pathway (e.g., LpxH [548] or LpxK [549]) causes toxic
accumulation of lipid A synthetic pathway intermediates and is not tolerated.
Furthermore, this mechanism does not apply across all A. baumannii, as only a
subset appears to tolerate loss of lipid A biosynthesis (e.g., ATCC 19606). The rea-
sons for this are not fully understood, but a recent study showed potentially com-
pensatory transcriptomic changes in response to loss of lpxA [550], whereas others
showed that expression of penicillin-binding protein (PBP)-1A rendered lipid A
loss lethal in strains that could otherwise tolerate lipid A loss and that cells lacking
both PBP 1A and lipid A had increased expression of lipoproteins on their surface
that may compensate for lipid A loss [114]. Although lipid A loss and colistin resis-
tance can be readily selected in vitro, it stimulates debate about its relevance in the
clinic, both in terms of colistin resistance and, as discussed above in Sect. 4.2.1.5,
with respect to the evaluation of novel antibacterial targets within the lipid A bio-
synthetic pathway (e.g., LpxC). This stems from the notion of whether A. bauman-
nii lacking lipid A (LPS) can survive during infection and therefore could be selected
during colistin treatment.
Since the Gram-negative OM provides protection from the host immune system,
it is generally thought that loss of lipid A (OM) would render the cells unfit in the
128 C. R. Dean et al.

host environment. Supporting this, colistin-resistant isolates with mutations in lpxA,

lpxC, or lpxD [112] were highly attenuated in C. elegans and mouse models of
infection [551], and an LpxC inhibitor that lacked in vitro antibacterial activity
against A. baumannii was efficacious in a mouse model of infection [117]. Overall
this is consistent with the very high detection of pmr mutations among colistin-­
resistant clinical isolates rather than loss of LPS and implies that total loss of lipid
A may be more of an in vitro phenomenon. More recent analyses of colistin-­resistant
A. baumannii XDR clinical isolates identified mutations in pmrA, lpxC, and lpxD
(and lpsB, involved in synthesis of the core region attached to lipid A and shown to
be involved in intrinsic polymyxin resistance [552]) occurring together [524].
Isolates selected for further study (AC12 and AC30) produced considerably less
LPS than either the laboratory strain ATCC 19606 or polymyxin-susceptible clinical
isolates [553]. This suggests a possibility that mutations reducing, but not abolish-
ing, lipid A biosynthesis may emerge over time in the clinic and that the combina-
tion of pEtN and/or galactosamine modification with reduced lipid A synthesis
could conspire to decrease susceptibility. These isolates were generally drug resis-
tant, suggesting that the reduction in lipid A may not be enough to severely compro-
mise the OM permeability barrier, also allowing for their survival in the host. It is
tempting to speculate that the mutation in lpsB may serve to further stabilize the
reduced levels of modified lipid A core present in these cells and that other factors
are likely involved in determining the overall susceptibility of A. baumannii to
colistin [552, 554], including the level of lipid A acylation. A full understanding of
this mechanism awaits further study. Whether this phenomenon can extend to other
Gram-negative pathogens that strictly require LPS for growth or viability remains to
be seen, but one recent study forcing the in vitro evolution of colistin resistance in
P. aeruginosa using a morbidostat approach generated mutations in pmr genes and
lpxC among others [555].

4.2.6.3 Plasmid-Mediated Modification of LPS

Mutations mediating resistance to colistin can occur fairly rapidly in some Gram-­
negative pathogens as described above, but there were initially no reports of hori-
zontal transfer of mobile elements carrying genes mediating LPS modification and
colistin resistance. This changed in 2015 with reporting of the plasmid-borne mcr-1
gene encoding a pEtN transferase in E. coli strains in China and its distribution in
strains isolated from raw meat, animals (pigs), and humans [556]. It was quickly
established that: plasmid-borne mcr-1 was widespread in many regions of the world;
occurred in isolates from food animals, meat and vegetables, the environment, and
humans; was found mainly in E. coli but also occurred in other bacteria; and was
detected in isolate collections dating back to the 1980s [557, 558]. The first identi-
fication of the mcr-1 gene in E. coli from a patient in the United States was reported
in 2016 [559]. Additional mcr-1.2 [560], mcr-2 [561], and mcr-3 [562] variants have
now also been identified. The specific impact of the mcr-1 gene in all four Gram-­
negative ESKAPE pathogens has very recently been reported. Mcr-1 mediates pEtN
4 Resistance of Gram-negative Bacilli to Antimicrobials 129

modification in E. coli, K. pneumoniae, A. baumannii, and P. aeruginosa; however

it only seems to shift susceptibility in the first three [563]. Given that colistin was
reintroduced into clinical use primarily as a last-line therapy for treating MDR
Gram-negative infections, including carbapenem-resistant Enterobacteriaceae
(CRE), the identification of a mobile element conferring colistin resistance raised
immediate concern as to its potential dissemination into strains such as CREs, and
indeed this has already occurred. For example, the mcr-1.2 gene was originally
found in a KPC carbapenemase-producing K. pneumoniae human clinical isolate of
the clinically important ST512 lineage, isolated in Italy [560]. Two multidrug-­
resistant K. pneumoniae human clinical isolates were shown to harbor both the
blandm-5 metallo-β-lactamase and mcr-1 genes [564]. E. coli isolates from food and
human origin had both the blandm-9 and mcr-1 genes [565, 566], and an isolate from
a human urinary tract infection in the United States had blandm-5 and mcr-1 [567].
The first death known to result from such an infection in the United States occurred
in Nevada in 2016 and was attributed to an untreatable K. pneumoniae harboring an
NDM metallo-β-lactamase and plasmid-borne mcr-1 [568]. Although still generally
of lower incidence worldwide [569], the spread of these untreatable strains is inevi-
table. If colistin therapy will continue to be used, vigilance in detection and surveil-
lance of both carbapenem and colistin resistance and corresponding implementation
of effective infection control and stewardship procedures are very important [570].
Furthermore, the need for new antibiotics to treat these infections has now become
an extreme priority. Finally, the widespread distribution of mcr-1 in food animals
and food products is entirely consistent with the use of large amounts of colistin in
agriculture [557] and therefore the dissemination of these antibiotics generally into
the environment, particularly localized around farms. MCR-1 and variants are
related to a resistance protein from natural producers of polymyxin and to another
pEtN transferase, LptA from Neisseria [571, 572]. The evolutionary history remains
to be fully understood in this case, but it is difficult not to speculate that this process
was enhanced by extensive agricultural and veterinary use of colistin. Now that it
has occurred, such continued use of colistin will continue to facilitate the mainte-
nance and spread of mcr-1-containing strains, and so significant benefit may be
derived from finding creative ways to address this issue (for more information on
agricultural use of antibiotics, see Chap. 10 in this volume). It is interesting to note
that since colistin was out of favor for some time in human clinical usage, the dis-
semination of mcr-1-containing strains into reservoirs, such as the human gut, and
diversity of mcr-1-containing genetic elements may be underestimated [573].

4.2.6.4 Novel Approaches and New Agents

Current efforts in the area of novel polymyxins are aimed at the design of non-­
antibacterial polymyxin analogs for use as potentiators of currently used antibiotics
or the design of new antibacterial analogs with reduced toxicity allowing for a
higher therapeutic index or with increased antibacterial activity against emerging
polymyxin-resistant isolates. Several of these efforts exploit the earlier finding that
130 C. R. Dean et al.

Fig. 4.9 Chemical structure of polymyxin B. Acyl chain depicted in blue is a key determinant of
antibacterial activity but not outer membrane disruption activity

the N-terminal acyl chain of polymyxin B (Fig. 4.9) is involved in both antibacterial
activity and toxicity.
A derivative of polymyxin B, polymyxin B nonapetide (PMBN), lacks this moi-
ety and is less toxic and less potent as an antibacterial but retains the ability to
interact with the bacterial OM and permeabilize cells. PMBN itself has been the
subject of much interest and research over the years as a possible potentiating mol-
ecule for use in combination with other antibiotics, but the potential for unaccept-
able residual toxicity still exists. The number of positive charges on polymyxin has
also been associated with toxicity. Northern Antibiotics/Spero has exploited this to
design a PMBN derivative (SPR741) that contains an N-acetyl-threonine-d-serine
side chain, thereby reducing the number of positive charges from five to three rela-
tive to PMBN [121]. This molecule does not have significant antibacterial activity
and is reported to be less toxic but retains antibiotic potentiation activity in E. coli,
K. pneumoniae, and A. baumannii, although it does not potentiate in P. aeruginosa
[574–576]. As of this writing, SPR741 has entered Phase I clinical trials. Cubist
Pharmaceuticals (now Merck) have designed a polymyxin decapeptide derivative
containing a halo-aryl moiety at its N-terminus (CB-182804) to pursue a reduction
in toxicity [577]. CB-182804 exhibited slightly lower antibacterial potency relative
to polymyxin B but was efficacious in animal models of infection and is reported to
have reduced toxicity. CB-182804 entered Phase I clinical trials but appears to be
discontinued. Along the same lines, Pfizer reported a series of analogs replacing the
N-terminal acyl chain with biaryl moieties and substituting the diamino-butyrate
moiety at amino acid 3 with diamino-propionate. One of these, named 5 X, had
slightly improved antibacterial activity and indications of reduced toxicity, but
based on studies in dogs, the therapeutic index was not significantly better than
polymyxin B [578]. Researchers at Monash University are exploring novel poly-
myxin lipopeptides to define structure activity relationships for gaining activity
against colistin-resistant isolates [579] and have presented data on other less toxic
polymyxin derivatives in conjunction with the Medicines Company [580]. Cantab
4 Resistance of Gram-negative Bacilli to Antimicrobials 131

has also reported on the piperazine derivative that showed reduced cytotoxicity and
improved in vivo efficacy over polymyxin B in A. baumannii and P. aeruginosa
lung infection models. Additional details and chemical structures for these and
other novel polymyxins can be found in the review by Brown and Dawson [581].
Finally, there is renewed interest in the octapeptin natural products which also inter-
act with and traverse the Gram-negative OM but do so via a different mechanism
and therefore may be active against polymyxin-resistant strains [582–584].

4.3 Concluding Remarks

The discovery of antibiotics, along with vaccines and improved concepts in hygiene,
could be considered the greatest achievement in healthcare-related science in his-
tory. Unfortunately, decades of antibiotic use and perhaps misuse, in both medicine
and agriculture, have enriched for resistant bacteria in the clinical setting, eroding
the effectiveness of the antibiotics upon which we have relied and setting the stage
for a potentially very different reality in medicine from what most of us had grown
accustomed to. This is especially unfortunate since so much of medical practice, for
example, surgery, has relied on antibiotics for success. As can be seen from the
above discussions, antibacterial resistance is complex and multifactorial. However
there are key mechanisms that affect susceptibility to certain classes, such as
β-lactamases for β-lactams and AMEs for aminoglycosides, which may provide
specific strategies for next-generation versions of these antibiotics that address
those mechanisms. It is clear that no effort should be spared on these approaches for
the near term and that new agents directed at previously unexploited novel targets
should be aggressively pursued where they show promise. Hopefully the new aware-
ness of the issue of antimicrobial resistance, and the various incentivizing efforts
spawned from this, will be successful in moving us in the right direction to address
this threat. Finally, even if new agents come along in the near term, it is imperative
that complacency in antibiotic discovery never again sets in. New agents will likely
be the last line of defense, and as such, when resistance to them emerges, the overall
issue of untreatable infections would again be upon us.

Major Points
• Gram-negative pathogens have a unique additional asymmetric outer membrane
(OM). This membrane establishes a significant permeability barrier (reduces
influx) to toxic molecules including antibiotics. Mutations decreasing compound
permeability can be selected under antibiotic exposure.
• Gram-negative pathogens have unique RND family efflux pumps that extrude
most antibiotics and other toxic molecules; these work together with the OM to
reduce intracellular compound accumulation. Upregulation of efflux pump
expression, or changes in compound specificity, can be selected under antibiotic
exposure.
132 C. R. Dean et al.

• An understanding of the design of new compounds that accumulate sufficiently

(overcome efflux) in Gram-negative bacteria is lacking.
• Mechanisms that cause resistance to specific antibiotics include mutations that
alter the antibiotic target, acquisition of proteins that bind and protect the target,
or acquisition of enzymes that modify antibiotics.
• Clinical resistance to specific classes of antibiotics often results from combina-
tions of efflux and OM changes together with compound-specific mechanisms.
• Multidrug resistance arises from various combinations of all of the above.

References

1. Boucher HW, Talbot GH, Bradley JS, Edwards JE, Gilbert D, Rice LB, et al. Bad bugs, no
drugs: no ESKAPE! An update from the Infectious Diseases Society of America. Clin Infect
Dis. 2009;48(1):1–12.
2. Turnidge J, Paterson DL. Setting and revising antibacterial susceptibility breakpoints. Clin
Microbiol Rev. 2007;20(3):391–408.
3. Li XZ, Plesiat P, Nikaido H. The challenge of efflux-mediated antibiotic resistance in Gram-­
negative bacteria. Clin Microbiol Rev. 2015;28(2):337–418.
4. Yu EW, Aires JR, Nikaido H. AcrB multidrug efflux pump of Escherichia coli: compos-
ite substrate-binding cavity of exceptional flexibility generates its extremely wide substrate
specificity. J Bacteriol. 2003;185(19):5657–64.
5. Yu EW, McDermott G, Zgurskaya HI, Nikaido H, Koshland DE Jr. Structural basis of multiple
drug-binding capacity of the AcrB multidrug efflux pump. Science. 2003;300(5621):976–80.
6. Tal N, Schuldiner S. A coordinated network of transporters with overlapping specificities
provides a robust survival strategy. Proc Natl Acad Sci U S A. 2009;106(22):9051–6.
7. Kumar A, Schweizer HP. Bacterial resistance to antibiotics: active efflux and reduced uptake.
Adv Drug Deliv Rev. 2005;57(10):1486–513.
8. Winsor GL, Griffiths EJ, Lo R, Dhillon BK, Shay JA, Brinkman FS. Enhanced annotations
and features for comparing thousands of Pseudomonas genomes in the Pseudomonas genome
database. Nucleic Acids Res. 2016;44(D1):D646–53.
9. Du D, Wang Z, James NR, Voss JE, Klimont E, Ohene-Agyei T, et al. Structure of the AcrAB-­
TolC multidrug efflux pump. Nature. 2014;509(7501):512–5.
10. Tamura N, Murakami S, Oyama Y, Ishiguro M, Yamaguchi A. Direct interaction of multidrug
efflux transporter AcrB and outer membrane channel TolC detected via site-directed disulfide
cross-linking. Biochemistry. 2005;44(33):11115–21.
11. Kim JS, Jeong H, Song S, Kim HY, Lee K, Hyun J, et al. Structure of the tripartite mul-
tidrug efflux pump AcrAB-TolC suggests an alternative assembly mode. Mol Cells.
2015;38(2):180–6.
12. Daury L, Orange F, Taveau JC, Verchere A, Monlezun L, Gounou C, et al. Tripartite assembly
of RND multidrug efflux pumps. Nat Commun. 2016;7:10731.
13. Murakami S, Nakashima R, Yamashita E, Matsumoto T, Yamaguchi A. Crystal struc-
tures of a multidrug transporter reveal a functionally rotating mechanism. Nature.
2006;443(7108):173–9.
14. Seeger MA, Schiefner A, Eicher T, Verrey F, Diederichs K, Pos KM. Structural asymmetry of
AcrB trimer suggests a peristaltic pump mechanism. Science. 2006;313(5791):1295–8.
15. Sennhauser G, Amstutz P, Briand C, Storchenegger O, Grutter MG. Drug export pathway of
multidrug exporter AcrB revealed by DARPin inhibitors. PLoS Biol. 2007;5(1):e7.
16. Ruggerone P, Vargiu AV, Collu F, Fischer N, Kandt C. Molecular dynamics computer simula-
tions of multidrug RND efflux pumps. Comput Struct Biotechnol J. 2013;5:e201302008.
4 Resistance of Gram-negative Bacilli to Antimicrobials 133

17. Schulz R, Vargiu AV, Collu F, Kleinekathofer U, Ruggerone P. Functional rotation of the
transporter AcrB: insights into drug extrusion from simulations. PLoS Comput Biol.
2010;6(6):e1000806.
18. Li XZ, Ma D, Livermore DM, Nikaido H. Role of efflux pump(s) in intrinsic resistance
of Pseudomonas aeruginosa: active efflux as a contributing factor to β-lactam resistance.
Antimicrob Agents Chemother. 1994;38(8):1742–52.
19. Nikaido H, Basina M, Nguyen V, Rosenberg EY. Multidrug efflux pump AcrAB of
Salmonella typhimurium excretes only those β-lactam antibiotics containing lipophilic side
chains. J Bacteriol. 1998;180(17):4686–92.
20. Li XZ, Zhang L, Srikumar R, Poole K. Β-lactamase inhibitors are substrates for the
multidrug efflux pumps of Pseudomonas aeruginosa. Antimicrob Agents Chemother.
1998;42(2):399–403.
21. Kobayashi N, Tamura N, van Veen HW, Yamaguchi A, Murakami S. β-Lactam selectivity of
multidrug transporters AcrB and AcrD resides in the proximal binding pocket. J Biol Chem.
2014;289(15):10680–90.
22. Elkins CA, Nikaido H. Substrate specificity of the RND-type multidrug efflux pumps AcrB
and AcrD of Escherichia coli is determined predominantly by two large periplasmic loops.
J Bacteriol. 2002;184(23):6490–8.
23. Eda S, Maseda H, Nakae T. An elegant means of self-protection in gram-negative bacte-
ria by recognizing and extruding xenobiotics from the periplasmic space. J Biol Chem.
2003;278(4):2085–8.
24. Lee A, Mao W, Warren MS, Mistry A, Hoshino K, Okumura R, et al. Interplay between efflux
pumps may provide either additive or multiplicative effects on drug resistance. J Bacteriol.
2000;182(11):3142–50.
25. Raetz CR, Whitfield C. Lipopolysaccharide endotoxins. Annu Rev Biochem.
2002;71:635–700.
26. Walsh AG, Matewish MJ, Burrows LL, Monteiro MA, Perry MB, Lam JS. Lipopolysaccharide
core phosphates are required for viability and intrinsic drug resistance in Pseudomonas
aeruginosa. Mol Microbiol. 2000;35(4):718–27.
27. Pages JM, James CE, Winterhalter M. The porin and the permeating antibiotic: a selective
diffusion barrier in Gram-negative bacteria. Nat Rev Microbiol. 2008;6(12):893–903.
28. Nikaido H. Porins and specific diffusion channels in bacterial outer membranes. J Biol Chem.
1994;269(6):3905–8.
29. Sugawara E, Nagano K, Nikaido H. Alternative folding pathways of the major porin OprF of
Pseudomonas aeruginosa. FEBS J. 2012;279(6):910–8.
30. Hancock RE, Woodruff WA. Roles of porin and β-lactamase in β-lactam resistance of
Pseudomonas aeruginosa. Rev Infect Dis. 1988;10(4):770–5.
31. Isabella VM, Campbell AJ, Manchester J, Sylvester M, Nayar AS, Ferguson KE, et al.
Toward the rational design of carbapenem uptake in Pseudomonas aeruginosa. Chem Biol.
2015;22(4):535–47.
32. Faraldo-Gomez JD, Sansom MS. Acquisition of siderophores in gram-negative bacteria. Nat
Rev Mol Cell Biol. 2003;4(2):105–16.
33. Kadner RJ. Vitamin B12 transport in Escherichia coli: energy coupling between membranes.
Mol Microbiol. 1990;4(12):2027–33.
34. Nikaido H. The role of outer membrane and efflux pumps in the resistance of gram-negative
bacteria. Can we improve drug access? Drug Resist Updat. 1998;1(2):93–8.
35. Nikaido H. Molecular basis of bacterial outer membrane permeability revisited. Microbiol
Mol Biol Rev. 2003;67(4):593–656.
36. Lim SP, Nikaido H. Kinetic parameters of efflux of penicillins by the multidrug efflux trans-
porter AcrAB-TolC of Escherichia coli. Antimicrob Agents Chemother. 2010;54(5):1800–6.
37. Mazzariol A, Cornaglia G, Nikaido H. Contributions of the AmpC β-lactamase and the
AcrAB multidrug efflux system in intrinsic resistance of Escherichia coli K-12 to β-lactams.
Antimicrob Agents Chemother. 2000;44(5):1387–90.
134 C. R. Dean et al.

38. Li XZ, Zhang L, Poole K. Interplay between the MexA-MexB-OprM multidrug efflux system
and the outer membrane barrier in the multiple antibiotic resistance of Pseudomonas aerugi-
nosa. J Antimicrob Chemother. 2000;45(4):433–6.
39. Krishnamoorthy G, Wolloscheck D, Weeks JW, Croft C, Rybenkov VV, Zgurskaya
HI. Breaking the permeability barrier of Escherichia coli by controlled hyperporination of
the outer membrane. Antimicrob Agents Chemother. 2016;60(12):7372–81.
40. Dean CR, Visalli MA, Projan SJ, Sum PE, Bradford PA. Efflux-mediated resistance to tige-
cycline (GAR-936) in Pseudomonas aeruginosa PAO1. Antimicrob Agents Chemother.
2003;47(3):972–8.
41. Visalli MA, Murphy E, Projan SJ, Bradford PA. AcrAB multidrug efflux pump is associ-
ated with reduced levels of susceptibility to tigecycline (GAR-936) in Proteus mirabilis.
Antimicrob Agents Chemother. 2003;47(2):665–9.
42. Ruzin A, Visalli MA, Keeney D, Bradford PA. Influence of transcriptional activator RamA
on expression of multidrug efflux pump AcrAB and tigecycline susceptibility in Klebsiella
pneumoniae. Antimicrob Agents Chemother. 2005;49(3):1017–22.
43. Keeney D, Ruzin A, McAleese F, Murphy E, Bradford PA. MarA-mediated overexpression of
the AcrAB efflux pump results in decreased susceptibility to tigecycline in Escherichia coli.
J Antimicrob Chemother. 2008;61(1):46–53.
44. Deng M, Zhu MH, Li JJ, Bi S, Sheng ZK, Hu FS, et al. Molecular epidemiology and mecha-
nisms of tigecycline resistance in clinical isolates of Acinetobacter baumannii from a Chinese
university hospital. Antimicrob Agents Chemother. 2014;58(1):297–303.
45. Li X, Zolli-Juran M, Cechetto JD, Daigle DM, Wright GD, Brown ED. Multicopy suppres-
sors for novel antibacterial compounds reveal targets and drug efflux susceptibility. Chem
Biol. 2004;11(10):1423–30.
46. Brown DG, May-Dracka TL, Gagnon MM, Tommasi R. Trends and exceptions of physical
properties on antibacterial activity for Gram-positive and Gram-negative pathogens. J Med
Chem. 2014;57(23):10144–61.
47. Dean CR, Narayan S, Daigle DM, Dzink-Fox JL, Puyang X, Bracken KR, et al. Role of
the AcrAB-TolC efflux pump in determining susceptibility of Haemophilus influen-
zae to the novel peptide deformylase inhibitor LBM415. Antimicrob Agents Chemother.
2005;49(8):3129–35.
48. Caughlan RE, Jones AK, Delucia AM, Woods AL, Xie L, Ma B, et al. Mechanisms decreas-
ing in vitro susceptibility to the LpxC inhibitor CHIR-090 in the gram-negative pathogen
Pseudomonas aeruginosa. Antimicrob Agents Chemother. 2012;56(1):17–27.
49. Jones AK, Woods AL, Takeoka KT, Shen X, Wei JR, Caughlan RE, et al. Determinants of
antibacterial spectrum and resistance potential of the elongation factor G inhibitor argyrin B
in key gram-negative pathogens. Antimicrob Agents Chemother. 2017;61(4):e02400–16.
50. Caughlan RE, Sriram S, Daigle DM, Woods AL, Buco J, Peterson RL, et al. Fmt bypass in
Pseudomonas aeruginosa causes induction of MexXY efflux pump expression. Antimicrob
Agents Chemother. 2009;53(12):5015–21.
51. Balibar CJ, Iwanowicz D, Dean CR. Elongation factor P is dispensable in Escherichia coli
and Pseudomonas aeruginosa. Curr Microbiol. 2013;67(3):293–9.
52. Morita Y, Gilmour C, Metcalf D, Poole K. Translational control of the antibiotic inducibility
of the PA5471 gene required for mexXY multidrug efflux gene expression in Pseudomonas
aeruginosa. J Bacteriol. 2009;191(15):4966–75.
53. Morita Y, Sobel ML, Poole K. Antibiotic inducibility of the MexXY multidrug efflux system
of Pseudomonas aeruginosa: involvement of the antibiotic-inducible PA5471 gene product.
J Bacteriol. 2006;188(5):1847–55.
54. Chen W, Wang D, Zhou W, Sang H, Liu X, Ge Z, et al. Novobiocin binding to NalD induces
the expression of the MexAB-OprM pump in Pseudomonas aeruginosa. Mol Microbiol.
2016;100(5):749–58.
55. Srikumar R, Paul CJ, Poole K. Influence of mutations in the mexR repressor gene on expres-
sion of the MexA-MexB-oprM multidrug efflux system of Pseudomonas aeruginosa.
J Bacteriol. 2000;182(5):1410–4.
4 Resistance of Gram-negative Bacilli to Antimicrobials 135

56. Daigle DM, Cao L, Fraud S, Wilke MS, Pacey A, Klinoski R, et al. Protein modulator of multi-
drug efflux gene expression in Pseudomonas aeruginosa. J Bacteriol. 2007;189(15):5441–51.
57. Morita Y, Cao L, Gould VC, Avison MB, Poole K. nalD encodes a second repressor of
the mexAB-oprM multidrug efflux operon of Pseudomonas aeruginosa. J Bacteriol.
2006;188(24):8649–54.
58. Fetar H, Gilmour C, Klinoski R, Daigle DM, Dean CR, Poole K. mexEF-oprN multidrug
efflux operon of Pseudomonas aeruginosa: regulation by the MexT activator in response to
nitrosative stress and chloramphenicol. Antimicrob Agents Chemother. 2011;55(2):508–14.
59. Sobel ML, Neshat S, Poole K. Mutations in PA2491 (mexS) promote MexT-dependent
mexEF-oprN expression and multidrug resistance in a clinical strain of Pseudomonas aeru-
ginosa. J Bacteriol. 2005;187(4):1246–53.
60. Baranova N, Nikaido H. The baeSR two-component regulatory system activates transcription
of the yegMNOB (mdtABCD) transporter gene cluster in Escherichia coli and increases its
resistance to novobiocin and deoxycholate. J Bacteriol. 2002;184(15):4168–76.
61. Lin MF, Lin YY, Yeh HW, Lan CY. Role of the BaeSR two-component system in the regula-
tion of Acinetobacter baumannii adeAB genes and its correlation with tigecycline suscepti-
bility. BMC Microbiol. 2014;14:119.
62. Rosner JL, Martin RG. Reduction of cellular stress by TolC-dependent efflux pumps in
Escherichia coli indicated by BaeSR and CpxARP activation of spy in efflux mutants.
J Bacteriol. 2013;195(5):1042–50.
63. Tian ZX, Yi XX, Cho A, O”Gara F, Wang YP. CpxR activates MexAB-OprM efflux pump
expression and enhances antibiotic resistance in both laboratory and clinical nalB-type iso-
lates of Pseudomonas aeruginosa. PLoS Pathog. 2016;12(10):e1005932.
64. Adewoye L, Sutherland A, Srikumar R, Poole K. The mexR repressor of the mexAB-oprM
multidrug efflux operon in Pseudomonas aeruginosa: characterization of mutations compro-
mising activity. J Bacteriol. 2002;184(15):4308–12.
65. Sobel ML, Hocquet D, Cao L, Plesiat P, Poole K. Mutations in PA3574 (nalD) lead to
increased MexAB-OprM expression and multidrug resistance in laboratory and clinical iso-
lates of Pseudomonas aeruginosa. Antimicrob Agents Chemother. 2005;49(5):1782–6.
66. Cao L, Srikumar R, Poole K. MexAB-OprM hyperexpression in NalC-type multidrug-­
resistant Pseudomonas aeruginosa: identification and characterization of the nalC gene
encoding a repressor of PA3720-PA3719. Mol Microbiol. 2004;53(5):1423–36.
67. Wang H, Dzink-Fox JL, Chen M, Levy SB. Genetic characterization of highly fluoroquinolone-­
resistant clinical Escherichia coli strains from China: role of acrR mutations. Antimicrob
Agents Chemother. 2001;45(5):1515–21.
68. Purssell A, Poole K. Functional characterization of the NfxB repressor of the mexCD-­
oprJ multidrug efflux operon of Pseudomonas aeruginosa. Microbiology. 2013;159(Pt
10):2058–73.
69. Poole K. Efflux-mediated antimicrobial resistance. J Antimicrob Chemother.
2005;56(1):20–51.
70. Blair JM, Webber MA, Baylay AJ, Ogbolu DO, Piddock LJ. Molecular mechanisms of anti-
biotic resistance. Nat Rev Microbiol. 2015;13(1):42–51.
71. Poulou A, Voulgari E, Vrioni G, Koumaki V, Xidopoulos G, Chatzipantazi V, et al. Outbreak
caused by an ertapenem-resistant, CTX-M-15-producing Klebsiella pneumoniae sequence
type 101 clone carrying an OmpK36 porin variant. J Clin Microbiol. 2013;51(10):3176–82.
72. Wozniak A, Villagra NA, Undabarrena A, Gallardo N, Keller N, Moraga M, et al. Porin alter-
ations present in non-carbapenemase-producing Enterobacteriaceae with high and intermedi-
ate levels of carbapenem resistance in Chile. J Med Microbiol. 2012;61(Pt 9):1270–9.
73. Lavigne JP, Sotto A, Nicolas-Chanoine MH, Bouziges N, Pages JM, Davin-Regli A. An adap-
tive response of Enterobacter aerogenes to imipenem: regulation of porin balance in clinical
isolates. Int J Antimicrob Agents. 2013;41(2):130–6.
74. Bajaj H, Scorciapino MA, Moynie L, Page MG, Naismith JH, Ceccarelli M, et al. Molecular
basis of filtering carbapenems by porins from β-Lactam-resistant clinical strains of
Escherichia coli. J Biol Chem. 2016;291(6):2837–47.
136 C. R. Dean et al.

75. Muller C, Plesiat P, Jeannot K. A two-component regulatory system interconnects resistance

to polymyxins, aminoglycosides, fluoroquinolones, and β-lactams in Pseudomonas aerugi-
nosa. Antimicrob Agents Chemother. 2011;55(3):1211–21.
76. Lomovskaya O, Warren MS, Lee A, Galazzo J, Fronko R, Lee M, et al. Identification and char-
acterization of inhibitors of multidrug resistance efflux pumps in Pseudomonas aeruginosa:
novel agents for combination therapy. Antimicrob Agents Chemother. 2001;45(1):105–16.
77. Venter H, Mowla R, Ohene-Agyei T, Ma S. RND-type drug e ffl ux pumps from Gram-­
negative bacteria: molecular mechanism and inhibition. Front Microbiol. 2015;6:377.
78. Peters JU, Hert J, Bissantz C, Hillebrecht A, Gerebtzoff G, Bendels S, et al. Can we dis-
cover pharmacological promiscuity early in the drug discovery process? Drug Discov Today.
2012;17(7-8):325–35.
79. Epand RM, Epand RF. Lipid domains in bacterial membranes and the action of antimicrobial
agents. Biochim Biophys Acta. 2009;1788(1):289–94.
80. Yoshida K, Nakayama K, Ohtsuka M, Kuru N, Yokomizo Y, Sakamoto A, et al. MexAB-­
OprM specific efflux pump inhibitors in Pseudomonas aeruginosa. Part 7: highly soluble and
in vivo active quaternary ammonium analogue D13-9001, a potential preclinical candidate.
Bioorg Med Chem. 2007;15(22):7087–97.
81. Nakashima R, Sakurai K, Yamasaki S, Hayashi K, Nagata C, Hoshino K, et al. Structural
basis for the inhibition of bacterial multidrug exporters. Nature. 2013;500(7460):102–6.
82. Zuo Z, Weng J, Wang W. Insights into the inhibitory mechanism of D13-9001 to the
multidrug transporter AcrB through molecular dynamics simulations. J Phys Chem B.
2016;120(9):2145–54.
83. Nguyen ST, Kwasny SM, Ding X, Cardinale SC, McCarthy CT, Kim HS, et al. Structure-­
activity relationships of a novel pyranopyridine series of Gram-negative bacterial efflux
pump inhibitors. Bioorg Med Chem. 2015;23(9):2024–34.
84. Opperman TJ, Kwasny SM, Kim HS, Nguyen ST, Houseweart C, D’Souza S, et al.
Characterization of a novel pyranopyridine inhibitor of the AcrAB efflux pump of Escherichia
coli. Antimicrob Agents Chemother. 2014;58(2):722–33.
85. Sjuts H, Vargiu AV, Kwasny SM, Nguyen ST, Kim HS, Ding X, et al. Molecular basis for
inhibition of AcrB multidrug efflux pump by novel and powerful pyranopyridine derivatives.
Proc Natl Acad Sci U S A. 2016;113(13):3509–14.
86. Abdali N, Parks JM, Haynes KM, Chaney JL, Green AT, Wolloscheck D, et al. Reviving
antibiotics: efflux pump inhibitors that interact with AcrA, a membrane fusion protein of the
AcrAB-TolC multidrug efflux pump. ACS Infect Dis. 2017;3(1):89–98.
87. Chalhoub H, Pletzer D, Weingart H, Braun Y, Tunney MM, Elborn JS, et al. Mechanisms of
intrinsic resistance and acquired susceptibility of Pseudomonas aeruginosa isolated from
cystic fibrosis patients to temocillin, a revived antibiotic. Sci Rep. 2017;7:40208.
88. Vuorio R, Vaara M. The lipid A biosynthesis mutation lpxA2 of Escherichia coli results in
drastic antibiotic supersusceptibility. Antimicrob Agents Chemother. 1992;36(4):826–9.
89. Young K, Silver LL, Bramhill D, Cameron P, Eveland SS, Raetz CR, et al. The envA perme-
ability/cell division gene of Escherichia coli encodes the second enzyme of lipid A biosyn-
thesis. UDP-3-O-(R-3-hydroxymyristoyl)-N-acetylglucosamine deacetylase. J Biol Chem.
1995;270(51):30384–91.
90. Sampson BA, Misra R, Benson SA. Identification and characterization of a new
gene of Escherichia coli K-12 involved in outer membrane permeability. Genetics.
1989;122(3):491–501.
91. Chng SS, Ruiz N, Chimalakonda G, Silhavy TJ, Kahne D. Characterization of the two-­
protein complex in Escherichia coli responsible for lipopolysaccharide assembly at the outer
membrane. Proc Natl Acad Sci U S A. 2010;107(12):5363–8.
92. Wu T, McCandlish AC, Gronenberg LS, Chng SS, Silhavy TJ, Kahne D. Identification of a
protein complex that assembles lipopolysaccharide in the outer membrane of Escherichia
coli. Proc Natl Acad Sci U S A. 2006;103(31):11754–9.
4 Resistance of Gram-negative Bacilli to Antimicrobials 137

93. Balibar CJ, Grabowicz M. Mutant alleles of lptD increase the permeability of Pseudomonas
aeruginosa and define determinants of intrinsic resistance to antibiotics. Antimicrob Agents
Chemother. 2016;60(2):845–54.
94. Shen X, Johnson AK, Jones AK, Barnes SW, Walker JR, Ranjitkar S, Woods AL, Six DA,
Dean CR. Genetic characterization of the hypersusceptible Pseudomonas aeruginosa strain
Z61: identification of a defect in LptE. Abstr C-105 Abstr 54th Intersci Conf Antimicrob
Agents Chemother American Society for Microbiology, Washington, DC; 2014.
95. Piizzi G, Parker DT, Peng Y, Dobler M, Patnaik A, Wattanasin S, et al. Design, synthesis, and
properties of a potent inhibitor of Pseudomonas aeruginosa deacetylase LpxC. J Med Chem.
2017;60(12):5002–14.
96. Barb AW, Jiang L, Raetz CR, Zhou P. Structure of the deacetylase LpxC bound to the antibi-
otic CHIR-090: time-dependent inhibition and specificity in ligand binding. Proc Natl Acad
Sci U S A. 2007;104(47):18433–8.
97. Barb AW, McClerren AL, Snehelatha K, Reynolds CM, Zhou P, Raetz CR. Inhibition of lipid
A biosynthesis as the primary mechanism of CHIR-090 antibiotic activity in Escherichia
coli. Biochemist. 2007;46(12):3793–802.
98. Brown MF, Reilly U, Abramite JA, Arcari JT, Oliver R, Barham RA, et al. Potent inhibitors
of LpxC for the treatment of Gram-negative infections. J Med Chem. 2012;55(2):914–23.
99. Hale MR, Hill P, Lahiri S, Miller MD, Ross P, Alm R, et al. Exploring the UDP pocket of
LpxC through amino acid analogs. Bioorg Med Chem Lett. 2013;23(8):2362–7.
100. Liang X, Lee CJ, Chen X, Chung HS, Zeng D, Raetz CR, et al. Syntheses, structures and
antibiotic activities of LpxC inhibitors based on the diacetylene scaffold. Bioorg Med Chem.
2011;19(2):852–60.
101. Liang X, Lee CJ, Zhao J, Toone EJ, Zhou P. Synthesis, structure, and antibiotic activity of
aryl-substituted LpxC inhibitors. J Med Chem. 2013;56(17):6954–66.
102. Mansoor UF, Vitharana D, Reddy PA, Daubaras DL, McNicholas P, Orth P, et al. Design
and synthesis of potent Gram-negative specific LpxC inhibitors. Bioorg Med Chem Lett.
2011;21(4):1155–61.
103. McAllister LA, Montgomery JI, Abramite JA, Reilly U, Brown MF, Chen JM, et al.
Heterocyclic methylsulfone hydroxamic acid LpxC inhibitors as Gram-negative antibacterial
agents. Bioorg Med Chem Lett. 2012;22(22):6832–8.
104. Montgomery JI, Brown MF, Reilly U, Price LM, Abramite JA, Arcari J, et al. Pyridone meth-
ylsulfone hydroxamate LpxC inhibitors for the treatment of serious gram-negative infections.
J Med Chem. 2012;55(4):1662–70.
105. Murphy-Benenato KE, Olivier N, Choy A, Ross PL, Miller MD, Thresher J, et al. Synthesis,
structure, and SAR of tetrahydropyran-Based LpxC Inhibitors. ACS Med Chem Lett.
2014;5(11):1213–8.
106. Titecat M, Liang X, Lee CJ, Charlet A, Hocquet D, Lambert T, et al. High susceptibility
of MDR and XDR Gram-negative pathogens to biphenyl-diacetylene-based difluoromethyl-
allo-­threonyl-hydroxamate LpxC inhibitors. J Antimicrob Chemother. 2016;71(10):2874–82.
107. Kasar R, Linsell MS, Aggen JB, Lu Q, Wang D, Church T, et al. Hydroxamic acid derivatives
and their use in the treatment of bacterial infections. Google Patents. 2012.
108. Cigana C, Bernardini F, Facchini M, Alcala-Franco B, Riva C, De Fino I, et al. Efficacy of
the novel antibiotic POL7001 in preclinical models of Pseudomonas aeruginosa pneumonia.
Antimicrob Agents Chemother. 2016;60(8):4991–5000.
109. Kalinin DV, Holl R. Insights into the zinc-dependent deacetylase LpxC: biochemical proper-
ties and inhibitor design. Curr Top Med Chem. 2016;16(21):2379–430.
110. Srinivas N, Jetter P, Ueberbacher BJ, Werneburg M, Zerbe K, Steinmann J, et al.
Peptidomimetic antibiotics target outer-membrane biogenesis in Pseudomonas aeruginosa.
Science. 2010;327(5968):1010–3.
111. Steeghs L, de Cock H, Evers E, Zomer B, Tommassen J, van der Ley P. Outer membrane
composition of a lipopolysaccharide-deficient Neisseria meningitidis mutant. EMBO
J. 2001;20(24):6937–45.
138 C. R. Dean et al.

112. Moffatt JH, Harper M, Harrison P, Hale JD, Vinogradov E, Seemann T, et al. Colistin resis-
tance in Acinetobacter baumannii is mediated by complete loss of lipopolysaccharide pro-
duction. Antimicrob Agents Chemother. 2010;54(12):4971–7.
113. Garcia-Quintanilla M, Carretero-Ledesma M, Moreno-Martinez P, Martin-Pena R, Pachon
J, McConnell MJ. Lipopolysaccharide loss produces partial colistin dependence and collat-
eral sensitivity to azithromycin, rifampicin and vancomycin in Acinetobacter baumannii. Int
J Antimicrob Agents. 2015;46(6):696–702.
114. Boll JM, Crofts AA, Peters K, Cattoir V, Vollmer W, Davies BW, et al. A penicillin-binding
protein inhibits selection of colistin-resistant, lipooligosaccharide-deficient Acinetobacter
baumannii. Proc Natl Acad Sci U S A. 2016;113(41):E6228–E37.
115. Bojkovic J, Richie DL, Six DA, Rath CM, Sawyer WS, Hu Q, et al. Characterization of an
Acinetobacter baumannii lptD deletion strain: permeability defects and response to inhibition
of lipopolysaccharide and fatty acid biosynthesis. J Bacteriol. 2016;198(4):731–41.
116. Garcia-Quintanilla M, Caro-Vega JM, Pulido MR, Moreno-Martinez P, Pachon J, McConnell
MJ. Inhibition of LpxC Increases antibiotic susceptibility in Acinetobacter baumannii.
Antimicrob Agents Chemother. 2016;60(8):5076–9.
117. Lin L, Tan B, Pantapalangkoor P, Ho T, Baquir B, Tomaras A, et al. Inhibition of LpxC pro-
tects mice from resistant Acinetobacter baumannii by modulating inflammation and enhanc-
ing phagocytosis. MBio. 2012;3(5):e00312.
118. Clements JM, Coignard F, Johnson I, Chandler S, Palan S, Waller A, et al. Antibacterial
activities and characterization of novel inhibitors of LpxC. Antimicrob Agents Chemother.
2002;46(6):1793–9.
119. Kimura Y, Matsunaga H, Vaara M. Polymyxin B octapeptide and polymyxin B hepta-
peptide are potent outer membrane permeability-increasing agents. J Antibiot (Tokyo).
1992;45(5):742–9.
120. Vaara M. Agents that increase the permeability of the outer membrane. Microbiol Rev.
1992;56(3):395–411.
121. Corbett D, Wise A, Langley T, Skinner K, Trimby E, Birchall S, et al. Potentiation of anti-
biotic activity by a novel cationic peptide: potency and spectrum of activity of SPR741.
Antimicrob Agents Chemother. 2017;61(8):e00200–17.
122. Murray B, Pillar C, Pucci M, Shinabarger D. Mechanism of action of SPR741, a potentiator
molecule for Gram-negative pathogens. Abstr Saturday-491. New Orleans: Microbe; 2017.
123. Urfer M, Bogdanovic J, Lo Monte F, Moehle K, Zerbe K, Omasits U, et al. A peptidomimetic
antibiotic targets outer membrane proteins and disrupts selectively the outer membrane in
Escherichia coli. J Biol Chem. 2016;291(4):1921–32.
124. DiGiandomenico A, Keller AE, Gao C, Rainey GJ, Warrener P, Camara MM, et al. A mul-
tifunctional bispecific antibody protects against Pseudomonas aeruginosa. Sci Transl Med.
2014;6(262):262ra155.
125. Thanabalasuriar A, Surewaard BG, Willson ME, Neupane AS, Stover CK, Warrener P, et al.
Bispecific antibody targets multiple Pseudomonas aeruginosa evasion mechanisms in the
lung vasculature. J Clin Invest. 2017;127(6):2249–61.
126. Tran QT, Williams S, Farid R, Erdemli G, Pearlstein R. The translocation kinetics of antibiot-
ics through porin OmpC: insights from structure-based solvation mapping using WaterMap.
Proteins. 2013;81(2):291–9.
127. Tran QT, Pearlstein RA, Williams S, Reilly J, Krucker T, Erdemli G. Structure-kinetic rela-
tionship of carbapenem antibacterials permeating through E. coli OmpC porin. Proteins.
2014;82(11):2998–3012.
128. Bajaj H, Acosta Gutierrez S, Bodrenko I, Malloci G, Scorciapino MA, Winterhalter M, et al.
Bacterial outer membrane porins as electrostatic nanosieves: exploring transport rules of
small polar molecules. ACS Nano. 2017;11(6):4598–606.
129. Ghai I, Winterhalter M, Wagner R. Probing transport of charged β-lactamase inhibi-
tors through OmpC, a membrane channel from E. coli. Biochem Biophys Res Commun.
2017;484(1):51–5.
4 Resistance of Gram-negative Bacilli to Antimicrobials 139

130. Giammanco A, Cala C, Fasciana T, Dowzicky MJ. Global assessment of the activity of tige-
cycline against multidrug-resistant Gram-negative pathogens between 2004 and 2014 as part
of the tigecycline evaluation and surveillance trial. mSphere. 2017;2(1):e00310.
131. Lomovskaya O, Watkins WJ. Efflux pumps: their role in antibacterial drug discovery. Curr
Med Chem. 2001;8(14):1699–711.
132. O’Shea R, Moser HE. Physicochemical properties of antibacterial compounds: implications
for drug discovery. J Med Chem. 2008;51(10):2871–8.
133. Reck F, Ehmann DE, Dougherty TJ, Newman JV, Hopkins S, Stone G, et al. Optimization
of physicochemical properties and safety profile of novel bacterial topoisomerase type II
inhibitors (NBTIs) with activity against Pseudomonas aeruginosa. Bioorg Med Chem.
2014;22(19):5392–409.
134. Richter MF, Drown BS, Riley AP, Garcia A, Shirai T, Svec RL, et al. Predictive compound
accumulation rules yield a broad-spectrum antibiotic. Nature. 2017;545(7654):299–304.
135. Hong S, Moritz TJ, Rath CM, Tamrakar P, Lee P, Krucker T, et al. Assessing antibiotic perme-
ability of Gram-negative bacteria via nanofluidics. ACS Nano. 2017;11(7):6959–67.
136. Tian H, Six DA, Krucker T, Leeds JA, Winograd N. Subcellular chemical imaging of
antibiotics in single bacteria using C60-secondary ion mass spectrometry. Anal Chem.
2017;89(9):5050–7.
137. Ito A, Kohira N, Bouchillon SK, West J, Rittenhouse S, Sader HS, et al. In vitro antimicrobial
activity of S-649266, a catechol-substituted siderophore cephalosporin, when tested against
non-fermenting Gram-negative bacteria. J Antimicrob Chemother. 2016;71(3):670–7.
138. Fleming A. On the antibacterial action of cultures of a penicillium, with special reference to
their use in the isolation of B. influenzae. Br J Exp Pathol. 1929;10:226–36.
139. Bush K, Bradford PA. β-Lactams and β-Lactamase Inhibitors: an overview. Cold Spring Harb
Perspect Med. 2016;6(8):a025247.
140. Spratt BG. Penicillin-binding proteins and the future of β-lactam antibiotics. The seventh
fleming lecture. J Gen Microbiol. 1983;129(5):1247–60.
141. Sauvage E, Kerff F, Terrak M, Ayala JA, Charlier P. The penicillin-binding proteins: structure
and role in peptidoglycan biosynthesis. FEMS Microbiol Rev. 2008;32(2):234–58.
142. van Heijenoort J. Peptidoglycan hydrolases of Escherichia coli. Microbiol Mol Biol Rev.
2011;75(4):636–63.
143. Cho H, Uehara T, Bernhardt TG. Β-lactam antibiotics induce a lethal malfunctioning of the
bacterial cell wall synthesis machinery. Cell. 2014;159(6):1300–11.
144. Massova I, Mobashery S. Kinship and diversification of bacterial penicillin-binding proteins
and β-lactamases. Antimicrob Agents Chemother. 1998;42(1):1–17.
145. Bush K, Jacoby GA. Updated functional classification of β-lactamases. Antimicrob Agents
Chemother. 2010;54(3):969–76.
146. Medeiros AA. Evolution and dissemination of β-lactamases accelerated by generations of
β-lactam antibiotics. Clin Infect Dis. 1997;24(Suppl 1):S19–45.
147. Rossolini GM, Docquier JD. New β-lactamases: a paradigm for the rapid response of bacte-
rial evolution in the clinical setting. Future Microbiol. 2006;1(3):295–308.
148. De Pascale G, Wright GD. Antibiotic resistance by enzyme inactivation: from mechanisms to
solutions. Chembiochem. 2010;11(10):1325–34.
149. Datta N, Kontomichalou P. Penicillinase synthesis controlled by infectious R factors in
Enterobacteriaceae. Nature. 1965;208(5007):239–41.
150. Bonomo RA. β-Lactamases: a focus on current challenges. Cold Spring Harb Perspect Med.
2017;7(1):a025239.
151. Knothe H, Shah P, Krcmery V, Antal M, Mitsuhashi S. Transferable resistance to cefotaxime,
cefoxitin, cefamandole and cefuroxime in clinical isolates of Klebsiella pneumoniae and
Serratia marcescens. Infection. 1983;11(6):315–7.
152. Ramphal R, Ambrose PG. Extended-spectrum β-lactamases and clinical outcomes: current
data. Clin Infect Dis. 2006;42(Suppl 4):S164–72.
140 C. R. Dean et al.

153. Bush K. Extended-spectrum β-lactamases in North America, 1987–2006. Clin Microbiol

Infect. 2008;14(Suppl 1):134–43.
154. Bush K, Fisher JF. Epidemiological expansion, structural studies, and clinical challenges of
new β-lactamases from gram-negative bacteria. Annu Rev Microbiol. 2011;65:455–78.
155. Livermore DM. Defining an extended-spectrum β-lactamase. Clin Microbiol Infect.
2008;14(Suppl 1):3–10.
156. Fournier D, Hocquet D, Dehecq B, Cholley P, Plesiat P. Detection of a new extended-­spectrum
oxacillinase in Pseudomonas aeruginosa. J Antimicrob Chemother. 2010;65(2):364–5.
157. Castanheira M, Mendes RE, Jones RN, Sader HS. Changes in the frequencies of β-Lactamase
genes among Enterobacteriaceae isolates in U.S. hospitals, 2012 to 2014: activity of
ceftazidime-avibactam tested against β-Lactamase-producing isolates. Antimicrob Agents
Chemother. 2016;60(8):4770–7.
158. European Centre for Disease Prevention and Control. Antimicrobial resistance surveillance in
Europe 2015. Annual report of the European Antimicrobial Resistance Surveillance Network
(EARS-Net). Stockholm: ECDC; 2017.
159. Livermore DM, Canton R, Gniadkowski M, Nordmann P, Rossolini GM, Arlet G, et al. CTX-­
M: changing the face of ESBLs in Europe. J Antimicrob Chemother. 2007;59(2):165–74.
160. Naas T, Poirel L, Nordmann P. Minor extended-spectrum β-lactamases. Clin Microbiol
Infect. 2008;14(Suppl 1):42–52.
161. Viale P, Giannella M, Bartoletti M, Tedeschi S, Lewis R. Considerations about antimicro-
bial stewardship in settings with epidemic extended-spectrum β-lactamase-producing or
carbapenem-­resistant Enterobacteriaceae. Infect Dis Ther. 2015;4(Suppl 1):65–83.
162. Jacoby GA. AmpC β-lactamases. Clin Microbiol Rev. 2009;22(1):161–82.
163. Drawz SM, Taracila M, Caselli E, Prati F, Bonomo RA. Exploring sequence requirements
for C(3)/C(4) carboxylate recognition in the Pseudomonas aeruginosa cephalosporinase:
Insights into plasticity of the AmpC β-lactamase. Protein Sci. 2011;20(6):941–58.
164. Perez-Gallego M, Torrens G, Castillo-Vera J, Moya B, Zamorano L, Cabot G, et al. Impact
of AmpC derepression on fitness and virulence: the mechanism or the pathway? MBio.
2016;7(5):e01783.
165. Caroff N, Espaze E, Gautreau D, Richet H, Reynaud A. Analysis of the effects of -42 and -32
ampC promoter mutations in clinical isolates of Escherichia coli hyperproducing ampC. J
Antimicrob Chemother. 2000;45(6):783–8.
166. Forward KR, Willey BM, Low DE, McGeer A, Kapala MA, Kapala MM, et al. Molecular
mechanisms of cefoxitin resistance in Escherichia coli from the Toronto area hospitals.
Diagn Microbiol Infect Dis. 2001;41(1-2):57–63.
167. Olsson O, Bergstrom S, Lindberg FP, Normark S. ampC β-lactamase hyperproduction in
Escherichia coli: natural ampicillin resistance generated by horizontal chromosomal DNA
transfer from Shigella. Proc Natl Acad Sci U S A. 1983;80(24):7556–60.
168. Bauernfeind A, Stemplinger I, Jungwirth R, Wilhelm R, Chong Y. Comparative characteriza-
tion of the cephamycinase blaCMY-1 gene and its relationship with other β-lactamase genes.
Antimicrob Agents Chemother. 1996;40(8):1926–30.
169. Philippon A, Arlet G, Jacoby GA. Plasmid-determined AmpC-type β-lactamases. Antimicrob
Agents Chemother. 2002;46(1):1–11.
170. Alvarez M, Tran JH, Chow N, Jacoby GA. Epidemiology of conjugative plasmid-mediated
AmpC β-lactamases in the United States. Antimicrob Agents Chemother. 2004;48(2):533–7.
171. Tsakris A, Poulou A, Markou F, Pitiriga V, Piperaki ET, Kristo I, et al. Dissemination of clini-
cal isolates of Klebsiella oxytoca harboring CMY-31, VIM-1, and a New OXY-2-type variant
in the community. Antimicrob Agents Chemother. 2011;55(7):3164–8.
172. Pitout JD. Enterobacteriaceae that produce extended-spectrum β-lactamases and AmpC
β-lactamases in the community: the tip of the iceberg? Curr Pharm Des. 2013;19(2):257–63.
173. Wang XD, Cai JC, Zhou HW, Zhang R, Chen GX. Reduced susceptibility to carbapenems in
Klebsiella pneumoniae clinical isolates associated with plasmid-mediated β-lactamase pro-
duction and OmpK36 porin deficiency. J Med Microbiol. 2009;58(Pt 9):1196–202.
4 Resistance of Gram-negative Bacilli to Antimicrobials 141

174. Martinez-Martinez L, Pascual A, Hernandez-Alles S, Alvarez-Diaz D, Suarez AI, Tran J,

et al. Roles of β-lactamases and porins in activities of carbapenems and cephalosporins
against Klebsiella pneumoniae. Antimicrob Agents Chemother. 1999;43(7):1669–73.
175. Armand-Lefevre L, Leflon-Guibout V, Bredin J, Barguellil F, Amor A, Pages JM, et al.
Imipenem resistance in Salmonella enterica serovar Wien related to porin loss and CMY-4
β-lactamase production. Antimicrob Agents Chemother. 2003;47(3):1165–8.
176. Shin SY, Bae IK, Kim J, Jeong SH, Yong D, Kim JM, et al. Resistance to carbapenems in
sequence type 11 Klebsiella pneumoniae is related to DHA-1 and loss of OmpK35 and/or
OmpK36. J Med Microbiol. 2012;61(Pt 2):239–45.
177. Tsai YK, Liou CH, Fung CP, Lin JC, Siu LK. Single or in combination antimicrobial resis-
tance mechanisms of Klebsiella pneumoniae contribute to varied susceptibility to different
carbapenems. PLoS One. 2013;8(11):e79640.
178. Queenan AM, Bush K. Carbapenemases: the versatile β-lactamases. Clin Microbiol Rev.
2007;20(3):440–58.
179. Walsh TR. Emerging carbapenemases: a global perspective. Int J Antimicrob Agents.
2010;36(Suppl 3):S8–14.
180. Yang YJ, Wu PJ, Livermore DM. Biochemical characterization of a β-lactamase that hydro-
lyzes penems and carbapenems from two Serratia marcescens isolates. Antimicrob Agents
Chemother. 1990;34(5):755–8.
181. Meletis G. Carbapenem resistance: overview of the problem and future perspectives. Ther
Adv Infect Dis. 2016;3(1):15–21.
182. Archibald L, Phillips L, Monnet D, McGowan JE Jr, Tenover F, Gaynes R. Antimicrobial
resistance in isolates from inpatients and outpatients in the United States: increasing impor-
tance of the intensive care unit. Clin Infect Dis. 1997;24(2):211–5.
183. Yigit H, Queenan AM, Anderson GJ, Domenech-Sanchez A, Biddle JW, Steward CD, et al.
Novel carbapenem-hydrolyzing β-lactamase, KPC-1, from a carbapenem-resistant strain of
Klebsiella pneumoniae. Antimicrob Agents Chemother. 2001;45(4):1151–61.
184. Cristina ML, Sartini M, Ottria G, Schinca E, Cenderello N, Crisalli MP, et al. Epidemiology
and biomolecular characterization of carbapenem-resistant klebsiella pneumoniae in an
Italian hospital. J Prev Med Hyg. 2016;57(3):E149–E56.
185. Evans BA, Amyes SG. OXA β-lactamases. Clin Microbiol Rev. 2014;27(2):241–63.
186. Donald HM, Scaife W, Amyes SG, Young HK. Sequence analysis of ARI-1, a novel OXA
β-lactamase, responsible for imipenem resistance in Acinetobacter baumannii 6B92.
Antimicrob Agents Chemother. 2000;44(1):196–9.
187. Potter RF, D’Souza AW, Dantas G. The rapid spread of carbapenem-resistant
Enterobacteriaceae. Drug Resist Updat. 2016;29:30–46.
188. Poirel L, Heritier C, Tolun V, Nordmann P. Emergence of oxacillinase-mediated resistance to
imipenem in Klebsiella pneumoniae. Antimicrob Agents Chemother. 2004;48(1):15–22.
189. Branas P, Villa J, Viedma E, Mingorance J, Orellana MA, Chaves F. Molecular epidemiol-
ogy of carbapenemase-producing Klebsiella pneumoniae in a hospital in Madrid: Successful
establishment of an OXA-48 ST11 clone. Int J Antimicrob Agents. 2015;46(1):111–6.
190. Karampatakis T, Antachopoulos C, Iosifidis E, Tsakris A, Roilides E. Molecular epide-
miology of carbapenem-resistant Klebsiella pneumoniae in Greece. Future Microbiol.
2016;11:809–23.
191. Walsh TR, Toleman MA, Poirel L, Nordmann P. Metallo-β-lactamases: the quiet before the
storm? Clin Microbiol Rev. 2005;18(2):306–25.
192. Lauretti L, Riccio ML, Mazzariol A, Cornaglia G, Amicosante G, Fontana R, et al.
Cloning and characterization of blaVIM, a new integron-borne metallo-β-lactamase
gene from a Pseudomonas aeruginosa clinical isolate. Antimicrob Agents Chemother.
1999;43(7):1584–90.
193. Giakkoupi P, Petrikkos G, Tzouvelekis LS, Tsonas S, Legakis NJ, Vatopoulos AC, et al.
Spread of integron-associated VIM-type metallo-β-lactamase genes among imipenem-­
nonsusceptible Pseudomonas aeruginosa strains in Greek hospitals. J Clin Microbiol.
2003;41(2):822–5.
142 C. R. Dean et al.

194. Zmarlicka MT, Nailor MD, Nicolau DP. Impact of the New Delhi metallo-β-lactamase on
β-lactam antibiotics. Infect Drug Resist. 2015;8:297–309.
195. Borgia S, Lastovetska O, Richardson D, Eshaghi A, Xiong J, Chung C, et al. Outbreak of
carbapenem-­ resistant Enterobacteriaceae containing blaNDM-1, Ontario, Canada. Clin
Infect Dis. 2012;55(11):e109–17.
196. Sugawara E, Kojima S, Nikaido H. Klebsiella pneumoniae major porins OmpK35 and
OmpK36 allow more efficient diffusion of β-Lactams than their Escherichia coli Homologs
OmpF and OmpC. J Bacteriol. 2016;198(23):3200–8.
197. Adams-Sapper S, Nolen S, Donzelli GF, Lal M, Chen K, Justo da Silva LH, et al. Rapid
induction of high-level carbapenem resistance in heteroresistant KPC-producing Klebsiella
pneumoniae. Antimicrob Agents Chemother. 2015;59(6):3281–9.
198. Castanheira M, Mills JC, Farrell DJ, Jones RN. Mutation-driven β-lactam resistance mecha-
nisms among contemporary ceftazidime-nonsusceptible Pseudomonas aeruginosa isolates
from U.S. hospitals. Antimicrob Agents Chemother. 2014;58(11):6844–50.
199. Moya B, Beceiro A, Cabot G, Juan C, Zamorano L, Alberti S, et al. Pan-β-lactam resis-
tance development in Pseudomonas aeruginosa clinical strains: molecular mechanisms,
penicillin-binding protein profiles, and binding affinities. Antimicrob Agents Chemother.
2012;56(9):4771–8.
200. Bradford PA, Urban C, Mariano N, Projan SJ, Rahal JJ, Bush K. Imipenem resistance in
Klebsiella pneumoniae is associated with the combination of ACT-1, a plasmid-­mediated
AmpC β-lactamase, and the foss of an outer membrane protein. Antimicrob Agents
Chemother. 1997;41(3):563–9.
201. Yamachika S, Sugihara C, Kamai Y, Yamashita M. Correlation between penicillin-­binding
protein 2 mutations and carbapenem resistance in Escherichia coli. J Med Microbiol.
2013;62(Pt 3):429–36.
202. Alm RA, Johnstone MR, Lahiri SD. Characterization of Escherichia coli NDM isolates
with decreased susceptibility to aztreonam/avibactam: role of a novel insertion in PBP3.
J Antimicrob Chemother. 2015;70(5):1420–8.
203. Zhang Y, Kashikar A, Brown CA, Denys G, Bush K. Unusual Escherichia coli PBP 3
insertion sequence identified from a collection of carbapenem-resistant Enterobacteriaceae
tested In vitro with a combination of ceftazidime-, ceftaroline-, or aztreonam-avibactam.
Antimicrob Agents Chemother. 2017;61(8):e00389.
204. Durand-Reville TF, Guler S, Comita-Prevoir J, Chen B, Bifulco N, Huynh H, et al. ETX2514
is a broad-spectrum β-lactamase inhibitor for the treatment of drug-resistant Gram-negative
bacteria including Acinetobacter baumannii. Nat Microbiol. 2017;2:17104.
205. Reading C, Cole M. Clavulanic acid: a β-lactamase-inhiting β-lactam from Streptomyces
clavuligerus. Antimicrob Agents Chemother. 1977;11(5):852–7.
206. English AR, Retsema JA, Girard AE, Lynch JE, Barth WE. CP-45,899, a β-lactamase inhibi-
tor that extends the antibacterial spectrum of β-lactams: initial bacteriological characteriza-
tion. Antimicrob Agents Chemother. 1978;14(3):414–9.
207. Fisher J, Belasco JG, Charnas RL, Khosla S, Knowles JR. Β-lactamase inactivation by
mechanism-­based reagents. Philos Trans R Soc Lond Ser B Biol Sci. 1980;289(1036):309–19.
208. Cole M. Biochemistry and action of clavulanic acid. Scott Med J. 1982;27:S10–6.
209. Brown RP, Aplin RT, Schofield CJ. Inhibition of TEM-2 β-lactamase from Escherichia coli
by clavulanic acid: observation of intermediates by electrospray ionization mass spectrom-
etry. Biochemistry. 1996;35(38):12421–32.
210. Drawz SM, Bonomo RA. Three decades of β-lactamase inhibitors. Clin Microbiol Rev.
2010;23(1):160–201.
211. Sulton D, Pagan-Rodriguez D, Zhou X, Liu Y, Hujer AM, Bethel CR, et al. Clavulanic acid
inactivation of SHV-1 and the inhibitor-resistant S130G SHV-1 β-lactamase. Insights into the
mechanism of inhibition. J Biol Chem. 2005;280(42):35528–36.
212. Weber DA, Sanders CC. Diverse potential of β-lactamase inhibitors to induce class I enzymes.
Antimicrob Agents Chemother. 1990;34(1):156–8.
4 Resistance of Gram-negative Bacilli to Antimicrobials 143

213. Coleman K. Diazabicyclooctanes (DBOs): a potent new class of non-β-lactam β-lactamase

inhibitors. Curr Opin Microbiol. 2011;14(5):550–5.
214. Ehmann DE, Jahic H, Ross PL, Gu RF, Hu J, Kern G, et al. Avibactam is a covalent, revers-
ible, non-β-lactam β-lactamase inhibitor. Proc Natl Acad Sci U S A. 2012;109(29):11663–8.
215. Humphries RM, Hemarajata P. Resistance to ceftazidime-avibactam in Klebsiella pneu-
moniae due to porin mutations and the increased expression of KPC-3. Antimicrob Agents
Chemother. 2017;61(6):e00537.
216. Marshall S, Hujer AM, Rojas LJ, Papp-Wallace KM, Humphries RM, Spellberg B, et al. Can
ceftazidime-avibactam and aztreonam overcome β-lactam resistance conferred by metallo-β-­
lactamases in Enterobacteriaceae? Antimicrob Agents Chemother. 2017;61(4):e02243.
217. Fass RJ, Gregory WW, D’Amato RF, Matsen JM, Wright DN, Young LS. In vitro activities of
cefoperazone and sulbactam singly and in combination against cefoperazone-resistant mem-
bers of the family Enterobacteriaceae and nonfermenters. Antimicrob Agents Chemother.
1990;34(11):2256–9.
218. Hecker SJ, Reddy KR, Totrov M, Hirst GC, Lomovskaya O, Griffith DC, et al. Discovery of a
cyclic boronic acid β-lactamase inhibitor (RPX7009) with utility vs class A serine carbapen-
emases. J Med Chem. 2015;58(9):3682–92.
219. Lapuebla A, Abdallah M, Olafisoye O, Cortes C, Urban C, Quale J, et al. Activity of merope-
nem combined with RPX7009, a novel β-Lactamase inhibitor, against Gram-negative clinical
isolates in New York City. Antimicrob Agents Chemother. 2015;59(8):4856–60.
220. Castanheira M, Rhomberg PR, Flamm RK, Jones RN. Effect of the β-Lactamase inhibi-
tor vaborbactam combined with meropenem against serine carbapenemase-producing
Enterobacteriaceae. Antimicrob Agents Chemother. 2016;60(9):5454–8.
221. Lomovskaya O, Sun D, Rubio-Aparicio D, Nelson K, Tsivkovski R, Griffith DC, et al.
Vaborbactam: spectrum of Β-Lactamase inhibition and impact of resistance mechanisms on
activity in Enterobacteriaceae. Antimicrob Agents Chemother. 2017;61:e01443.
222. Drawz SM, Papp-Wallace KM, Bonomo RA. New β-lactamase inhibitors: a therapeutic
renaissance in an MDR world. Antimicrob Agents Chemother. 2014;58(4):1835–46.
223. Reck F, Bermingham A, Blais J, Capka V, Cariaga T, Casarez A, et al. Optimization of novel
monobactams with activity against carbapenem-resistant Enterobacteriaceae - Identification
of LYS228. Bioorg Med Chem Lett. 2018;28(4):748-755. doi: 10.1016/j.bmcl.2018.01.006.
224. Blais J, Lopez S, Ruzin A, Leeds JA, Dean CR, Simmons RL, Casarez A, Reck F. LYS228 is
a novel monobactam with potent activity in vitro against ESBL-producing and Carbapenem-­
resistant Enterobacteriaceae. Abstracts of microbe 2017. New Orleans, USA; 2017; June 1–5.
225. Mendes RE, Rhomberg PR, Schaeffer B, Huband MD, Flamm RK. In vitro activity of
LYS228 against Enterobacteriaceae, including molecularly characterized multidrug-resistant
isolates. Abstracts of microbe 2017. New Orleans; 2017; June 1–5.
226. King AM, Reid-Yu SA, Wang W, King DT, De Pascale G, Strynadka NC, et al.
Aspergillomarasmine A overcomes metallo-β-lactamase antibiotic resistance. Nature.
2014;510(7506):503–6.
227. Hinchliffe P, Gonzalez MM, Mojica MF, Gonzalez JM, Castillo V, Saiz C, et al. Cross-class
metallo-β-lactamase inhibition by bisthiazolidines reveals multiple binding modes. Proc Natl
Acad Sci U S A. 2016;113(26):E3745–54.
228. Hecker SJ, Reddy KR, Glinka TW, Totrov M, Lomovskaya O, Tsivkovski R, Rubio-Aparicio
D, Clifton MC, Atkins K, Fairman JF, Griffith DC, Dudley MN. Discovery of a new series
of broad-spectrum carbapenemase inhibitors (BCIs) with activity vs serine and metallo
Β-lactamases. Abstracts of the 55th interscience conference on antimicrobial agents and che-
motherapy. San Diego: 2015; F–270.
229. Jones-Dias D, Manageiro V, Ferreira E, Louro D. Antibiotic resistance surveillance pro-
gram in Portugal p, Canica M. Diversity of extended-spectrum and plasmid-mediated
AmpC β-lactamases in Enterobacteriaceae isolates from Portuguese health care facilities.
J Microbiol. 2014;52(6):496–503.
144 C. R. Dean et al.

230. Canton R, Morosini MI, de la Maza OM, de la Pedrosa EG. IRT and CMT β-lactamases and
inhibitor resistance. Clin Microbiol Infect. 2008;14(Suppl 1):53–62.
231. Martinez JL, Cercenado E, Rodriguez-Creixems M, Vincente-Perez MF, Delgado-Iribarren
A, Baquero F. Resistance to β-lactam/clavulanate. Lancet. 1987;2(8573):1473.
232. Martinez JL, Vicente MF, Delgado-Iribarren A, Perez-Diaz JC, Baquero F. Small plasmids
are involved in amoxicillin-clavulanate resistance in Escherichia coli. Antimicrob Agents
Chemother. 1989;33(4):595.
233. Lister PD, Gardner VM, Sanders CC. Clavulanate induces expression of the Pseudomonas
aeruginosa AmpC cephalosporinase at physiologically relevant concentrations and antago-
nizes the antibacterial activity of ticarcillin. Antimicrob Agents Chemother. 1999;43(4):882–9.
234. Derbyshire H, Kay G, Evans K, Vaughan C, Kavuri U, Winstanley T. A simple disc diffu-
sion method for detecting AmpC and extended-spectrum β-lactamases in clinical isolates of
Enterobacteriaceae. J Antimicrob Chemother. 2009;63(3):497–501.
235. Yan JJ, Ko WC, Jung YC, Chuang CL, Wu JJ. Emergence of Klebsiella pneumoniae Isolates
producing inducible DHA-1 -lactamase in a university hospital in Taiwan. J Clin Microbiol.
2002;40(9):3121–6.
236. Nikaido H, Pages JM. Broad-specificity efflux pumps and their role in multidrug resistance
of Gram-negative bacteria. FEMS Microbiol Rev. 2012;36(2):340–63.
237. Pages JM, Peslier S, Keating TA, Lavigne JP, Nichols WW. Role of the outer membrane
and porins in susceptibility of β-lactamase-oroducing Enterobacteriaceae to ceftazidime-­
avibactam. Antimicrob Agents Chemother. 2015;60(3):1349–59.
238. Lapuebla A, Abdallah M, Olafisoye O, Cortes C, Urban C, Landman D, et al. Activity of imi-
penem with relebactam against Gram-Negative pathogens from New York City. Antimicrob
Agents Chemother. 2015;59(8):5029–31.
239. Winkler ML, Papp-Wallace KM, Hujer AM, Domitrovic TN, Hujer KM, Hurless KN, et al.
Unexpected challenges in treating multidrug-resistant Gram-negative bacteria: resistance to
ceftazidime-avibactam in archived isolates of Pseudomonas aeruginosa. Antimicrob Agents
Chemother. 2015;59(2):1020–9.
240. Livermore DM, Mushtaq S, Barker K, Hope R, Warner M, Woodford N. Characterization of
β-lactamase and porin mutants of Enterobacteriaceae selected with ceftaroline + avibactam
(NXL104). J Antimicrob Chemother. 2012;67(6):1354–8.
241. Blazquez J, Baquero MR, Canton R, Alos I, Baquero F. Characterization of a new TEM-­
type β-lactamase resistant to clavulanate, sulbactam, and tazobactam in a clinical isolate of
Escherichia coli. Antimicrob Agents Chemother. 1993;37(10):2059–63.
242. Vedel G, Belaaouaj A, Gilly L, Labia R, Philippon A, Nevot P, et al. Clinical isolates of
Escherichia coli producing TRI β-lactamases: novel TEM-enzymes conferring resistance to
β-lactamase inhibitors. J Antimicrob Chemother. 1992;30(4):449–62.
243. Belaaouaj A, Lapoumeroulie C, Canica MM, Vedel G, Nevot P, Krishnamoorthy R, et al.
Nucleotide sequences of the genes coding for the TEM-like β-lactamases IRT-1 and IRT-2
(formerly called TRI-1 and TRI-2). FEMS Microbiol Lett. 1994;120(1-2):75–80.
244. Lemozy J, Sirot D, Chanal C, Huc C, Labia R, Dabernat H, et al. First characterization
of inhibitor-resistant TEM (IRT) β-lactamases in Klebsiella pneumoniae strains. Antimicrob
Agents Chemother. 1995;39(11):2580–2.
245. Bret L, Chanal C, Sirot D, Labia R, Sirot J. Characterization of an inhibitor-resistant enzyme
IRT-2 derived from TEM-2 β-lactamase produced by Proteus mirabilis strains. J Antimicrob
Chemother. 1996;38(2):183–91.
246. Sirot D, Chanal C, Bonnet R, De Champs C, Bret L. Inhibitor-resistant TEM-33 β-lactamase
in a Shigella sonnei isolate. Antimicrob Agents Chemother. 2001;45(7):2179–80.
247. Hunter JE, Corkill JE, McLennan AG, Fletcher JN, Hart CA. Plasmid encoded β-lactamases
resistant to inhibition by clavulanic acid produced by calf faecal coliforms. Res Vet Sci.
1993;55(3):367–70.
248. Manageiro V, Ferreira E, Cougnoux A, Albuquerque L, Canica M, Bonnet R. Characterization
of the inhibitor-resistant SHV β-lactamase SHV-107 in a clinical Klebsiella pneumoniae
strain coproducing GES-7 enzyme. Antimicrob Agents Chemother. 2012;56(2):1042–6.
4 Resistance of Gram-negative Bacilli to Antimicrobials 145

249. Bonomo RA, Rudin SA, Shlaes DM. Tazobactam is a potent inactivator of selected inhibitor-­
resistant class A β-lactamases. FEMS Microbiol Lett. 1997;148(1):59–62.
250. Chaibi EB, Sirot D, Paul G, Labia R. Inhibitor-resistant TEM β-lactamases: phenotypic,
genetic and biochemical characteristics. J Antimicrob Chemother. 1999;43(4):447–58.
251. Pagan-Rodriguez D, Zhou X, Simmons R, Bethel CR, Hujer AM, Helfand MS, et al.
Tazobactam inactivation of SHV-1 and the inhibitor-resistant Ser130 -->Gly SHV-1
β-lactamase: insights into the mechanism of inhibition. J Biol Chem. 2004;279(19):19494–501.
252. Nordmann P, Cuzon G, Naas T. The real threat of Klebsiella pneumoniae carbapenemase-­
producing bacteria. Lancet Infect Dis. 2009;9(4):228–36.
253. Papp-Wallace KM, Bethel CR, Distler AM, Kasuboski C, Taracila M, Bonomo RA. Inhibitor
resistance in the KPC-2 β-lactamase, a preeminent property of this class A β-lactamase.
Antimicrob Agents Chemother. 2010;54(2):890–7.
254. Livermore DM, Warner M, Jamrozy D, Mushtaq S, Nichols WW, Mustafa N, et al. In vitro
selection of ceftazidime-avibactam resistance in Enterobacteriaceae with KPC-3 carbapen-
emase. Antimicrob Agents Chemother. 2015;59(9):5324–30.
255. Winkler ML, Papp-Wallace KM, Taracila MA, Bonomo RA. Avibactam and inhibitor-­
resistant SHV β-lactamases. Antimicrob Agents Chemother. 2015;59(7):3700–9.
256. Shields RK, Chen L, Cheng S, Chavda KD, Press EG, Snyder A, et al. Emergence of
ceftazidime-­avibactam resistance due to plasmid-borne blaKPC-3 mutations during treatment
of carbapenem-resistant Klebsiella pneumoniae infections. Antimicrob Agents Chemother.
2017;61(3):e02097.
257. Lahiri SD, Bradford PA, Nichols WW, Alm RA. Structural and sequence analysis of class
A β-lactamases with respect to avibactam inhibition: impact of Omega-loop variations.
J Antimicrob Chemother. 2016;71(10):2848–55.
258. Drlica K, Malik M. Fluoroquinolones: action and resistance. Curr Top Med Chem.
2003;3(3):249–82.
259. Champoux JJ. DNA topoisomerases: structure, function, and mechanism. Annu Rev
Biochem. 2001;70:369–413.
260. Drlica K, Hiasa H, Kerns R, Malik M, Mustaev A, Zhao X. Quinolones: action and resistance
updated. Curr Top Med Chem. 2009;9(11):981–98.
261. Hiasa H, Yousef DO, Marians KJ. DNA strand cleavage is required for replication
fork arrest by a frozen topoisomerase-quinolone-DNA ternary complex. J Biol Chem.
1996;271(42):26424–9.
262. Willmott CJ, Critchlow SE, Eperon IC, Maxwell A. The complex of DNA gyrase and qui-
nolone drugs with DNA forms a barrier to transcription by RNA polymerase. J Mol Biol.
1994;242(4):351–63.
263. Shea ME, Hiasa H. Interactions between DNA helicases and frozen topoisomerase
IV-quinolone-DNA ternary complexes. J Biol Chem. 1999;274(32):22747–54.
264. Brazas MD, Hancock RE. Ciprofloxacin induction of a susceptibility determinant in
Pseudomonas aeruginosa. Antimicrob Agents Chemother. 2005;49(8):3222–7.
265. Economou V, Gousia P. Agriculture and food animals as a source of antimicrobial-resistant
bacteria. Infect Drug Resist. 2015;8:49–61.
266. Hooper DC, Jacoby GA. Mechanisms of drug resistance: quinolone resistance. Ann N Y
Acad Sci. 2015;1354:12–31.
267. Morais Cabral JH, Jackson AP, Smith CV, Shikotra N, Maxwell A, Liddington RC. Crystal
structure of the breakage-reunion domain of DNA gyrase. Nature. 1997;388(6645):903–6.
268. Yoshida H, Bogaki M, Nakamura M, Nakamura S. Quinolone resistance-determining
region in the DNA gyrase gyrA gene of Escherichia coli. Antimicrob Agents Chemother.
1990;34(6):1271–2.
269. Laponogov I, Veselkov DA, Crevel IM, Pan XS, Fisher LM, Sanderson MR. Structure of an
‘open’ clamp type II topoisomerase-DNA complex provides a mechanism for DNA capture
and transport. Nucleic Acids Res. 2013;41(21):9911–23.
270. Wohlkonig A, Chan PF, Fosberry AP, Homes P, Huang J, Kranz M, et al. Structural basis of
quinolone inhibition of type IIA topoisomerases and target-mediated resistance. Nat Struct
Mol Biol. 2010;17(9):1152–3.
146 C. R. Dean et al.

271. Willmott CJ, Maxwell A. A single point mutation in the DNA gyrase A protein greatly reduces
binding of fluoroquinolones to the gyrase-DNA complex. Antimicrob Agents Chemother.
1993;37(1):126–7.
272. Aldred KJ, Kerns RJ, Osheroff N. Mechanism of quinolone action and resistance.
Biochemistry. 2014;53(10):1565–74.
273. Bagel S, Hullen V, Wiedemann B, Heisig P. Impact of gyrA and parC mutations on quinolone
resistance, doubling time, and supercoiling degree of Escherichia coli. Antimicrob Agents
Chemother. 1999;43(4):868–75.
274. Komp Lindgren P, Marcusson LL, Sandvang D, Frimodt-Moller N, Hughes D. Biological
cost of single and multiple norfloxacin resistance mutations in Escherichia coli implicated in
urinary tract infections. Antimicrob Agents Chemother. 2005;49(6):2343–51.
275. Bruchmann S, Dotsch A, Nouri B, Chaberny IF, Haussler S. Quantitative contributions of tar-
get alteration and decreased drug accumulation to Pseudomonas aeruginosa fluoroquinolone
resistance. Antimicrob Agents Chemother. 2013;57(3):1361–8.
276. Kriengkauykiat J, Porter E, Lomovskaya O, Wong-Beringer A. Use of an efflux pump
inhibitor to determine the prevalence of efflux pump-mediated fluoroquinolone resistance
and multidrug resistance in Pseudomonas aeruginosa. Antimicrob Agents Chemother.
2005;49(2):565–70.
277. Paltansing S, Tengeler AC, Kraakman ME, Claas EC, Bernards AT. Exploring the contribu-
tion of efflux on the resistance to fluoroquinolones in clinical isolates of Escherichia coli.
Microb Drug Resist. 2013;19(6):469–76.
278. Mazzariol A, Zuliani J, Cornaglia G, Rossolini GM, Fontana R. AcrAB Efflux system:
expression and contribution to fluoroquinolone resistance in Klebsiella spp. Antimicrob
Agents Chemother. 2002;46(12):3984–6.
279. Le Thomas I, Couetdic G, Clermont O, Brahimi N, Plesiat P, Bingen E. In vivo selection
of a target/efflux double mutant of Pseudomonas aeruginosa by ciprofloxacin therapy.
J Antimicrob Chemother. 2001;48(4):553–5.
280. Chapman JS, Georgopapadakou NH. Routes of quinolone permeation in Escherichia coli.
Antimicrob Agents Chemother. 1988;32(4):438–42.
281. Chenia HY, Pillay B, Pillay D. Analysis of the mechanisms of fluoroquinolone resistance in
urinary tract pathogens. J Antimicrob Chemother. 2006;58(6):1274–8.
282. Miro E, Verges C, Garcia I, Mirelis B, Navarro F, Coll P, et al. Resistance to quinolones and
β-lactams in Salmonella enterica due to mutations in topoisomerase-encoding genes, altered
cell permeability and expression of an active efflux system. Enferm Infecc Microbiol Clin.
2004;22(4):204–11.
283. Randall LP, Woodward MJ. The multiple antibiotic resistance (mar) locus and its signifi-
cance. Res Vet Sci. 2002;72(2):87–93.
284. Martínez-Martínez L, Pascual A, Jacoby GA. Quinolone resistance from a transferable plas-
mid. Lancet. 1998;351(9105):797–9.
285. Jacoby GA, Walsh KE, Mills DM, Walker VJ, Oh H, Robicsek A, et al. qnrB, another plasmid-­
mediated gene for quinolone resistance. Antimicrob Agents Chemother. 2006;50(4):1178–82.
286. Hata M, Suzuki M, Matsumoto M, Takahashi M, Sato K, Ibe S, et al. Cloning of a novel
gene for quinolone resistance from a transferable plasmid in Shigella flexneri 2b. Antimicrob
Agents Chemother. 2005;49(2):801–3.
287. Vetting MW, Hegde SS, Fajardo JE, Fiser A, Roderick SL, Takiff HE, et al. Pentapeptide
repeat proteins. Biochemistry. 2006;45(1):1–10.
288. Tran JH, Jacoby GA. Mechanism of plasmid-mediated quinolone resistance. Proc Natl Acad
Sci U S A. 2002;99(8):5638–42.
289. Tran JH, Jacoby GA, Hooper DC. Interaction of the plasmid-encoded quinolone resistance
protein QnrA with Escherichia coli topoisomerase IV. Antimicrob Agents Chemother.
2005;49(7):3050–2.
290. Tran JH, Jacoby GA, Hooper DC. Interaction of the plasmid-encoded quinolone resis-
tance protein Qnr with Escherichia coli DNA gyrase. Antimicrob Agents Chemother.
2005;49(1):118–25.
4 Resistance of Gram-negative Bacilli to Antimicrobials 147

291. Vetting MW, Hegde SS, Wang M, Jacoby GA, Hooper DC, Blanchard JS. Structure of QnrB1,
a plasmid-mediated fluoroquinolone resistance factor. J Biol Chem. 2011;286(28):25265–73.
292. Jacoby GA, Hooper DC. Phylogenetic analysis of chromosomally determined qnr and related
proteins. Antimicrob Agents Chemother. 2013;57(4):1930–4.
293. Robicsek A, Strahilevitz J, Sahm DF, Jacoby GA, Hooper DC. qnr prevalence in ceftazidime-­
resistant Enterobacteriaceae isolates from the United States. Antimicrob Agents Chemother.
2006;50(8):2872–4.
294. Briales A, Rodriguez-Martinez JM, Velasco C, de Alba PD, Rodriguez-Bano J, Martinez-­
Martinez L, et al. Prevalence of plasmid-mediated quinolone resistance determinants qnr and
aac(6’)-Ib-cr in Escherichia coli and Klebsiella pneumoniae producing extended-spectrum
β-lactamases in Spain. Int J Antimicrob Agents. 2012;39(5):431–4.
295. Robicsek A, Jacoby GA, Hooper DC. The worldwide emergence of plasmid-mediated quino-
lone resistance. Lancet Infect Dis. 2006;6(10):629–40.
296. Jacoby GA, Strahilevitz J, Hooper DC. Plasmid-mediated quinolone resistance. Microbiol
Spectr. 2014;2(5). https://doi.org/10.1128/microbiolspec.
297. Guillard T, Grillon A, de Champs C, Cartier C, Madoux J, Bercot B, et al. Mobile insertion
cassette elements found in small non-transmissible plasmids in Proteeae may explain qnrD
mobilization. PLoS One. 2014;9(2):e87801.
298. Robicsek A, Strahilevitz J, Jacoby GA, Macielag M, Abbanat D, Park CH, et al.
Fluoroquinolone-modifying enzyme: a new adaptation of a common aminoglycoside acetyl-
transferase. Nat Med. 2006;12(1):83–8.
299. Hansen LH, Jensen LB, Sorensen HI, Sorensen SJ. Substrate specificity of the OqxAB
multidrug resistance pump in Escherichia coli and selected enteric bacteria. J Antimicrob
Chemother. 2007;60(1):145–7.
300. Hansen LH, Johannesen E, Burmolle M, Sorensen AH, Sorensen SJ. Plasmid-encoded multi-
drug efflux pump conferring resistance to olaquindox in Escherichia coli. Antimicrob Agents
Chemother. 2004;48(9):3332–7.
301. Sorensen AH, Hansen LH, Johannesen E, Sorensen SJ. Conjugative plasmid conferring resis-
tance to olaquindox. Antimicrob Agents Chemother. 2003;47(2):798–9.
302. Kim HB, Wang M, Park CH, Kim EC, Jacoby GA, Hooper DC. oqxAB encoding a multidrug
efflux pump in human clinical isolates of Enterobacteriaceae. Antimicrob Agents Chemother.
2009;53(8):3582–4.
303. Li L, Liao X, Yang Y, Sun J, Li L, Liu B, et al. Spread of oqxAB in Salmonella enterica
serotype Typhimurium predominantly by IncHI2 plasmids. J Antimicrob Chemother.
2013;68(10):2263–8.
304. Liu BT, Yang QE, Li L, Sun J, Liao XP, Fang LX, et al. Dissemination and characteriza-
tion of plasmids carrying oqxAB-bla CTX-M genes in Escherichia coli isolates from food-­
producing animals. PLoS One. 2013;8(9):e73947.
305. Zhao J, Chen Z, Chen S, Deng Y, Liu Y, Tian W, et al. Prevalence and dissemination of oqxAB
in Escherichia coli isolates from animals, farmworkers, and the environment. Antimicrob
Agents Chemother. 2010;54(10):4219–24.
306. Perez F, Rudin SD, Marshall SH, Coakley P, Chen L, Kreiswirth BN, et al. OqxAB, a quino-
lone and olaquindox efflux pump, is widely distributed among multidrug-resistant Klebsiella
pneumoniae isolates of human origin. Antimicrob Agents Chemother. 2013;57(9):4602–3.
307. Wong MH, Chan EW, Chen S. Evolution and dissemination of OqxAB-like efflux pumps,
an emerging quinolone resistance determinant among members of Enterobacteriaceae.
Antimicrob Agents Chemother. 2015;59(6):3290–7.
308. Bialek-Davenet S, Lavigne JP, Guyot K, Mayer N, Tournebize R, Brisse S, et al. Differential
contribution of AcrAB and OqxAB efflux pumps to multidrug resistance and virulence in
Klebsiella pneumoniae. J Antimicrob Chemother. 2015;70(1):81–8.
309. Yamane K, Wachino J, Suzuki S, Kimura K, Shibata N, Kato H, et al. New plasmid-mediated
fluoroquinolone efflux pump, QepA, found in an Escherichia coli clinical isolate. Antimicrob
Agents Chemother. 2007;51(9):3354–60.
148 C. R. Dean et al.

310. Cattoir V, Poirel L, Nordmann P. Plasmid-mediated quinolone resistance pump QepA2 in an

Escherichia coli isolate from France. Antimicrob Agents Chemother. 2008;52(10):3801–4.
311. Perichon B, Courvalin P, Galimand M. Transferable resistance to aminoglycosides by meth-
ylation of G1405 in 16S rRNA and to hydrophilic fluoroquinolones by QepA-mediated efflux
in Escherichia coli. Antimicrob Agents Chemother. 2007;51(7):2464–9.
312. Zhao X, Drlica K. Restricting the selection of antibiotic-resistant mutant bacteria: measure-
ment and potential use of the mutant selection window. J Infect Dis. 2002;185(4):561–5.
313. Singh R, Swick MC, Ledesma KR, Yang Z, Hu M, Zechiedrich L, et al. Temporal inter-
play between efflux pumps and target mutations in development of antibiotic resistance in
Escherichia coli. Antimicrob Agents Chemother. 2012;56(4):1680–5.
314. Lomovskaya O, Lee A, Hoshino K, Ishida H, Mistry A, Warren MS, et al. Use of a genetic
approach to evaluate the consequences of inhibition of efflux pumps in Pseudomonas aerugi-
nosa. Antimicrob Agents Chemother. 1999;43(6):1340–6.
315. Poole K. Efflux pumps as antimicrobial resistance mechanisms. Ann Med. 2007;39(3):162–76.
316. Hsu DI, Okamoto MP, Murthy R, Wong-Beringer A. Fluoroquinolone-resistant Pseudomonas
aeruginosa: risk factors for acquisition and impact on outcomes. J Antimicrob Chemother.
2005;55(4):535–41.
317. Neuhauser MM, Weinstein RA, Rydman R, Danziger LH, Karam G, Quinn JP. Antibiotic
resistance among gram-negative bacilli in US intensive care units: implications for fluoroqui-
nolone use. JAMA. 2003;289(7):885–8.
318. Poirel L, Cattoir V, Nordmann P. Is plasmid-mediated quinolone resistance a clinically sig-
nificant problem? Clin Microbiol Infect. 2008;14(4):295–7.
319. Cirz RT, Chin JK, Andes DR, de Crecy-Lagard V, Craig WA, Romesberg FE. Inhibition of
mutation and combating the evolution of antibiotic resistance. PLoS Biol. 2005;3(6):e176.
320. Bassetti M, Della Siega P, Pecori D, Scarparo C, Righi E. Delafloxacin for the treatment of
respiratory and skin infections. Expert Opin Investig Drugs. 2015;24(3):433–42.
321. Hoover R, Marbury TC, Preston RA, Quintas M, Lawrence LE, Paulson SK, et al. Clinical
pharmacology of delafloxacin in patients with hepatic impairment. J Clin Pharmacol.
2017;57(3):328–35.
322. Flamm RK, Rhomberg PR, Huband MD, Farrell DJ. In vitro activity of delafloxacin tested
against isolates of Streptococcus pneumoniae, Haemophilus influenzae, and Moraxella
catarrhalis. Antimicrob Agents Chemother. 2016;60(10):6381–5.
323. Nayar AS, Dougherty TJ, Reck F, Thresher J, Gao N, Shapiro AB, et al. Target-based resis-
tance in Pseudomonas aeruginosa and Escherichia coli to NBTI 5463, a novel bacterial type
II topoisomerase inhibitor. Antimicrob Agents Chemother. 2015;59(1):331–7.
324. Tessier PR, Nicolau DP. In vitro activity of novel gyrase inhibitors against a highly resistant
population of Pseudomonas aeruginosa. Antimicrob Agents Chemother. 2013;57(6):2887–9.
325. Tari LW, Li X, Trzoss M, Bensen DC, Chen Z, Lam T, et al. Tricyclic GyrB/ParE (TriBE)
inhibitors: a new class of broad-spectrum dual-targeting antibacterial agents. PLoS One.
2013;8(12):e84409.
326. Basarab GS, Brassil P, Doig P, Galullo V, Haimes HB, Kern G, et al. Novel DNA gyrase
inhibiting spiropyrimidinetriones with a benzisoxazole scaffold: SAR and in vivo character-
ization. J Med Chem. 2014;57(21):9078–95.
327. Basarab GS, Kern GH, McNulty J, Mueller JP, Lawrence K, Vishwanathan K, et al. Responding
to the challenge of untreatable gonorrhea: ETX0914, a first-in-class agent with a distinct
mechanism-of-action against bacterial Type II topoisomerases. Sci Rep. 2015;5:11827.
328. Foerster S, Golparian D, Jacobsson S, Hathaway LJ, Low N, Shafer WM, et al. Genetic
resistance determinants, In vitro time-kill curve analysis and pharmacodynamic functions for
the novel topoisomerase II inhibitor ETX0914 (AZD0914) in Neisseria gonorrhoeae. Front
Microbiol. 2015;6:1377.
329. Walker SS, Labroli M, Painter RE, Wiltsie J, Sherborne B, Murgolo N, et al. Antibacterial
small molecules targeting the conserved TOPRIM domain of DNA gyrase. PLoS One.
2017;12(7):e0180965.
4 Resistance of Gram-negative Bacilli to Antimicrobials 149

330. Hierowski M. Inhibition of protein synthesis by Chlortetracycline in the E. Coli in vitro sys-
tem. Proc Natl Acad Sci U S A. 1965;53:594–9.
331. Brodersen DE, Clemons WM Jr, Carter AP, Morgan-Warren RJ, Wimberly BT, Ramakrishnan
V. The structural basis for the action of the antibiotics tetracycline, pactamycin, and hygro-
mycin B on the 30S ribosomal subunit. Cell. 2000;103(7):1143–54.
332. Pioletti M, Schlunzen F, Harms J, Zarivach R, Gluhmann M, Avila H, et al. Crystal struc-
tures of complexes of the small ribosomal subunit with tetracycline, edeine and IF3. EMBO
J. 2001;20(8):1829–39.
333. Duggar BM. Aureomycin; a product of the continuing search for new antibiotics. Ann N Y
Acad Sci. 1948;51(Art. 2):177–81.
334. Sum PE, Lee VJ, Testa RT, Hlavka JJ, Ellestad GA, Bloom JD, et al. Glycylcyclines. 1. A
new generation of potent antibacterial agents through modification of 9-aminotetracyclines.
J Med Chem. 1994;37(1):184–8.
335. Solomkin J, Evans D, Slepavicius A, Lee P, Marsh A, Tsai L, et al. Assessing the efficacy and
safety of Eravacycline vs Ertapenem in complicated intra-abdominal infections in the inves-
tigating Gram-Negative infections treated with Eravacycline (IGNITE 1) trial: a randomized
clinical trial. JAMA Surg. 2017;152(3):224–32.
336. Livermore DM, Mushtaq S, Warner M, Woodford N. In Vitro Activity of Eravacycline against
Carbapenem-Resistant Enterobacteriaceae and Acinetobacter baumannii. Antimicrob Agents
Chemother. 2016;60(6):3840–4.
337. Schnappinger D, Hillen W. Tetracyclines: antibiotic action, uptake, and resistance mecha-
nisms. Arch Microbiol. 1996;165(6):359–69.
338. Chopra I, Roberts M. Tetracycline antibiotics: mode of action, applications, molecular biol-
ogy, and epidemiology of bacterial resistance. Microbiol Mol Biol Rev. 2001;65(2):232–60.
339. Thaker M, Spanogiannopoulos P, Wright GD. The tetracycline resistome. Cell Mol Life Sci.
2010;67(3):419–31.
340. Guillaume G, Ledent V, Moens W, Collard JM. Phylogeny of efflux-mediated tetracycline
resistance genes and related proteins revisited. Microb Drug Resist. 2004;10(1):11–26.
341. Levy SB. Active efflux mechanisms for antimicrobial resistance. Antimicrob Agents
Chemother. 1992;36(4):695–703.
342. McMurry L, Petrucci RE Jr, Levy SB. Active efflux of tetracycline encoded by four geneti-
cally different tetracycline resistance determinants in Escherichia coli. Proc Natl Acad Sci U
S A. 1980;77(7):3974–7.
343. Ohnuki T, Katoh T, Imanaka T, Aiba S. Molecular cloning of tetracycline resistance genes
from Streptomyces rimosus in Streptomyces griseus and characterization of the cloned genes.
J Bacteriol. 1985;161(3):1010–6.
344. Hillen W, Berens C. Mechanisms underlying expression of Tn10 encoded tetracycline resis-
tance. Annu Rev Microbiol. 1994;48:345–69.
345. Saenger W, Orth P, Kisker C, Hillen W, Hinrichs W. The tetracycline repressor-a paradigm for
a biological switch. Angew Chem Int Ed Engl. 2000;39(12):2042–52.
346. Hinrichs W, Kisker C, Duvel M, Muller A, Tovar K, Hillen W, et al. Structure of the
Tet repressor-tetracycline complex and regulation of antibiotic resistance. Science.
1994;264(5157):418–20.
347. Kisker C, Hinrichs W, Tovar K, Hillen W, Saenger W. The complex formed between Tet
repressor and tetracycline-Mg2+ reveals mechanism of antibiotic resistance. J Mol Biol.
1995;247(2):260–80.
348. Orth P, Schnappinger D, Hillen W, Saenger W, Hinrichs W. Structural basis of gene regulation
by the tetracycline inducible Tet repressor-operator system. Nat Struct Biol. 2000;7(3):215–9.
349. Petersen PJ, Jacobus NV, Weiss WJ, Sum PE, Testa RT. In vitro and in vivo antibacterial
activities of a novel glycylcycline, the 9-t-butylglycylamido derivative of minocycline (GAR-­
936). Antimicrob Agents Chemother. 1999;43(4):738–44.
350. Tally FT, Ellestad GA, Testa RT. Glycylcyclines: a new generation of tetracyclines.
J Antimicrob Chemother. 1995;35(4):449–52.
150 C. R. Dean et al.

351. Testa RT, Petersen PJ, Jacobus NV, Sum PE, Lee VJ, Tally FP. In vitro and in vivo antibacte-
rial activities of the glycylcyclines, a new class of semisynthetic tetracyclines. Antimicrob
Agents Chemother. 1993;37(11):2270–7.
352. Bradford PA, Weaver-Sands DT, Petersen PJ. In vitro activity of tigecycline against iso-
lates from patients enrolled in phase 3 clinical trials of treatment for complicated skin
and skin-structure infections and complicated intra-abdominal infections. Clin Infect Dis.
2005;41(Suppl 5):S315–32.
353. Fluit AC, Florijn A, Verhoef J, Milatovic D. Presence of tetracycline resistance determi-
nants and susceptibility to tigecycline and minocycline. Antimicrob Agents Chemother.
2005;49(4):1636–8.
354. Petersen PJ, Bradford PA. Effect of medium age and supplementation with the biocatalytic
oxygen-reducing reagent oxyrase on in vitro activities of tigecycline against recent clinical
isolates. Antimicrob Agents Chemother. 2005;49(9):3910–8.
355. Fritsche TR, Sader HS, Stilwell MG, Dowzicky MJ, Jones RN. Potency and spectrum of tige-
cycline tested against an international collection of bacterial pathogens associated with skin
and soft tissue infections (2000–2004). Diagn Microbiol Infect Dis. 2005;52(3):195–201.
356. Fritsche TR, Sader HS, Stilwell MG, Dowzicky MJ, Jones RN. Antimicrobial activity of
tigecycline tested against organisms causing community-acquired respiratory tract infection
and nosocomial pneumonia. Diagn Microbiol Infect Dis. 2005;52(3):187–93.
357. Fritsche TR, Strabala PA, Sader HS, Dowzicky MJ, Jones RN. Activity of tigecycline tested
against a global collection of Enterobacteriaceae, including tetracycline-resistant isolates.
Diagn Microbiol Infect Dis. 2005;52(3):209–13.
358. Sader HS, Jones RN, Dowzicky MJ, Fritsche TR. Antimicrobial activity of tigecycline tested
against nosocomial bacterial pathogens from patients hospitalized in the intensive care unit.
Diagn Microbiol Infect Dis. 2005;52(3):203–8.
359. Sader HS, Jones RN, Stilwell MG, Dowzicky MJ, Fritsche TR. Tigecycline activity tested
against 26,474 bloodstream infection isolates: a collection from 6 continents. Diagn Microbiol
Infect Dis. 2005;52(3):181–6.
360. Grossman TH, Starosta AL, Fyfe C, O’Brien W, Rothstein DM, Mikolajka A, et al. Target-
and resistance-based mechanistic studies with TP-434, a novel fluorocycline antibiotic.
Antimicrob Agents Chemother. 2012;56(5):2559–64.
361. Guay GG, Tuckman M, Rothstein DM. Mutations in the tetA(B) gene that cause a change
in substrate specificity of the tetracycline efflux pump. Antimicrob Agents Chemother.
1994;38(4):857–60.
362. Castanheira M, Mendes RE, Jones RN. Update on Acinetobacter species: mechanisms of
antimicrobial resistance and contemporary in vitro activity of minocycline and other treat-
ment options. Clin Infect Dis. 2014;59(Suppl 6):S367–73.
363. Stein GE, Craig WA. Tigecycline: a critical analysis. Clin Infect Dis. 2006;43(4):518–24.
364. Ruzin A, Keeney D, Bradford PA. AcrAB efflux pump plays a role in decreased susceptibility
to tigecycline in Morganella morganii. Antimicrob Agents Chemother. 2005;49(2):791–3.
365. Hentschke M, Wolters M, Sobottka I, Rohde H, Aepfelbacher M. ramR mutations in clini-
cal isolates of Klebsiella pneumoniae with reduced susceptibility to tigecycline. Antimicrob
Agents Chemother. 2010;54(6):2720–3.
366. Veleba M, Schneiders T. Tigecycline resistance can occur independently of the ramA gene in
Klebsiella pneumoniae. Antimicrob Agents Chemother. 2012;56(8):4466–7.
367. Keeney D, Ruzin A, Bradford PA. RamA, a transcriptional regulator, and AcrAB, an RND-­
type efflux pump, are associated with decreased susceptibility to tigecycline in Enterobacter
cloacae. Microb Drug Resist. 2007;13(1):1–6.
368. Veleba M, De Majumdar S, Hornsey M, Woodford N, Schneiders T. Genetic characterization
of tigecycline resistance in clinical isolates of Enterobacter cloacae and Enterobacter aero-
genes. J Antimicrob Chemother. 2013;68(5):1011–8.
369. Hentschke M, Christner M, Sobottka I, Aepfelbacher M, Rohde H. Combined ramR mutation
and presence of a Tn1721-associated tet(A) variant in a clinical isolate of Salmonella enterica
serovar Hadar resistant to tigecycline. Antimicrob Agents Chemother. 2010;54(3):1319–22.
4 Resistance of Gram-negative Bacilli to Antimicrobials 151

370. Ruzin A, Keeney D, Bradford PA. AdeABC multidrug efflux pump is associated with
decreased susceptibility to tigecycline in Acinetobacter calcoaceticus-Acinetobacter bau-
mannii complex. J Antimicrob Chemother. 2007;59(5):1001–4.
371. Bratu S, Landman D, Martin DA, Georgescu C, Quale J. Correlation of antimicrobial
resistance with β-lactamases, the OmpA-like porin, and efflux pumps in clinical isolates
of Acinetobacter baumannii endemic to New York City. Antimicrob Agents Chemother.
2008;52(9):2999–3005.
372. Coyne S, Guigon G, Courvalin P, Perichon B. Screening and quantification of the expres-
sion of antibiotic resistance genes in Acinetobacter baumannii with a microarray. Antimicrob
Agents Chemother. 2010;54(1):333–40.
373. Coyne S, Rosenfeld N, Lambert T, Courvalin P, Perichon B. Overexpression of resistance-­
nodulation-­cell division pump AdeFGH confers multidrug resistance in Acinetobacter bau-
mannii. Antimicrob Agents Chemother. 2010;54(10):4389–93.
374. Rumbo C, Gato E, Lopez M, Ruiz de Alegria C, Fernandez-Cuenca F, Martinez-Martinez L,
et al. Contribution of efflux pumps, porins, and β-lactamases to multidrug resistance in clinical
isolates of Acinetobacter baumannii. Antimicrob Agents Chemother. 2013;57(11):5247–57.
375. Connell SR, Tracz DM, Nierhaus KH, Taylor DE. Ribosomal protection proteins and their
mechanism of tetracycline resistance. Antimicrob Agents Chemother. 2003;47(12):3675–81.
376. Taylor DE. Plasmid-mediated tetracycline resistance in Campylobacter jejuni: expression
in Escherichia coli and identification of homology with streptococcal class M determinant.
J Bacteriol. 1986;165(3):1037–9.
377. Doyle D, McDowall KJ, Butler MJ, Hunter IS. Characterization of an oxytetracycline-­
resistance gene, otrA, of Streptomyces rimosus. Mol Microbiol. 1991;5(12):2923–33.
378. Sanchez-Pescador R, Brown JT, Roberts M, Urdea MS. Homology of the TetM with trans-
lational elongation factors: implications for potential modes of tetM-conferred tetracycline
resistance. Nucleic Acids Res. 1988;16(3):1218.
379. Burdett V. Tet(M)-promoted release of tetracycline from ribosomes is GTP dependent.
J Bacteriol. 1996;178(11):3246–51.
380. Spahn CM, Blaha G, Agrawal RK, Penczek P, Grassucci RA, Trieber CA, et al. Localization
of the ribosomal protection protein Tet(O) on the ribosome and the mechanism of tetracycline
resistance. Mol Cell. 2001;7(5):1037–45.
381. Connell SR, Trieber CA, Dinos GP, Einfeldt E, Taylor DE, Nierhaus KH. Mechanism of
Tet(O)-mediated tetracycline resistance. EMBO J. 2003;22(4):945–53.
382. Connell SR, Trieber CA, Stelzl U, Einfeldt E, Taylor DE, Nierhaus KH. The tetracycline
resistance protein Tet(o) perturbs the conformation of the ribosomal decoding centre. Mol
Microbiol. 2002;45(6):1463–72.
383. Donhofer A, Franckenberg S, Wickles S, Berninghausen O, Beckmann R, Wilson
DN. Structural basis for TetM-mediated tetracycline resistance. Proc Natl Acad Sci U S A.
2012;109(42):16900–5.
384. Li W, Atkinson GC, Thakor NS, Allas U, Lu CC, Chan KY, et al. Mechanism of tetracycline
resistance by ribosomal protection protein Tet(O). Nat Commun. 2013;4:1477.
385. Draper MP, Weir S, Macone A, Donatelli J, Trieber CA, Tanaka SK, et al. Mechanism
of action of the novel aminomethylcycline antibiotic omadacycline. Antimicrob Agents
Chemother. 2014;58(3):1279–83.
386. Macone AB, Caruso BK, Leahy RG, Donatelli J, Weir S, Draper MP, et al. In vitro and in vivo
antibacterial activities of omadacycline, a novel aminomethylcycline. Antimicrob Agents
Chemother. 2014;58(2):1127–35.
387. Ross JI, Eady EA, Cove JH, Cunliffe WJ. 16S rRNA mutation associated with tetracycline
resistance in a gram-positive bacterium. Antimicrob Agents Chemother. 1998;42(7):1702–5.
388. Gerrits MM, de Zoete MR, Arents NL, Kuipers EJ, Kusters JG. 16S rRNA mutation-­
mediated tetracycline resistance in Helicobacter pylori. Antimicrob Agents Chemother.
2002;46(9):2996–3000.
389. Trieber CA, Taylor DE. Mutations in the 16S rRNA genes of Helicobacter pylori mediate
resistance to tetracycline. J Bacteriol. 2002;184(8):2131–40.
152 C. R. Dean et al.

390. Hu M, Nandi S, Davies C, Nicholas RA. High-level chromosomally mediated tetracycline

resistance in Neisseria gonorrhoeae results from a point mutation in the rpsJ gene encod-
ing ribosomal protein S10 in combination with the mtrR and penB resistance determinants.
Antimicrob Agents Chemother. 2005;49(10):4327–34.
391. Villa L, Feudi C, Fortini D, Garcia-Fernandez A, Carattoli A. Genomics of KPC-producing
Klebsiella pneumoniae sequence type 512 clone highlights the role of RamR and ribosomal
S10 protein mutations in conferring tigecycline resistance. Antimicrob Agents Chemother.
2014;58(3):1707–12.
392. Snitkin ES, Zelazny AM, Thomas PJ, Stock F, Group NCSP, Henderson DK, et al. Tracking
a hospital outbreak of carbapenem-resistant Klebsiella pneumoniae with whole-genome
sequencing. Sci Transl Med. 2012;4(148):148ra16.
393. Speer BS, Bedzyk L, Salyers AA. Evidence that a novel tetracycline resistance gene found
on two Bacteroides transposons encodes an NADP-requiring oxidoreductase. J Bacteriol.
1991;173(1):176–83.
394. Yang W, Moore IF, Koteva KP, Bareich DC, Hughes DW, Wright GD. TetX is a flavin-­
dependent monooxygenase conferring resistance to tetracycline antibiotics. J Biol Chem.
2004;279(50):52346–52.
395. Volkers G, Damas JM, Palm GJ, Panjikar S, Soares CM, Hinrichs W. Putative dioxygen-­
binding sites and recognition of tigecycline and minocycline in the tetracycline-degrading
monooxygenase TetX. Acta Crystallogr D Biol Crystallogr. 2013;69(Pt 9):1758–67.
396. Volkers G, Palm GJ, Weiss MS, Wright GD, Hinrichs W. Structural basis for a new tetracycline
resistance mechanism relying on the TetX monooxygenase. FEBS Lett. 2011;585(7):1061–6.
397. Bergeron J, Ammirati M, Danley D, James L, Norcia M, Retsema J, et al. Glycylcyclines
bind to the high-affinity tetracycline ribosomal binding site and evade Tet(M)- and Tet(O)-
mediated ribosomal protection. Antimicrob Agents Chemother. 1996;40(9):2226–8.
398. Jenner L, Starosta AL, Terry DS, Mikolajka A, Filonava L, Yusupov M, et al. Structural basis
for potent inhibitory activity of the antibiotic tigecycline during protein synthesis. Proc Natl
Acad Sci U S A. 2013;110(10):3812–6.
399. Olson MW, Ruzin A, Feyfant E, Rush TS, O’Connell J, Bradford PA. Functional, biophysical,
and structural bases for antibacterial activity of tigecycline. Antimicrob Agents Chemother.
2006;50(6):2156–66.
400. Hirata T, Saito A, Nishino K, Tamura N, Yamaguchi A. Effects of efflux transporter genes on
susceptibility of Escherichia coli to tigecycline (GAR-936). Antimicrob Agents Chemother.
2004;48(6):2179–84.
401. Orth P, Schnappinger D, Sum PE, Ellestad GA, Hillen W, Saenger W, et al. Crystal struc-
ture of the tet repressor in complex with a novel tetracycline, 9-(N,N-dimethylglycylamido)-
6-demethyl-6-deoxy-tetracycline. J Mol Biol. 1999;285(2):455–61.
402. Sato T, Suzuki Y, Shiraishi T, Honda H, Shinagawa M, Yamamoto S, et al. Tigecycline nonsus-
ceptibility occurs exclusively in fluoroquinolone-resistant Escherichia coli clinical isolates,
Including the major multidrug-resistant lineages O25b:H4-ST131-H30R and O1-ST648.
Antimicrob Agents Chemother. 2017;61(2):e01654.
403. Sheng ZK, Hu F, Wang W, Guo Q, Chen Z, Xu X, et al. Mechanisms of tigecycline resis-
tance among Klebsiella pneumoniae clinical isolates. Antimicrob Agents Chemother.
2014;58(11):6982–5.
404. Trebosc V, Gartenmann S, Royet K, Manfredi P, Totzl M, Schellhorn B, et al. A novel genome-­
editing platform for drug-resistant Acinetobacter baumannii reveals an AdeR-unrelated tige-
cycline resistance mechanism. Antimicrob Agents Chemother. 2016;60(12):7263–71.
405. Linkevicius M, Sandegren L, Andersson DI. Potential of tetracycline resistance proteins to
evolve tigecycline resistance. Antimicrob Agents Chemother. 2016;60(2):789–96.
406. Grossman TH, Fyfe C, O’Brien W, Hackel M, Minyard MB, Waites KB, et al. Fluorocycline
TP-271 is potent against complicated community-acquired bacterial pneumonia pathogens.
mSphere. 2017;2(1):e00004.
4 Resistance of Gram-negative Bacilli to Antimicrobials 153

407. Boyer A, Gruson D, Bouchet S, Clouzeau B, Hoang-Nam B, Vargas F, et al. Aminoglycosides

in septic shock: an overview, with specific consideration given to their nephrotoxic risk. Drug
Saf. 2013;36(4):217–30.
408. Adler M, Anjum M, Andersson DI, Sandegren L. Combinations of mutations in envZ,
ftsI, mrdA, acrB and acrR can cause high-level carbapenem resistance in Escherichia coli.
J Antimicrob Chemother. 2016;71(5):1188–98.
409. Krause KM, Serio AW, Kane TR, Connolly LE. Aminoglycosides: an overview. Cold Spring
Harb Perspect Med. 2016;6(6):a027029.
410. Cunha BA. Aminoglycosides: current role in antimicrobial therapy. Pharmacotherapy.
1988;8(6):334–50.
411. Ogle JM, Brodersen DE, Clemons WM Jr, Tarry MJ, Carter AP, Ramakrishnan V. Recognition
of cognate transfer RNA by the 30S ribosomal subunit. Science. 2001;292(5518):897–902.
412. Moazed D, Noller HF. Interaction of antibiotics with functional sites in 16S ribosomal
RNA. Nature. 1987;327(6121):389–94.
413. Purohit P, Stern S. Interactions of a small RNA with antibiotic and RNA ligands of the 30S
subunit. Nature. 1994;370(6491):659–62.
414. Woodcock J, Moazed D, Cannon M, Davies J, Noller HF. Interaction of antibiotics with A-
and P-site-specific bases in 16S ribosomal RNA. EMBO J. 1991;10(10):3099–103.
415. Davies J, Davis BD. Misreading of ribonucleic acid code words induced by aminoglycoside
antibiotics. The effect of drug concentration. J Biol Chem. 1968;243(12):3312–6.
416. Gottfredsson M, Erlendsdottir H, Gudmundsson A, Gudmundsson S. Different patterns
of bacterial DNA synthesis during postantibiotic effect. Antimicrob Agents Chemother.
1995;39(6):1314–9.
417. Anderson NH, Bowman J, Erwin A, Harwood E, Kline T, Mdluli K, Ng S, Pfister KB, Shawar
R, Wagman A, Yabannavar A. Antibacterial agents. WO/2004/062601. July 29, 2004.
418. Stubbings W, Bostock J, Ingham E, Chopra I. Mechanisms of the post-antibiotic effects
induced by rifampicin and gentamicin in Escherichia coli. J Antimicrob Chemother.
2006;58(2):444–8.
419. Davis BD. Mechanism of bactericidal action of aminoglycosides. Microbiol Rev.
1987;51(3):341–50.
420. Kadar B, Kocsis B, Toth A, Damjanova I, Szasz M, Kristof K, et al. Synergistic antibiotic
combinations for colistin-resistant Klebsiella pneumoniae. Acta Microbiol Immunol Hung.
2013;60(2):201–9.
421. Clock SA, Tabibi S, Alba L, Kubin CJ, Whittier S, Saiman L. In vitro activity of doripenem
alone and in multi-agent combinations against extensively drug-resistant Acinetobacter bau-
mannii and Klebsiella pneumoniae. Diagn Microbiol Infect Dis. 2013;76(3):343–6.
422. Raja NS, Singh NN. Antimicrobial susceptibility pattern of clinical isolates of Pseudomonas
aeruginosa in a tertiary care hospital. J Microbiol Immunol Infect. 2007;40(1):45–9.
423. Ullmann U. In vitro experiments on the working of combinations of gentamicin and β-lactam
antibiotics against Pseudonomas aeruginosa (author’s transl). Immun Infekt. 1975;3(2):79–85.
424. Chaudhary M, Shrivastava SM, Varughese L, Sehgal R. Efficacy and safety evaluation of
fixed dose combination of cefepime and amikacin in comparison with cefepime alone in
treatment of nosocomial pneumonia patients. Curr Clin Pharmacol. 2008;3(2):118–22.
425. Hopefl AW. Overview of synergy with reference to double β-lactam combinations. DICP.
1991;25(9):972–7.
426. Livermore DM. Multiple mechanisms of antimicrobial resistance in Pseudomonas aerugi-
nosa: our worst nightmare? Clin Infect Dis. 2002;34(5):634–40.
427. Zembower TR, Noskin GA, Postelnick MJ, Nguyen C, Peterson LR. The utility of aminogly-
cosides in an era of emerging drug resistance. Int J Antimicrob Agents. 1998;10(2):95–105.
428. Tzouvelekis LS, Markogiannakis A, Piperaki E, Souli M, Daikos GL. Treating infec-
tions caused by carbapenemase-producing Enterobacteriaceae. Clin Microbiol Infect.
2014;20(9):862–72.
154 C. R. Dean et al.

429. Haidar G, Alkroud A, Cheng S, Churilla TM, Churilla BM, Shields RK, et al. Association
between the presence of aminoglycoside-modifying enzymes and in vitro activity of gentami-
cin, tobramycin, amikacin, and plazomicin against Klebsiella pneumoniae carbapenemase-
and extended-spectrum-β-lactamase-producing Enterobacter species. Antimicrob Agents
Chemother. 2016;60(9):5208–14.
430. Gonzalez-Padilla M, Torre-Cisneros J, Rivera-Espinar F, Pontes-Moreno A, Lopez-Cerero
L, Pascual A, et al. Gentamicin therapy for sepsis due to carbapenem-resistant and colistin-­
resistant Klebsiella pneumoniae. J Antimicrob Chemother. 2015;70(3):905–13.
431. Bercot B, Poirel L, Nordmann P. Updated multiplex polymerase chain reaction for detection
of 16S rRNA methylases: high prevalence among NDM-1 producers. Diagn Microbiol Infect
Dis. 2011;71(4):442–5.
432. Miro E, Grunbaum F, Gomez L, Rivera A, Mirelis B, Coll P, et al. Characterization of
aminoglycoside-­modifying enzymes in Enterobacteriaceae clinical strains and characteriza-
tion of the plasmids implicated in their diffusion. Microb Drug Resist. 2013;19(2):94–9.
433. Petrosillo N, Vranic-Ladavac M, Feudi C, Villa L, Fortini D, Barisic N, et al. Spread of
Enterobacter cloacae carrying blaNDM-1, blaCTX-M-15, blaSHV-12 and plasmid-mediated
quinolone resistance genes in a surgical intensive care unit in Croatia. J Glob Antimicrob
Resist. 2016;4:44–8.
434. Poirel L, Savov E, Nazli A, Trifonova A, Todorova I, Gergova I, et al. Outbreak caused by
NDM-1- and RmtB-producing Escherichia coli in Bulgaria. Antimicrob Agents Chemother.
2014;58(4):2472–4.
435. Poirel L, Yilmaz M, Istanbullu A, Arslan F, Mert A, Bernabeu S, et al. Spread of NDM-1-­
producing Enterobacteriaceae in a neonatal intensive care unit in Istanbul, Turkey. Antimicrob
Agents Chemother. 2014;58(5):2929–33.
436. Rahman M, Prasad KN, Pathak A, Pati BK, Singh A, Ovejero CM, et al. RmtC and RmtF 16S
rRNA methyltransferase in NDM-1-producing Pseudomonas aeruginosa. Emerg Infect Dis.
2015;21(11):2059–62.
437. Bradford PA, Dean CR. Resistance of Gram-Negative Bacilli to antimicrobials. In: Fong
IW, Drlica K, editors. Antimicrobial aesistance and implications for the twenty-first century.
Boston, MA: Springer US; 2008. p. 97–159.
438. Honore N, Marchal G, Cole ST. Novel mutation in 16S rRNA associated with strep-
tomycin dependence in Mycobacterium tuberculosis. Antimicrob Agents Chemother.
1995;39(3):769–70.
439. Nessar R, Reyrat JM, Murray A, Gicquel B. Genetic analysis of new 16S rRNA mutations
conferring aminoglycoside resistance in Mycobacterium abscessus. J Antimicrob Chemother.
2011;66(8):1719–24.
440. Criswell D, Tobiason VL, Lodmell JS, Samuels DS. Mutations conferring aminoglyco-
side and spectinomycin resistance in Borrelia burgdorferi. Antimicrob Agents Chemother.
2006;50(2):445–52.
441. Vakulenko SB, Mobashery S. Versatility of aminoglycosides and prospects for their future.
Clin Microbiol Rev. 2003;16(3):430–50.
442. Ramirez MS, Tolmasky ME. Aminoglycoside modifying enzymes. Drug Resist Updat.
2010;13(6):151–71.
443. Wright GD. Aminoglycoside-modifying enzymes. Curr Opin Microbiol. 1999;2(5):499–503.
444. Kosmidis C, Giannopoulou M, Flountzi A, Markogiannakis A, Goukos D, Petrikkos G, et al.
Genetic basis of aminoglycoside resistance following changes in aminoglycoside prescrip-
tion patterns. J Chemother. 2013;25(4):217–21.
445. Rather PN, Munayyer H, Mann PA, Hare RS, Miller GH, Shaw KJ. Genetic analysis of bacte-
rial acetyltransferases: identification of amino acids determining the specificities of the ami-
noglycoside 6’-N-acetyltransferase Ib and IIa proteins. J Bacteriol. 1992;174(10):3196–203.
446. Wohlleben W, Arnold W, Bissonnette L, Pelletier A, Tanguay A, Roy PH, et al. On the evo-
lution of Tn21-like multiresistance transposons: sequence analysis of the gene (aacC1) for
4 Resistance of Gram-negative Bacilli to Antimicrobials 155

gentamicin acetyltransferase-3-I(AAC(3)-I), another member of the Tn21-based expression

cassette. Mol Gen Genet. 1989;217(2-3):202–8.
447. Vetting MW, Park CH, Hegde SS, Jacoby GA, Hooper DC, Blanchard JS. Mechanistic and
structural analysis of aminoglycoside N-acetyltransferase AAC(6’)-Ib and its bifunctional,
fluoroquinolone-active AAC(6’)-Ib-cr variant. Biochemistry. 2008;47(37):9825–35.
448. Maurice F, Broutin I, Podglajen I, Benas P, Collatz E, Dardel F. Enzyme structural plasticity
and the emergence of broad-spectrum antibiotic resistance. EMBO Rep. 2008;9(4):344–9.
449. Grindley ND, Joyce CM. Analysis of the structure and function of the kanamycin-resistance
transposon Tn903. Cold Spring Harb Symp Quant Biol. 1981;45(Pt 1):125–33.
450. Stogios PJ, Spanogiannopoulos P, Evdokimova E, Egorova O, Shakya T, Todorovic N, et al.
Structure-guided optimization of protein kinase inhibitors reverses aminoglycoside antibiotic
resistance. Biochem J. 2013;454(2):191–200.
451. McKay GA, Thompson PR, Wright GD. Broad spectrum aminoglycoside phosphotrans-
ferase type III from Enterococcus: overexpression, purification, and substrate specificity.
Biochemistry. 1994;33(22):6936–44.
452. Wiedemann B. Mechanisms of antibiotic resistance and their dissemination of resistance
genes in the hospital environment. Infect Control. 1983;4(6):444–7.
453. Jacoby GA, Blaser MJ, Santanam P, Hachler H, Kayser FH, Hare RS, et al. Appearance
of amikacin and tobramycin resistance due to 4’-aminoglycoside nucleotidyltrans-
ferase [ANT(4’)-II] in gram-negative pathogens. Antimicrob Agents Chemother.
1990;34(12):2381–6.
454. Coggins BE, McClerren AL, Jiang L, Li X, Rudolph J, Hindsgaul O, et al. Refined solution
structure of the LpxC-TU-514 complex and pKa analysis of an active site histidine: insights
into the mechanism and inhibitor design. Biochemistry. 2005;44(4):1114–26.
455. Azucena E, Mobashery S. Aminoglycoside-modifying enzymes: mechanisms of catalytic
processes and inhibition. Drug Resist Updat. 2001;4(2):106–17.
456. Centron D, Roy PH. Presence of a group II intron in a multiresistant Serratia marcescens
strain that harbors three integrons and a novel gene fusion. Antimicrob Agents Chemother.
2002;46(5):1402–9.
457. Green KD, Garneau-Tsodikova S. Domain dissection and characterization of the aminogly-
coside resistance enzyme ANT(3”)-Ii/AAC(6’)-IId from Serratia marcescens. Biochimie.
2013;95(6):1319–25.
458. Thompson J, Skeggs PA, Cundliffe E. Methylation of 16S ribosomal RNA and resistance
to the aminoglycoside antibiotics gentamicin and kanamycin determined by DNA from the
gentamicin-producer, Micromonospora purpurea. Mol Gen Genet. 1985;201(2):168–73.
459. Beauclerk AA, Cundliffe E. Sites of action of two ribosomal RNA methylases responsible for
resistance to aminoglycosides. J Mol Biol. 1987;193(4):661–71.
460. Yamane K, Wachino J, Doi Y, Kurokawa H, Arakawa Y. Global spread of multiple aminogly-
coside resistance genes. Emerg Infect Dis. 2005;11(6):951–3.
461. Galimand M, Courvalin P, Lambert T. Plasmid-mediated high-level resistance to aminogly-
cosides in Enterobacteriaceae due to 16S rRNA methylation. Antimicrob Agents Chemother.
2003;47(8):2565–71.
462. Yokoyama K, Doi Y, Yamane K, Kurokawa H, Shibata N, Shibayama K, et al. Acquisition of
16S rRNA methylase gene in Pseudomonas aeruginosa. Lancet. 2003;362(9399):1888–93.
463. Wachino J, Shibayama K, Kurokawa H, Kimura K, Yamane K, Suzuki S, et al. Novel
plasmid-­mediated 16S rRNA m1A1408 methyltransferase, NpmA, found in a clinically iso-
lated Escherichia coli strain resistant to structurally diverse aminoglycosides. Antimicrob
Agents Chemother. 2007;51(12):4401–9.
464. Macfarlane EL, Kwasnicka A, Hancock RE. Role of Pseudomonas aeruginosa PhoP-­
phoQ in resistance to antimicrobial cationic peptides and aminoglycosides. Microbiology.
2000;146(Pt 10):2543–54.
465. Poole K. Aminoglycoside resistance in Pseudomonas aeruginosa. Antimicrob Agents
Chemother. 2005;49(2):479–87.
156 C. R. Dean et al.

466. Aires JR, Nikaido H. Aminoglycosides are captured from both periplasm and cytoplasm by
the AcrD multidrug efflux transporter of Escherichia coli. J Bacteriol. 2005;187(6):1923–9.
467. Rosenberg EY, Ma D, Nikaido H. AcrD of Escherichia coli is an aminoglycoside efflux
pump. J Bacteriol. 2000;182(6):1754–6.
468. Chan YY, Tan TM, Ong YM, Chua KL. BpeAB-OprB, a multidrug efflux pump in
Burkholderia pseudomallei. Antimicrob Agents Chemother. 2004;48(4):1128–35.
469. Aires JR, Kohler T, Nikaido H, Plesiat P. Involvement of an active efflux system in the natural
resistance of Pseudomonas aeruginosa to aminoglycosides. Antimicrob Agents Chemother.
1999;43(11):2624–8.
470. Westbrock-Wadman S, Sherman DR, Hickey MJ, Coulter SN, Zhu YQ, Warrener P, et al.
Characterization of a Pseudomonas aeruginosa efflux pump contributing to aminoglycoside
impermeability. Antimicrob Agents Chemother. 1999;43(12):2975–83.
471. Jeannot K, Sobel ML, El Garch F, Poole K, Plesiat P. Induction of the MexXY efflux
pump in Pseudomonas aeruginosa is dependent on drug-ribosome interaction. J Bacteriol.
2005;187(15):5341–6.
472. Karlowsky JA, Hoban DJ, Zelenitsky SA, Zhanel GG. Altered denA and anr gene expression
in aminoglycoside adaptive resistance in Pseudomonas aeruginosa. J Antimicrob Chemother.
1997;40(3):371–6.
473. Yamamoto M, Ueda A, Kudo M, Matsuo Y, Fukushima J, Nakae T, et al. Role of MexZ and
PA5471 in transcriptional regulation of mexXY in Pseudomonas aeruginosa. Microbiology.
2009;155(Pt 10):3312–21.
474. Lau CH, Krahn T, Gilmour C, Mullen E, Poole K. AmgRS-mediated envelope stress-­
inducible expression of the mexXY multidrug efflux operon of Pseudomonas aeruginosa.
Microbiologyopen. 2015;4(1):121–35.
475. Guenard S, Muller C, Monlezun L, Benas P, Broutin I, Jeannot K, et al. Multiple muta-
tions lead to MexXY-OprM-dependent aminoglycoside resistance in clinical strains of
Pseudomonas aeruginosa. Antimicrob Agents Chemother. 2014;58(1):221–8.
476. Pirnay JP, Bilocq F, Pot B, Cornelis P, Zizi M, Van Eldere J, et al. Pseudomonas aeruginosa
population structure revisited. PLoS One. 2009;4(11):e7740.
477. Taccetti G, Campana S, Neri AS, Boni V, Festini F. Antibiotic therapy against Pseudomonas
aeruginosa in cystic fibrosis. J Chemother. 2008;20(2):166–9.
478. Ciofu O, Mandsberg LF, Wang H, Hoiby N. Phenotypes selected during chronic lung infec-
tion in cystic fibrosis patients: implications for the treatment of Pseudomonas aeruginosa
biofilm infections. FEMS Immunol Med Microbiol. 2012;65(2):215–25.
479. Sadovskaya I, Vinogradov E, Li J, Hachani A, Kowalska K, Filloux A. High-level antibi-
otic resistance in Pseudomonas aeruginosa biofilm: the ndvB gene is involved in the pro-
duction of highly glycerol-phosphorylated β-(1->3)-glucans, which bind aminoglycosides.
Glycobiology. 2010;20(7):895–904.
480. Mah TF, Pitts B, Pellock B, Walker GC, Stewart PS, O’Toole GA. A genetic basis for
Pseudomonas aeruginosa biofilm antibiotic resistance. Nature. 2003;426(6964):306–10.
481. Crabbe A, Liu Y, Matthijs N, Rigole P, De La Fuente-Nunez C, Davis R, et al. Antimicrobial
efficacy against Pseudomonas aeruginosa biofilm formation in a three-dimensional lung epi-
thelial model and the influence of fetal bovine serum. Sci Rep. 2017;7:43321.
482. Nagant C, Tre-Hardy M, El-Ouaaliti M, Savage P, Devleeschouwer M, Dehaye JP. Interaction
between tobramycin and CSA-13 on clinical isolates of Pseudomonas aeruginosa in a model
of young and mature biofilms. Appl Microbiol Biotechnol. 2010;88(1):251–63.
483. Burk DL, Hon WC, Leung AK, Berghuis AM. Structural analyses of nucleotide binding to an
aminoglycoside phosphotransferase. Biochemistry. 2001;40(30):8756–64.
484. Hon WC, McKay GA, Thompson PR, Sweet RM, Yang DS, Wright GD, et al. Structure of
an enzyme required for aminoglycoside antibiotic resistance reveals homology to eukaryotic
protein kinases. Cell. 1997;89(6):887–95.
485. Sakon J, Liao HH, Kanikula AM, Benning MM, Rayment I, Holden HM. Molecular struc-
ture of kanamycin nucleotidyltransferase determined to 3.0-A resolution. Biochemistry.
1993;32(45):11977–84.
4 Resistance of Gram-negative Bacilli to Antimicrobials 157

486. Wolf E, Vassilev A, Makino Y, Sali A, Nakatani Y, Burley SK. Crystal structure of a GCN5-­
related N-acetyltransferase: Serratia marcescens aminoglycoside 3-N-acetyltransferase. Cell.
1998;94(4):439–49.
487. Wybenga-Groot LE, Draker K, Wright GD, Berghuis AM. Crystal structure of an aminogly-
coside 6’-N-acetyltransferase: defining the GCN5-related N-acetyltransferase superfamily
fold. Structure. 1999;7(5):497–507.
488. Caldwell SJ, Berghuis AM. Small-angle X-ray scattering analysis of the bifunctional anti-
biotic resistance enzyme aminoglycoside (6’) acetyltransferase-ie/aminoglycoside (2”)
phosphotransferase-ia reveals a rigid solution structure. Antimicrob Agents Chemother.
2012;56(4):1899–906.
489. Burk DL, Ghuman N, Wybenga-Groot LE, Berghuis AM. X-ray structure of the AAC(6’)-Ii
antibiotic resistance enzyme at 1.8 A resolution; examination of oligomeric arrangements in
GNAT superfamily members. Protein Sci. 2003;12(3):426–37.
490. Burk DL, Xiong B, Breitbach C, Berghuis AM. Structures of aminoglycoside acetyltransfer-
ase AAC(6’)-Ii in a novel crystal form: structural and normal-mode analyses. Acta Crystallogr
D Biol Crystallogr. 2005;61(Pt 9):1273–9.
491. Nurizzo D, Shewry SC, Perlin MH, Brown SA, Dholakia JN, Fuchs RL, et al. The crystal
structure of aminoglycoside-3’-phosphotransferase-IIa, an enzyme responsible for antibiotic
resistance. J Mol Biol. 2003;327(2):491–506.
492. Young PG, Walanj R, Lakshmi V, Byrnes LJ, Metcalf P, Baker EN, et al. The crystal struc-
tures of substrate and nucleotide complexes of Enterococcus faecium aminoglycoside-­2″-
phosphotransferase-IIa [APH(2″)-IIa] provide insights into substrate selectivity in the
APH(2″) subfamily. J Bacteriol. 2009;191(13):4133–43.
493. Labby KJ, Garneau-Tsodikova S. Strategies to overcome the action of aminoglycoside-­
modifying enzymes for treating resistant bacterial infections. Future Med Chem.
2013;5(11):1285–309.
494. Lee JH, Lee CS. Clinical usefulness of arbekacin. Infect Chemother. 2016;48(1):1–11.
495. Sader HS, Rhomberg PR, Farrell DJ, Jones RN. Arbekacin activity against contemporary
clinical bacteria isolated from patients hospitalized with pneumonia. Antimicrob Agents
Chemother. 2015;59(6):3263–70.
496. Livermore DM, Mushtaq S, Warner M, Zhang JC, Maharjan S, Doumith M, et al. Activity
of aminoglycosides, including ACHN-490, against carbapenem-resistant Enterobacteriaceae
isolates. J Antimicrob Chemother. 2011;66(1):48–53.
497. Ago K, Umemura E, Takahashi Y, Igarashi M, Hayashi C, Shibasaki M, Yamada K, Ida T,
Yonezawa M. TS3112, a novel aminoglycoside antibiotic active against multidrug-resistant
pathogens producing 16S rRNA methyltransferases: synthesis and structure-activity relation-
ships ASM microbe 2017; New Orleans; 2017 Jun 4.
498. Hamed K, Debonnett L. Tobramycin inhalation powder for the treatment of pulmonary
Pseudomonas aeruginosa infection in patients with cystic fibrosis: a review based on clinical
evidence. Ther Adv Respir Dis. 2017;11(5):193–209.
499. McKeage K. Tobramycin inhalation powder: a review of its use in the treatment of
chronic Pseudomonas aeruginosa infection in patients with cystic fibrosis. Drugs.
2013;73(16):1815–27.
500. Niederman MS, Chastre J, Corkery K, Fink JB, Luyt CE, Garcia MS. BAY41-6551 achieves
bactericidal tracheal aspirate amikacin concentrations in mechanically ventilated patients
with Gram-negative pneumonia. Intensive Care Med. 2012;38(2):263–71.
501. Luyt CE, Eldon MA, Stass H, Gribben D, Corkery K, Chastre J. Pharmacokinetics and toler-
ability of amikacin administered as BAY41-6551 aerosol in mechanically ventilated patients
with gram-negative pneumonia and acute renal failure. J Aerosol Med Pulm Drug Deliv.
2011;24(4):183–90.
502. Adams PD, Grosse-Kunstleve RW, Hung LW, Ioerger TR, McCoy AJ, Moriarty NW, et al.
PHENIX: building new software for automated crystallographic structure determination.
Acta Crystallogr D. 2002;58:1948–54.
158 C. R. Dean et al.

503. Koch-Weser J, Sidel VW, Federman EB, Kanarek P, Finer DC, Eaton AE. Adverse effects
of sodium colistimethate. Manifestations and specific reaction rates during 317 courses of
therapy. Ann Intern Med. 1970;72(6):857–68.
504. Yow EM, Moyer JH. Toxicity of polymyxin B. II. Human studies with particular reference to
evaluation of renal function. AMA Arch Intern Med. 1953;92(2):248–57.
505. Velkov T, Thompson PE, Nation RL, Li J. Structure--activity relationships of polymyxin
antibiotics. J Med Chem. 2010;53(5):1898–916.
506. Pristovsek P, Kidric J. The search for molecular determinants of LPS inhibition by proteins
and peptides. Curr Top Med Chem. 2004;4(11):1185–201.
507. Sampson TR, Liu X, Schroeder MR, Kraft CS, Burd EM, Weiss DS. Rapid killing of
Acinetobacter baumannii by polymyxins is mediated by a hydroxyl radical death pathway.
Antimicrob Agents Chemother. 2012;56(11):5642–9.
508. Deris ZZ, Akter J, Sivanesan S, Roberts KD, Thompson PE, Nation RL, et al. A second-
ary mode of action of polymyxins against Gram-negative bacteria involves the inhibition of
NADH-quinone oxidoreductase activity. J Antibiot (Tokyo). 2014;67(2):147–51.
509. Olaitan AO, Morand S, Rolain JM. Mechanisms of polymyxin resistance: acquired and
intrinsic resistance in bacteria. Front Microbiol. 2014;5:643.
510. Nowicki EM, O’Brien JP, Brodbelt JS, Trent MS. Extracellular zinc induces phosphoethanol-
amine addition to Pseudomonas aeruginosa lipid A via the ColRS two-component system.
Mol Microbiol. 2015;97(1):166–78.
511. Chin CY, Gregg KA, Napier BA, Ernst RK, Weiss DS. A PmrB-regulated deacetylase
required for lipid A modification and polymyxin resistance in Acinetobacter baumannii.
Antimicrob Agents Chemother. 2015;59(12):7911–4.
512. Rubin EJ, Herrera CM, Crofts AA, Trent MS. PmrD is required for modifications to esch-
erichia coli endotoxin that promote antimicrobial resistance. Antimicrob Agents Chemother.
2015;59(4):2051–61.
513. Prost LR, Daley ME, Le Sage V, Bader MW, Le Moual H, Klevit RE, et al. Activation of the
bacterial sensor kinase PhoQ by acidic pH. Mol Cell. 2007;26(2):165–74.
514. Bader MW, Sanowar S, Daley ME, Schneider AR, Cho U, Xu W, et al. Recognition of anti-
microbial peptides by a bacterial sensor kinase. Cell. 2005;122(3):461–72.
515. Gunn JS, Richards SM. Recognition and integration of multiple environmental signals by the
bacterial sensor kinase PhoQ. Cell Host Microbe. 2007;1(3):163–5.
516. Schurek KN, Sampaio JL, Kiffer CR, Sinto S, Mendes CM, Hancock RE. Involvement
of pmrAB and phoPQ in polymyxin B adaptation and inducible resistance in non-cystic
fibrosis clinical isolates of Pseudomonas aeruginosa. Antimicrob Agents Chemother.
2009;53(10):4345–51.
517. Pragasam AK, Shankar C, Veeraraghavan B, Biswas I, Nabarro LE, Inbanathan FY, et al.
Molecular mechanisms of colistin resistance in Klebsiella pneumoniae causing bacteremia
from India-a first report. Front Microbiol. 2016;7:2135.
518. Cannatelli A, Di Pilato V, Giani T, Arena F, Ambretti S, Gaibani P, et al. In vivo evolution
to colistin resistance by PmrB sensor kinase mutation in KPC-producing Klebsiella pneu-
moniae is associated with low-dosage colistin treatment. Antimicrob Agents Chemother.
2014;58(8):4399–403.
519. Cheng HY, Chen YF, Peng HL. Molecular characterization of the PhoPQ-PmrD-PmrAB
mediated pathway regulating polymyxin B resistance in Klebsiella pneumoniae CG43.
J Biomed Sci. 2010;17:60.
520. Jayol A, Poirel L, Brink A, Villegas MV, Yilmaz M, Nordmann P. Resistance to colistin asso-
ciated with a single amino acid change in protein PmrB among Klebsiella pneumoniae iso-
lates of worldwide origin. Antimicrob Agents Chemother. 2014;58(8):4762–6.
521. Moskowitz SM, Brannon MK, Dasgupta N, Pier M, Sgambati N, Miller AK, et al. PmrB
mutations promote polymyxin resistance of Pseudomonas aeruginosa isolated from colistin-­
treated cystic fibrosis patients. Antimicrob Agents Chemother. 2012;56(2):1019–30.
4 Resistance of Gram-negative Bacilli to Antimicrobials 159

522. Barrow K, Kwon DH. Alterations in two-component regulatory systems of phoPQ and
pmrAB are associated with polymyxin B resistance in clinical isolates of Pseudomonas aeru-
ginosa. Antimicrob Agents Chemother. 2009;53(12):5150–4.
523. Dahdouh E, Gomez-Gil R, Sanz S, Gonzalez-Zorn B, Daoud Z, Mingorance J, et al. A novel
mutation in pmrB mediates colistin resistance during therapy of Acinetobacter baumannii.
Int J Antimicrob Agents. 2017;49(6):727–33.
524. Lean SS, Suhaili Z, Ismail S, Rahman NI, Othman N, Abdullah FH, et al. Prevalence and
genetic characterization of carbapenem- and polymyxin-resistant Acinetobacter baumannii
isolated from a tertiary hospital in Terengganu, Malaysia. ISRN Microbiol. 2014;2014:953417.
525. Adams MD, Nickel GC, Bajaksouzian S, Lavender H, Murthy AR, Jacobs MR, et al.
Resistance to colistin in Acinetobacter baumannii associated with mutations in the PmrAB
two-component system. Antimicrob Agents Chemother. 2009;53(9):3628–34.
526. Arroyo LA, Herrera CM, Fernandez L, Hankins JV, Trent MS, Hancock RE. The pmrCAB
operon mediates polymyxin resistance in Acinetobacter baumannii ATCC 17978 and clini-
cal isolates through phosphoethanolamine modification of lipid A. Antimicrob Agents
Chemother. 2011;55(8):3743–51.
527. Lesho E, Yoon EJ, McGann P, Snesrud E, Kwak Y, Milillo M, et al. Emergence of colistin-­
resistance in extremely drug-resistant Acinetobacter baumannii containing a novel pmrCAB
operon during colistin therapy of wound infections. J Infect Dis. 2013;208(7):1142–51.
528. Jayol A, Nordmann P, Brink A, Poirel L. Heteroresistance to colistin in Klebsiella pneu-
moniae associated with alterations in the PhoPQ regulatory system. Antimicrob Agents
Chemother. 2015;59(5):2780–4.
529. Cheng YH, Lin TL, Pan YJ, Wang YP, Lin YT, Wang JT. Colistin resistance mecha-
nisms in Klebsiella pneumoniae strains from Taiwan. Antimicrob Agents Chemother.
2015;59(5):2909–13.
530. Wright MS, Suzuki Y, Jones MB, Marshall SH, Rudin SD, van Duin D, et al. Genomic and
transcriptomic analyses of colistin-resistant clinical isolates of Klebsiella pneumoniae reveal
multiple pathways of resistance. Antimicrob Agents Chemother. 2015;59(1):536–43.
531. Miller AK, Brannon MK, Stevens L, Johansen HK, Selgrade SE, Miller SI, et al. PhoQ
mutations promote lipid A modification and polymyxin resistance of Pseudomonas aeru-
ginosa found in colistin-treated cystic fibrosis patients. Antimicrob Agents Chemother.
2011;55(12):5761–9.
532. Park YK, Choi JY, Shin D, Ko KS. Correlation between overexpression and amino acid
substitution of the PmrAB locus and colistin resistance in Acinetobacter baumannii. Int
J Antimicrob Agents. 2011;37(6):525–30.
533. Poirel L, Jayol A, Bontron S, Villegas MV, Ozdamar M, Turkoglu S, et al. The mgrB gene
as a key target for acquired resistance to colistin in Klebsiella pneumoniae. J Antimicrob
Chemother. 2015;70(1):75–80.
534. Cannatelli A, D’Andrea MM, Giani T, Di Pilato V, Arena F, Ambretti S, et al. In vivo emer-
gence of colistin resistance in Klebsiella pneumoniae producing KPC-type carbapenemases
mediated by insertional inactivation of the PhoQ/PhoP mgrB regulator. Antimicrob Agents
Chemother. 2013;57(11):5521–6.
535. Cannatelli A, Giani T, D’Andrea MM, Di Pilato V, Arena F, Conte V, et al. MgrB inactivation
is a common mechanism of colistin resistance in KPC-producing Klebsiella pneumoniae of
clinical origin. Antimicrob Agents Chemother. 2014;58(10):5696–703.
536. Olaitan AO, Diene SM, Kempf M, Berrazeg M, Bakour S, Gupta SK, et al. Worldwide emer-
gence of colistin resistance in Klebsiella pneumoniae from healthy humans and patients in Lao
PDR, Thailand, Israel, Nigeria and France owing to inactivation of the PhoP/PhoQ regulator
mgrB: an epidemiological and molecular study. Int J Antimicrob Agents. 2014;44(6):500–7.
537. Jayol A, Nordmann P, Desroches M, Decousser JW, Poirel L. Acquisition of broad-spectrum
cephalosporin resistance leading to colistin resistance in Klebsiella pneumoniae. Antimicrob
Agents Chemother. 2016;60(5):3199–201.
160 C. R. Dean et al.

538. Cheng YH, Lin TL, Lin YT, Wang JT. Amino acid substitutions of CrrB responsible for resis-
tance to colistin through CrrC in Klebsiella pneumoniae. Antimicrob Agents Chemother.
2016;60(6):3709–16.
539. Fernandez L, Gooderham WJ, Bains M, McPhee JB, Wiegand I, Hancock RE. Adaptive resis-
tance to the “last hope” antibiotics polymyxin B and colistin in Pseudomonas aeruginosa
is mediated by the novel two-component regulatory system ParR-ParS. Antimicrob Agents
Chemother. 2010;54(8):3372–82.
540. Gutu AD, Sgambati N, Strasbourger P, Brannon MK, Jacobs MA, Haugen E, et al. Polymyxin
resistance of Pseudomonas aeruginosa phoQ mutants is dependent on additional two-­
component regulatory systems. Antimicrob Agents Chemother. 2013;57(5):2204–15.
541. Fernandez L, Jenssen H, Bains M, Wiegand I, Gooderham WJ, Hancock RE. The two-­
component system CprRS senses cationic peptides and triggers adaptive resistance in
Pseudomonas aeruginosa independently of ParRS. Antimicrob Agents Chemother.
2012;56(12):6212–22.
542. Jeannot K, Bolard A, Plesiat P. Resistance to polymyxins in Gram-negative organisms. Int
J Antimicrob Agents. 2017;49(5):526–35.
543. Halaby T, Kucukkose E, Janssen AB, Rogers MR, Doorduijn DJ, van der Zanden AG, et al.
Genomic characterization of colistin heteroresistance in Klebsiella pneumoniae during a
nosocomial outbreak. Antimicrob Agents Chemother. 2016;60(11):6837–43.
544. Bardet L, Baron S, Leangapichart T, Okdah L, Diene SM, Rolain JM. Deciphering heterore-
sistance to Colistin in a Klebsiella pneumoniae isolate from Marseille, France. Antimicrob
Agents Chemother. 2017;61(6):e00356.
545. Li J, Rayner CR, Nation RL, Owen RJ, Spelman D, Tan KE, et al. Heteroresistance to
colistin in multidrug-resistant Acinetobacter baumannii. Antimicrob Agents Chemother.
2006;50(9):2946–50.
546. Moffatt JH, Harper M, Adler B, Nation RL, Li J, Boyce JD. Insertion sequence ISAba11 is
involved in colistin resistance and loss of lipopolysaccharide in Acinetobacter baumannii.
Antimicrob Agents Chemother. 2011;55(6):3022–4.
547. Bojkovic J, Richie DL, Six DA, Rath CM, Sawyer WS, Hu Q, et al. Characterization of an
Acinetobacter baumannii lptD deletion strain: permeability defects and response to inhibition
of lipopolysaccharide and fatty acid biosynthesis. J Bacteriol. 2015;198(4):731–41.
548. Richie DL, Takeoka KT, Bojkovic J, Metzger LE, Rath CM, Sawyer WS, et al. Toxic accu-
mulation of LPS pathway intermediates underlies the requirement of LpxH for growth of
Acinetobacter baumannii ATCC 19606. PLoS One. 2016;11(8):e0160918.
549. Wei J-R, Richie DL, Mostafavi M, Metzger LE IV, Rath CM, Sawyer WS, Takeoka KT, Dean
CR. LpxK is essential for growth of Acinetobacter baumannii ATCC 19606: relationship
to toxic accumulation of lipid A pathway intermediates. mSPHERE. 2017;4(2). https://doi.
org/10.1128/mSphere.00199-17.
550. Henry R, Vithanage N, Harrison P, Seemann T, Coutts S, Moffatt JH, et al. Colistin-resistant,
lipopolysaccharide-deficient Acinetobacter baumannii responds to lipopolysaccharide loss
through increased expression of genes involved in the synthesis and transport of lipopro-
teins, phospholipids, and poly-β-1,6-N-acetylglucosamine. Antimicrob Agents Chemother.
2012;56(1):59–69.
551. Beceiro A, Moreno A, Fernandez N, Vallejo JA, Aranda J, Adler B, et al. Biological cost of
different mechanisms of colistin resistance and their impact on virulence in Acinetobacter
baumannii. Antimicrob Agents Chemother. 2014;58(1):518–26.
552. Hood MI, Becker KW, Roux CM, Dunman PM, Skaar EP. genetic determinants of intrinsic
colistin tolerance in Acinetobacter baumannii. Infect Immun. 2013;81(2):542–51.
553. Lean SS, Yeo CC, Suhaili Z, Thong KL. Comparative genomics of two ST 195 carbapenem-­
resistant Acinetobacter baumannii with different susceptibility to polymyxin revealed under-
lying resistance mechanism. Front Microbiol. 2015;6:1445.
554. Thi Khanh Nhu N, Riordan DW, Do Hoang Nhu T, Thanh DP, Thwaites G, Huong Lan NP,
et al. The induction and identification of novel Colistin resistance mutations in Acinetobacter
baumannii and their implications. Sci Rep. 2016;6:28291.
4 Resistance of Gram-negative Bacilli to Antimicrobials 161

555. Regenbogen B, Willmann M, Steglich M, Bunk B, Nubel U, Peter S, et al. Rapid and con-
sistent evolution of colistin resistance in XDR Pseudomonas aeruginosa during morbidostat
culture. Antimicrob Agents Chemother. 2017; 61(9):e00043–17.
556. Liu Y-Y, Wang Y, Walsh TR, Yi L-X, Zhang R, Spencer J, et al. Emergence of plasmid-­
mediated colistin resistance mechanism MCR-1 in animals and human beings in China: a
microbiological and molecular biological study. Lancet Infect Dis. 2016;16(2):161–8.
557. Skov RL, Monnet DL. Plasmid-mediated colistin resistance (mcr-1 gene): three months later,
the story unfolds. Euro Surveill. 2016;21(9):30155.
558. Al-Tawfiq JA, Laxminarayan R, Mendelson M. How should we respond to the emergence of
plasmid-mediated colistin resistance in humans and animals? Int J Infect Dis. 2017;54:77–84.
559. McGann P, Snesrud E, Maybank R, Corey B, Ong AC, Clifford R, et al. Escherichia coli
Harboring mcr-1 and blaCTX-M on a novel IncF plasmid: first report of mcr-1 in the United
States. Antimicrob Agents Chemother. 2016;60(7):4420–1.
560. Di Pilato V, Arena F, Tascini C, Cannatelli A, Henrici De Angelis L, Fortunato S, et al.
mcr-1.2, a new mcr variant carried on a transferable plasmid from a colistin-resistant KPC
carbapenemase-producing Klebsiella pneumoniae strain of sequence type 512. Antimicrob
Agents Chemother. 2016;60(9):5612–5.
561. Xavier BB, Lammens C, Ruhal R, Kumar-Singh S, Butaye P, Goossens H, et al. Identification
of a novel plasmid-mediated colistin-resistance gene, mcr-2, in Escherichia coli,
Belgium, June 2016. Euro Surveillance. 2016;21(27). https://doi.org/10.2807/1560-7917.
ES.2016.21.27.30280.
562. Yin W, Li H, Shen Y, Liu Z, Wang S, Shen Z, et al. Novel plasmid-mediated colistin resis-
tance gene mcr-3 in Escherichia coli. MBio. 2017;8(3):e00543.
563. Liu YY, Chandler CE, Leung LM, McElheny CL, Mettus RT, Shanks RMQ, et al. Structural
modification of lipopolysaccharide conferred by mcr-1 in Gram-negative ESKAPE
Pathogens. Antimicrob Agents Chemother. 2017;61(6):e00580.
564. Du H, Chen L, Tang Y-W, Kreiswirth BN. Emergence of the mcr-1 colistin resistance gene in
carbapenem-resistant Enterobacteriaceae. Lancet Infect Dis. 2016;16(3):287–8.
565. Yao X, Doi Y, Zeng L, Lv L, Liu J-H. Carbapenem-resistant and colistin-resistant Escherichia
coli co-producing NDM-9 and MCR-1. Lancet Infect Dis. 2016;16(3):288–9.
566. Lai CC, Chuang YC, Chen CC, Tang HJ. Coexistence of MCR-1 and NDM-9 in a clinical
carbapenem-resistant Escherichia coli isolate. Int J Antimicrob Agents. 2017;49(4):517–8.
567. Mediavilla JR, Patrawalla A, Chen L, Chavda KD, Mathema B, Vinnard C, et al. Colistin- and
carbapenem-resistant Escherichia coli harboring mcr-1 and blaNDM-5, causing a compli-
cated urinary tract infection in a patient from the United States. MBio. 2016;7(4):e01191.
568. Guh AY, Bulens SN, Mu Y, Jacob JT, Reno J, Scott J, et al. Epidemiology of
carbapenem-­ resistant Enterobacteriaceae in 7 US communities, 2012–2013. JAMA.
2015;314(14):1479–87.
569. Castanheira M, Griffin MA, Deshpande LM, Mendes RE, Jones RN, Flamm RK. Detection
of mcr-1 among Escherichia coli clinical isolates collected worldwide as part of the SENTRY
antimicrobial surveillance program in 2014 and 2015. Antimicrob Agents Chemother.
2016;60(9):5623–4.
570. Caniaux I, van Belkum A, Zambardi G, Poirel L, Gros MF. MCR: modern colistin resistance.
Eur J Clin Microbiol Infect Dis. 2017;36(3):415–20.
571. Gao R, Hu Y, Li Z, Sun J, Wang Q, Lin J, et al. Dissemination and mechanism for the MCR-1
colistin resistance. PLoS Pathog. 2016;12(11):e1005957.
572. Sun J, Xu Y, Gao R, Lin J, Wei W, Srinivas S, et al. Deciphering MCR-2 colistin resistance.
MBio. 2017;8(3):e00625.
573. Ye H, Li Y, Li Z, Gao R, Zhang H, Wen R, et al. Diversified mcr-1-harbouring plasmid reser-
voirs confer resistance to colistin in human gut microbiota. MBio. 2016;7(2):e00177.
574. Vaara M, Fox J, Loidl G, Siikanen O, Apajalahti J, Hansen F, et al. Novel polymyxin deriva-
tives carrying only three positive charges are effective antibacterial agents. Antimicrob
Agents Chemother. 2008;52(9):3229–36.
162 C. R. Dean et al.

575. Vaara M, Siikanen O, Apajalahti J, Fox J, Frimodt-Moller N, He H, et al. A novel polymyxin

derivative that lacks the fatty acid tail and carries only three positive charges has strong syn-
ergism with agents excluded by the intact outer membrane. Antimicrob Agents Chemother.
2010;54(8):3341–6.
576. Zabawa TP, Pucci MJ, Parr TR Jr, Lister T. Treatment of Gram-negative bacterial infections
by potentiation of antibiotics. Curr Opin Microbiol. 2016;33:7–12.
577. Quale J, Shah N, Kelly P, Babu E, Backer M, Rosas-Garcia G, et al. Activity of polymyxin B
and the novel polymyxin analogue CB-182,804 against contemporary Gram-negative patho-
gens in New York City. Microb Drug Resist. 2012;18(2):132–6.
578. Magee TV, Brown MF, Starr JT, Ackley DC, Abramite JA, Aubrecht J, et al. Discovery of
Dap-3 polymyxin analogues for the treatment of multidrug-resistant Gram-negative nosoco-
mial infections. J Med Chem. 2013;56(12):5079–93.
579. Velkov T, Roberts KD, Nation RL, Wang J, Thompson PE, Li J. Teaching ‘old’ polymyxins
new tricks: new-generation lipopeptides targeting gram-negative ‘superbugs’. ACS Chem
Biol. 2014;9(5):1172–7.
580. Lomovskaya O, Rubio-Aparicio D, Nelson K, Roberts KD, Thompson PE, Nation RL,
Velkov T, Li J, Hecker SJ, Griffith DC, Dudley MN. In vitro activity of Faddi-287, a repre-
sentative of a novel series of polymyxins (Pm) with reduced nephrotoxic potential. Abstracts
of Microbe 2016. Boston; 2016; Saturday-495.
581. Brown P, Dawson MJ. Development of new polymyxin derivatives for multi-drug resistant
Gram-negative infections. J Antibiot (Tokyo). 2017;70(4):386–94.
582. Velkov T, Roberts KD, Li J. Rediscovering the octapeptins. Nat Prod Rep.
2017;34(3):295–309.
583. Becker B, Butler MS, Hansford KA, Gallardo-Godoy A, Elliott AG, Huang JX, et al.
Synthesis of octapeptin C4 and biological profiling against NDM-1 and polymyxin-resistant
bacteria. Bioorg Med Chem Lett. 2017;27(11):2407–9.
584. Han ML, Shen HH, Hansford KA, Schneider EK, Sivanesan S, Roberts KD, et al.
Investigating the interaction of octapeptin A3 with model bacterial membranes. ACS Infect
Dis. 2017;3(8):606–19.
585. Krishnamoorthy G, Leus IV, Weeks JW, Wolloscheck D, Rybenkov VV, Zgurskaya
HI, Bonomo RA. Synergy between Active Efflux and Outer Membrane Diffusion Defines
Rules of Antibiotic Permeation into Gram-Negative Bacteria. mBio. 2017;8(5):e01172–17.
586. Steeghs L, den Hartog R, den Boer A, Zomer B, Roholl P, van der Ley P. Meningitis bacte-
rium is viable without endotoxin. Nature. 1998;392(6675):449–449.
587. Peng D, Hong W, Choudhury BP, Carlson, RW, Gu XX. Moraxella catarrhalis
Bacterium without Endotoxin, a Potential Vaccine Candidate. Infection and Immunity.
2005;73(11):7569–7577.
Chapter 5
Drug Resistance in Tuberculosis

Neil W. Schluger

5.1  reatment of Tuberculosis and the Generation

T
of Resistance

The problem of drug resistance in tuberculosis was apparent from the first experi-
ment involving antibiotic treatment of this disease. Streptomycin, the first antibiotic
with activity against M. tuberculosis, was discovered by Selman Waksman and
Albert Schatz of Rutgers University in the 1940s. The first rigorous investigation of
its use was conducted by the British Medical Research Council and reported in a
paper in the British Medical Journal published in 1948 [1]. In that experiment, gen-
erally also acknowledged as the first randomized controlled trial ever to be pub-
lished, 100 men were chosen to receive either bed rest, the standard of care at the
time for tuberculosis (TB), or bed rest plus injections of streptomycin. The results
were striking: they demonstrated that streptomycin was clearly effective in patients
with respect to improvements in symptoms, chest radiographic findings, and results
of sputum bacteriology (most of the study participants converted to negative sputum
cultures within a few months after beginning streptomycin injections). However,
after (often very soon after) converting to negative sputum cultures, essentially all
the patients in the trial relapsed and again developed positive sputum cultures. In all
these relapsed cases, cultures that had initially been susceptible to streptomycin had
become resistant. That very first trial, which clearly established that antibiotic treat-
ment of tuberculosis was effective, also demonstrated that drug resistance could
easily emerge when the disease was treated with a single antibiotic for more than a
few days. The need for multidrug regimens, in order to prevent the emergence of
resistance during treatment of tuberculosis, was suggested [2]. Combination

N. W. Schluger (*)
Departments of Medicine, Epidemiology and Environmental Health Sciences,
Columbia University Medical Center, New York, NY, USA
e-mail: [email protected]

© Springer International Publishing AG, part of Springer Nature 2018 163

I. W. Fong et al. (eds.), Antimicrobial Resistance in the 21st Century, Emerging Infectious
Diseases of the 21st Century, https://doi.org/10.1007/978-3-319-78538-7_5
164 N. W. Schluger

treatment became possible in the early 1950s with the introduction of isoniazid, and
the development of other so-called “first-line” drugs (rifampin, ethambutol, pyra-
zinamide) followed.
As noted above, isoniazid (INH) was initially introduced in the early 1950s and
was recognized immediately as an effective antituberculosis drug [3–7]. Although
the BMRC streptomycin trial had certainly brought the emergence of drug resis-
tance to the fore, the highly active properties of INH led some to think that this drug
could be used as a single agent in the treatment of TB disease in poor countries with
limited resources. That would eliminate the need to obtain and use more expensive
and complex agents such as streptomycin, which had to be given by injection. Trials
in the late 1950s and early 1960s in Africa and India using INH monotherapy for TB
showed a high rate of favorable responses, but also a very high rate—as high as
53%—of the development of INH resistance. This certainly (and perhaps predict-
ably) underscored the findings in the initial streptomycin trial.
These early experiments clearly demonstrated the potential for drug-resistant
strains of M. tuberculosis to emerge in a relatively short period of time following
exposure to a single drug. Consequently, the use of single-drug regimens was dis-
couraged and fell out of practice on a programmatic basis. On a worldwide level,
relatively little attention was paid to the issue of drug resistance as a public health
issue for the first several decades of the chemotherapy era.
By the 1980s, a short-course regimen consisting of isoniazid, rifampin, pyrazin-
amide, and ethambutol for 2 months followed by a continuation phase of isoniazid
and rifampin for an additional 4 months (a 6-month course in total) had been shown
to be effective in achieving cure in nearly all cases, and this became and has
remained the standard regimen for treating tuberculosis around the world [8]. When
administered correctly, this regimen should easily guard against the emergence of
drug resistance.

5.2  mergence and Recognition of Drug-Resistant

E
Tuberculosis as a Public Health Issue

Aside from occasional cases of streptomycin-resistant tuberculosis (no longer a

clinical problem since the development of effective multiagent regimens that did not
include this drug) and a low background level of isoniazid monoresistant strains of
M. tuberculosis, drug-resistant tuberculosis, and especially multidrug-resistant
tuberculosis, MDR-TB—defined as tuberculosis caused by strains resistant to at
least INH and rifampin—was not perceived as a significant global problem until
cases started to accumulate in a few cities in the late 1980s and early 1990s. In fact,
in the United States, a series of papers were published in the 1960s and 1970s indi-
cating that drug resistance was becoming less and less of a problem [9–12]. That
trend abruptly reversed in the late 1980s and early 1990s.
The emergence of MDR-TB was perhaps best described in a series of reports
from New York City beginning in the early 1990s [13, 14]. This coincided with a
5 Drug Resistance in Tuberculosis 165

rapid rise in overall TB cases in the United States that began in the mid-1980s and
peaked in 1992. This rise in cases has been attributed to several factors, including
the emergence of the HIV epidemic; the deterioration and neglect of the public
health tuberculosis control infrastructure; social conditions such as drug use, prison
overcrowding, and homelessness; and poor infection control in hospitals and other
congregate facilities.
In 1992, there were 3811 cases of tuberculosis in New York City (an incidence
rate of 50/100,000), and of those roughly 12%, an astonishing percentage, had
MDR-TB [15]. Patients infected with drug-resistant strains were more likely to have
been HIV-infected and to have been previously treated for TB than patients with
drug-susceptible strains. Several instances of nosocomial transmission of MDR-TB,
usually among patients with HIV infection, were also documented using molecular
epidemiology techniques, in several American cities. Extremely worrisome was the
very high mortality rate associated with MDR-TB. In an early report from New York
City, MDR was almost always a fatal disease, with reported mortality rates in excess
of 80% [16].
Following these reports concerning the emergence of drug resistance in New York
City, a landmark global survey was commissioned by the World Health Organization
(WHO) and published in 1998 [17]. In many countries, particularly those with the
highest burden of tuberculosis cases, resources were (and are still) not sufficient to
perform drug susceptibility testing on all M. tuberculosis isolates. The global
MDR-TB survey effort involved a network of 14 reference laboratories around the
world and the development of a standardized methodology for defining and estimat-
ing the number of patients with MDR. The results of the first global MDR-TB sur-
vey were disturbing. Nearly 10% of patients with TB were infected with strains that
had at least one type of drug resistance. Resistance to INH alone was found in 7.3%
of cases. Although only 1.4% of the total cases in the survey were MDR, this distri-
bution was far from uniform, as several hotspots were identified. In the original
survey, prevalence rates of MDR-TB in Estonia and Latvia exceeded 10% among
patients who had never been treated for TB. In patients previously treated, rates of
MDR exceeded 10% in Argentina, Cuba, the Dominican Republic, England and
Wales, Estonia, Latvia, Peru, Portugal, Puerto Rico, South Korea, Romania, Russia,
Sierra Leone, and Spain. Over half of Latvian re-treatment patients were infected
with MDR strains!
Subsequently, regular annual surveys of the prevalence of drug resistance have
been published, and the resulting information has been included in the World Health
Organization’s annual Global TB Report [18]. A more complete, and extremely
frightening, picture of drug-resistant TB around the world has emerged. Several
countries of the former Soviet Union, including Russia, Belarus, Ukraine,
Kazakhstan, and nearby Mongolia, all have rates of drug resistance that surpass
18% in previously untreated cases. In previously treated cases, rates exceed 50% in
several of these countries.
In terms of sheer numbers, China contributes a large percentage of the world’s
drug-resistant cases, even though the percentage of MDR cases there is much lower
than in Russia and the countries of the former Soviet Union.
166 N. W. Schluger

As has been detailed in the WHO’s annual Global TB Report, precise knowledge
about the prevalence of drug-resistant tuberculosis is limited by the inability of
many national TB control programs to conduct drug susceptibility testing for all
diagnosed cases of TB [19]. This limitation is due to a combination of factors,
including clinical diagnosis where culture methods are unavailable. Even when cul-
turing is performed, comprehensive drug susceptibility testing is too complex and
expensive in many resource-constrained settings. Testing has been a particular prob-
lem in Africa, where WHO maps indicate that no data is available for many of the
countries. It is hoped that introduction of genotypic drug susceptibility testing,
based either on nucleic acid amplification or whole-genome sequencing, will pro-
vide more robust and accessible drug susceptibility testing than traditional pheno-
typic testing using solid or liquid media.

5.3 Consequences of Drug Resistance in Tuberculosis

At the time of this writing, isoniazid monoresistance is quite common around the
world, with a prevalence of above 10% in many regions [19]. Evidence of the clini-
cal consequences of INH monoresistance is somewhat mixed. Some data suggest
that patients with this pattern of resistance can be treated with the standard regimen
for drug-susceptible tuberculosis and achieve cure and relapse rates similar to
patients with fully susceptible TB. But other studies have suggested that the stan-
dard short-course regimen is associated with higher treatment failure and relapse
rates when used for INH-monoresistant disease [20–23].
Rifampin monoresistance tends to be lower than with other agents. Many experts
feel that this is because rifampin was introduced relatively late compared with the
other first-line drugs and has never been used as a single agent to treat
tuberculosis.

5.4 Patterns and Treatment of Drug-Resistant Tuberculosis

As noted above, isoniazid monoresistance is fairly common, though it has generally

been thought that INH resistance alone does not portend a poor outcome when the
standard regimen for treating drug-susceptible tuberculosis (2 months of isoniazid,
rifampin, pyrazinamide, and rifampin, followed by 4 months of isoniazid and
rifampin) is used [24, 25]. In recent years, as INH monoresistance has increased in
prevalence, some clinicians have advocated adding or substituting a fluoroquino-
lone into the regimen [20–23, 26].
Streptomycin monoresistance is also reasonably common, but this resistance
pattern is essentially of no practical consequence, as streptomycin is not part of the
standard treatment regimen for drug-susceptible tuberculosis [27]. Rifampin mono-
resistance is less common; it is most often seen in patients with HIV infection who
5 Drug Resistance in Tuberculosis 167

have received intermittent high doses of rifampin [28–31]. As rifampin is the most
active drug in the standard treatment regimen, treatment of rifampin monoresistance
generally falls under the umbrella of treatment for multidrug-resistant
tuberculosis.
Pyrazinamide (PZA) monoresistance has been somewhat difficult to detect
because of the stringent acidic conditions under which PZA phenotypic testing must
be performed. Thus, PZA monoresistance may be more common than previously
thought [32]. Patients with PZA monoresistance require 9 months of treatment with
INH and rifampin rather than the shorter course 6-month regimen.
The term polydrug resistance is used to describe patients who are infected with
isolates that are resistant to isoniazid and one of the other first-line drugs (pyrazin-
amide, ethambutol, streptomycin) but not rifampin. Some data indicate that persons
with polydrug resistance have worse outcomes than patients with fully drug-­
susceptible tuberculosis when treated with standard regimens, though all these data
are retrospective. Optimal treatment regimens have yet to be defined for polyresis-
tant strains of M. tuberculosis.
Multidrug-resistant tuberculosis (MDR) is generally treated with regimens that
include one of the advanced fluoroquinolones (levofloxacin, moxifloxacin, or gati-
floxacin) and an injectable agent (streptomycin, amikacin, kanamycin, or capreo-
mycin). In carefully monitored settings, favorable outcomes for patients with
MDR-TB can be achieved in 80–85% of cases using drug therapy alone [33].
The term pre-XDR-TB has come into use recently to describe isolates of M.
tuberculosis that are resistant to INH, rifampin, and either a fluoroquinolone or one
of the injectable drugs but not both. XDR-TB (extensively drug-resistant TB) indi-
cates an isolate that is resistant to INH, rifampin, fluoroquinolones, and the inject-
ables [34]. Both pre-XDR and XDR-TB are extremely difficult to treat successfully
and require long duration of therapy with drugs that have a high rate of potentially
serious adverse effects. Newer agents, such as bedaquiline, delamanid, and line-
zolid, may be useful in treating these infections, although the optimal dosing regi-
mens with these newer drugs remain to be defined. In cases of XDR and pre-XDR-TB,
surgery may be useful as an adjunctive therapy if the disease is localized.

5.5 Detection of Drug Resistance in M. tuberculosis

The detection and diagnosis of drug resistance in M. tuberculosis is a rapidly evolv-

ing field in which molecular approaches may dramatically change clinical
practice.
For many years, detection of resistance was confined to demonstrating pheno-
typic resistance to growth inhibition with either solid or liquid antibiotic-containing
medium. This type of drug susceptibility testing (DST) remains the gold standard
against which all other techniques are measured [35]. Although approaches to DST
on solid or liquid medium are generally well-standardized, different labs occasion-
ally use different concentrations of drugs to define resistance, or they define
168 N. W. Schluger

r­ esistance as high-level (e.g., if a concentration of INH in the medium is 0.1 μgm/

ml) or low-level resistance, if there is growth at an INH concentration of 0.1 μgm/
ml but inhibition of growth at a concentration of 1.0 μgm/ml. Currently, the most
widely used method of culture for M. tuberculosis is the broth-based non-radiomet-
ric mycobacterial growth indicator tube (MGIT) system, which is used in most labs
in high-resource settings and in reference labs in many high-burden, low-resource
settings.
The biggest change in the diagnosis of tuberculosis drug resistance came with
the identification and understanding of molecular targets of the various antitubercu-
losis drugs and the development of reliable, accurate, and robust methods for iden-
tifying, amplifying, and/or sequencing relevant regions of DNA in clinical samples
[36–41]. The methods most useful in diagnosis of drug resistance in isolates of M.
tuberculosis rely on nucleic acid amplification, most recently whole-genome
sequencing (WGS).
A test called the line probe assay uses the polymerase chain reaction (PCR) to
amplify regions of genes in M. tuberculosis DNA that are associated with drug
resistance [42]. Prior to amplification, clinical samples must be treated and DNA
extracted. These amplified segments of DNA are then hybridized with DNA frag-
ments (probes) that have been immobilized on a strip of nitrocellulose. The cap-
tured, hybridized DNA fragments are detected by a colorimetric method that allows
identification of wild-type and mutated regions of genes associated with drug resis-
tance. The colorimetric assay is read by eye. Many mutations in many genes can be
identified by the line probe assay, and it has been popular as a relatively low-cost
approach to molecular detection of drug resistance in many parts of the world.
A significant advance in the detection of drug resistance occurred with the devel-
opment of the PCR-based nucleic acid amplification GeneXpert system for tubercu-
losis diagnostics [43–45]. This is a highly automated system in which nucleic acid
amplification of the critical segment of the gene associated with rifampin resistance
is linked to molecular beacons that allow semiquantitative detection of the wild-­
type or mutated gene. This is a rapid (~2 h) and nearly completely self-enclosed
system that can detect both the presence of M. tuberculosis in a clinical sample and
the presence or absence of rifampin resistance. Most notably, very little sputum
processing is required. A sputum sample is placed in a cartridge that is then placed
in the GeneXpert machine. Every ensuing step is completely automated. The sensi-
tivity and specificity of rifampin resistance detection, as compared with phenotypic
resistance detection using bacterial culture systems, are very high, and GeneXpert
has rapidly become a widely used technology, even in relatively low-resource set-
tings. Over 20,000 units are in place globally, and roughly half of those machines
are in South Africa [46–48]. The unit must be protected against extremes of tem-
perature, and the cost of the individual cartridges is not inconsiderable. Several
further iterations of the original GeneXpert are in development, including the Omni,
a battery-operated, small portable unit that is meant to be taken into the field, the
Ultra that has sensitivity that should approach that of bacterial culture, and a version
5 Drug Resistance in Tuberculosis 169

of Xpert that tests for resistance to a large number of drugs: isoniazid, rifampin,
fluoroquinolones, and aminoglycosides.
As molecular biology techniques have become more automated and less expen-
sive, the realistic prospect of using whole-genome sequencing (WGS) has emerged
as a means for detecting drug resistance [49]. WGS in theory should be the most
complete means of examining DNA sequences for mutations associated with drug
resistance, whereas PCR-based methods have not, up to now, been able to interro-
gate all mutations in all genes associated with antibiotic resistance. WGS can exam-
ine the entire genome and in theory can detect any mutation known to be associated
with drug resistance. Since this technique has an inherent error rate that could
impair its robustness, the method will have to be evaluated in field trials to move
forward.

5.6  ew Developments in the Diagnosis of Tuberculosis Drug

N
Resistance

It is worth exploring in some depth the role of genotypic detection of drug resistance
in M. tuberculosis in light of developments in molecular biology, gene sequencing,
and highly automated detection of mutations and single nucleotide polymorphisms
that hold the promise for inexpensive, rapid, and accurate detection. As noted above,
the Cepheid GeneXpert platform has now been used for several years to identify M.
tuberculosis in sputum samples and for the rapid detection of rifampin resistance.
Quite recently, a significant expansion of this platform holds the promise of rapid
determination of drug susceptibility for a large number of antibiotics.
Xie and colleagues recently evaluated a rapid molecular drug susceptibility plat-
form for detection of resistance to several anti-TB drugs [45]. This study provides
some of the best data, using the most up-to-date methodology, for evaluating
genotypic and phenotypic approaches to identifying drug-resistant strains of
­
M. tuberculosis. Using a platform derived from the Cepheid MTB/RIF instrument,
samples, obtained from patients in China and South Korea, were interrogated for
the presence of mutations in several genes associated with antibiotic resistance
(katG, inhA, gyrA, and rrs). In theory, this assay could detect resistance to isonia-
zid, fluoroquinolones, and the aminoglycosides kanamycin and amikacin. Results
from the molecular assay were compared with phenotypic testing performed with
the widely used BACTEC MGIT 960 system, using commonly employed break-
point MICs for the determination of resistance: 0.1 μgm/ml for INH, 0.5 and 2.0
μgm/ml for moxifloxacin, 2 μgm/ml for ofloxacin, 1 μgm/ml for amikacin, and 2.5
μgm/ml for kanamycin. As a further check, sequencing of the target genes was car-
ried out using the Sanger method on all the clinical isolates (aliquots of the same
samples that were used for the phenotypic testing).
Using phenotypic testing as the gold standard, the investigational nucleic acid
amplification platform had a sensitivity and specificity for detection of resistance to
170 N. W. Schluger

drugs as follows: for isoniazid, 83.3% and 99.2%, respectively; for ofloxacin, 88.4%
and 96.6%; for moxifloxacin at an MIC of 0.5 μgm/ml, 87.6% and 94.3%; for moxi-
floxacin at an MIC of 2.0 μgm/ml, 96.2% and 84%; for kanamycin 71.4% and
98.4%; and for amikacin, 70.7% and 99.6%. As compared to direct sequencing of
resistance genes, the nucleic acid amplification method had the following sensitivity
and specificity for individual drugs: for isoniazid, 98.1% and 100%, respectively;
for fluoroquinolones, 95.8% and 100%; for kanamycin 92.7% and 99.6%; and for
amikacin, 96.8% and 100%. There were 13 specimens out of 304 that had mutations
that were not detected by the investigational assay.
These results indicate that the highly automated investigational platform, mod-
eled after the current Cepheid MTB/RIF GeneXpert, is a highly accurate means of
identifying mutations in genes that are associated with resistance to the drugs that
define MDR-TB, pre-XDR-TB, and XDR-TB. As compared with phenotypic test-
ing, specificity for drug resistance was quite good. Thus, a result from the investiga-
tional platform indicating drug resistance could be reliably used to exclude that drug
from a therapeutic regimen in nearly every case. However, the sensitivity of the
investigational platform was not quite as good as the specificity. It was excellent for
high-­level resistance to moxifloxacin, somewhat less reliable for low-level resis-
tance to moxifloxacin, for ofloxacin, and for isoniazid, and probably unsatisfactory
for amikacin and kanamycin. Looked at another way, the positive predictive value
for determining drug resistance using the investigational platform was generally
excellent, but the negative predictive value fell short of clinical desirability. If only
the molecular testing were used to guide selection of an antibiotic regimen, as many
as 20% of patients might be treated with drugs, such as aminoglycosides, that would
not be clinically useful and could have serious adverse consequences, such as hear-
ing loss.
The results described above underscore the discrepancies that are sometimes
observed between phenotypic and genotypic drug susceptibility testing. These dis-
crepancies point to at least two possible explanations. First, it is certainly possible
that as yet unidentified mutations in genes not examined are responsible for drug
resistance. Second, it is possible that there are limitations in the critical concentra-
tions used to test drug susceptibility and that the phenotypic results obtained in the
laboratory are in fact not entirely reflective of what might be achieved in clinical
practice using standard dosing of the drugs studied in this experiment. At present,
either of these two possibilities, or both, may be operative (see also Chap. 9 on
heteroresistance).
The investigational nucleic acid platform returned results faster than would have
been obtained using the WHO-recommended line probe assay, and it is more suit-
able for use in local labs, hospitals, and clinics. The line probe assay is generally
confined to use in reference centers in most high-prevalence, low-resource settings.
Since many of these settings already have experience using the MTB/RIF version of
the platform, the only change would be in the cartridge that is used; the remaining
hardware is identical. Also, for the purposes of the study, cultures using MGIT were
highly controlled for quality; that level of quality might not be seen under field con-
ditions, whereas the performance of the DNA amplification platform is fairly robust
under most conditions.
5 Drug Resistance in Tuberculosis 171

It is unclear if, from a programmatic point of view, it is necessary to test initially

for more than rifampin resistance. If an isolate is susceptible to rifampin, patients
can be treated satisfactorily with a regimen for drug-susceptible TB, and further
resistance testing is not likely to provide clinically meaningful information, as long
as the background rates of resistance to isoniazid, pyrazinamide, and fluoroquino-
lones are known to be low.

5.7  iological and Molecular Basis of Drug Resistance

B
in M. tuberculosis

The following discussion focuses on the most clinically important antituberculosis

drugs: isoniazid, rifampin, pyrazinamide, streptomycin and other injectables, fluo-
roquinolones, and the newer agent bedaquiline.

5.7.1 Isoniazid

Isoniazid, or isonicotinic acid hydrazide (INH), was one of the earliest antitubercu-
losis drugs to be developed, coming into clinical use in the 1950s. Eventually,
Winder and Collins demonstrated convincingly that INH worked by inhibiting syn-
thesis of mycolic acids, the long-chain α-alkyl β-hydroxy fatty acids that are essen-
tial components of the mycobacterial cell wall [50, 51]. INH is a prodrug that is
metabolized by the catalase-peroxidase enzyme KatG (encoded by the katG gene)
and then binds to NAD to form an INH-NAD adduct as an intermediate form [52].
The INH-NAD adduct binds to and inhibits a reductase called InhA. InhA plays an
important role in fatty acid synthesis leading to mycolic acid production; conse-
quently, when its action is blocked by the binding of the INH-NAD adduct, cell wall
synthesis is interrupted, thereby accounting for the anti-TB action of isoniazid.
Resistance to INH occurs more commonly than resistance for any other first-line
agent used for the treatment of tuberculosis, and it can develop in a number of ways.
First, initial activation of the prodrug can be inhibited by mutations in the catalase-­
peroxidase enzymes that are required to create the active form of the molecule.
Thus, katG mutations cause INH resistance; they are found in a high percentage of
clinical M. tuberculosis isolates, ranging from 30% to 90% of INH-resistant strains
[53, 54]. Second, the inhibition of InhA by the binding of the INH-NAD adduct can
be inhibited by mutations in the InhA gene, which can also lead to resistance to
INH. The latter mechanism of resistance is observed less commonly than katG
mutations, although it is nonetheless clinically significant [55]. Overall, mutations
responsible for isoniazid resistance occur at a frequency of 10−5 to 10−6 bacilli,
probably the highest frequency of naturally occurring resistance to any of the first-­
line antituberculosis drugs.
172 N. W. Schluger

Although complete deletion of katG has been known to account for resistance to
INH in clinical isolates, mutation within that gene is a much more common mode of
resistance. Over 300 different mutations in katG have been identified. Most muta-
tions, however, occur at codon S315, where each base (AGC) can be found mutated
(mutants have Thr, Asn, Arg, Ile, Gly, or Leu residues). Interestingly, certain muta-
tions are more likely associated with monoresistant strains, while others are more
likely to be found in multidrug-resistant strains [56]. InhA mutations are less often
the cause of INH resistance; they are often seen in cases of low- rather than high-­
level resistance, but they are clinically significant nonetheless [53].
Mutations in several other genes that are involved in the action of INH have been
reported to cause resistance to isoniazid [57]. These include mutations in furA (a
gene whose product regulates katG expression), sigI (a sigma factor that regulates
katG expression), and glf (which encodes an NAD+ − and flavin adenine
dinucleotide-­dependent UDP galactopyranose mutase). These seem to be less com-
mon or important clinically as mediators of isoniazid resistance. As will be dis-
cussed below, the relative contribution of these various mutations in clinical isolates
is important for evaluating diagnostic tests that rely on genotypic rather than pheno-
typic approaches to identifying drug-resistant strains.

5.7.2 Rifampin and Other Rifamycins

Rifampin was first studied as a potential antituberculosis drug in the 1960s, and
resistant strains were quickly identified. Rifampin is a potent, sterilizing drug, with
MICs for susceptible strains generally less than 1.0 μgm/ml [58]. It is the corner-
stone of all modern short-course regimens for the treatment of tuberculosis, and its
introduction and use allowed treatment of tuberculosis to be reduced first to 9- and
then to the current standard 6-month short-course regimen. The introduction of
rifampin made the quite toxic injectable agents, notably streptomycin, obsolete in
the vast majority of cases of drug-susceptible tuberculosis.
Unlike isoniazid, which has several genes that can be associated with the devel-
opment of resistance, resistance to rifampin (and the other clinically used rifamy-
cins, rifabutin, and rifapentine) is controlled almost exclusively by a single gene,
rpoB [59, 60]. Rifampin interrupts RNA synthesis through binding to the beta sub-
unit of RNA polymerase, and mutations in the gene that encodes that subunit, rpoB,
prevent the binding of rifampin and allow RNA synthesis to continue unimpeded.
The binding site of rifampin onto RpoB occurs upstream from its catalytic center,
and elongation of the RNA chain cannot occur.
Mutations of rpoB occur at a natural frequency of 10−7–10−8, much less com-
monly than mutations associated with isoniazid resistance. Detection of rifampin
resistance by genotypic methods has been aided by the fact that nearly all the rpoB
mutations that are associated with resistance occur in a very small hotspot in the
gene, in an 81-base-pair segment [59, 60]. It is felt that mutations in this hotspot
region account for greater than 95% of clinical cases of rifampin resistance. Thus,
5 Drug Resistance in Tuberculosis 173

genotypic tests for rifampin resistance were among the first to come into widespread
clinical use. This is particularly important given rifampin’s importance in the treat-
ment of tuberculosis. Testing for rifampin resistance alone allows one to exclude the
possibility of MDR-TB (if rpoB is wild-type) or to alter therapy to include the use
of potent second-line agents, such as injectable aminoglycosides and fluoroquino-
lones, if rpoB mutations are detected. In general, mutations in rpoB produce high-­
level resistance, requiring MICs of greater than 32 μgm/ml, a serum concentration
not generally achievable with safe and well-tolerated dosing of rifampin. As noted
above, all rifamycins have the same mechanism of action, so any detected mutation
in rpoB should be taken as evidence for resistance. However, it has been noted that
particular mutations, at codons 511, 516, 518, and 522, are associated with retention
of susceptibility to rifabutin [61]. These mutations are less common, however.
Rifampin was the last of the so-called first-line drugs to be developed and to
come into widespread clinical use. It was added to already effective regimens to
allow treatment shortening, but it was never used alone. Likely because of this rea-
son, rifampin monoresistance is substantially less common than multidrug resis-
tance. Recent surveys bear this out. In Peru, a country where there had previously
been a high prevalence of MDR-TB, only 2% of cases were found to have rifampin
monoresistance [62]. Among a recently reported California cohort of HIV-infected
patients (a group previously identified as being at higher risk of rifampin monore-
sistance), only 0.4% were found to have strains with this pattern of resistance [63].
MDR-TB was seen in only 1.5%. As if to underscore the effect of immunocompro-
mise on the development of rifampin monoresistance (RMR), rates of RMR and
MDR were sharply lower in the era of highly effective antiviral therapy than they
were in the pre-HAART era. Although uncommon, in both the Peruvian and
California cohorts, RMR strains were significantly associated with an increased risk
of death, at least double that of patients infected with fully drug-susceptible strains.
In the Western Cape province of South Africa, a region burdened by extraordinarily
high rates of both tuberculosis and HIV infection, the number of cases of RMR-TB
seems to be rising, as a recent report documents a tripling of the number of cases in
a relatively short time frame. This seems to be recapitulating the experience of
rifampin monoresistance in New York City [31, 64] from the early to mid-1990s,
where a sudden rise in cases was noted among patients with HIV infection. At the
time, the incidence of tuberculosis in the city had risen to 50/100,000, the equivalent
of a medium-burden country, and roughly one-third of all persons with tuberculosis
were also infected with HIV at a time when effective antiretroviral therapy was
unavailable for most patients.

5.7.3 Pyrazinamide

Pyrazinamide (PZA) is another critical drug in modern short-course therapy [65].

After rifampin was introduced, it became possible, as noted above, to shorten ther-
apy from 18 months to 9, using only INH and rifampin. By adding PZA for the first
174 N. W. Schluger

2 months of therapy, the duration of treatment could be further shortened to a total

of 6 months. PZA seems to kill a population of bacilli known as persisters, bacteria
that are otherwise apparently impervious to the immediate action of other first-line
drugs.
PZA has a complicated mechanism of action [66–68]. Most notably, it is active
only at an acid pH of 5.5, and it has no activity at neutral pH. Even at pH 5.5, MICs
for PZA are relatively high, in the range of 6–50 μgm/ml. As with INH, PZA is a
prodrug that requires metabolism to an active form for its antibacterial effects to
emerge. The active form of PZA is pyrazinoic acid (POA); the conversion of PZA
to POA is under the control of the pyrazinamidase/nicotinamidase enzyme, which is
encoded by the gene pncA. Interestingly, although nicotinamidase is a ubiquitous
enzyme in prokaryotes, PZA has activity against no other bacteria aside from M.
tuberculosis. POA apparently accumulates in the cell and disrupts the membrane
potential of M. tuberculosis.
The usual function of nicotinamidase is to convert nicotinamide to nicotinic acid
(i.e., niacin), which is then recycled to nicotinamide adenine dinucleotide (NAD)
through a metabolic chain known as the Preiss-Handler pathway. In M. tuberculosis,
the Preiss-Handler pathway is defective, so that nicotinic acid is secreted.
Interestingly, a mycobacterial species closely related to M. tuberculosis, M. bovis, is
lacking a key enzyme in the Preiss-Handler pathway (nicotinic acid phosphoribos-
yltransferase or PncB), and as a result M. bovis does not secrete niacin. This is the
basis for the now largely outmoded niacin laboratory test for the speciation of
mycobacteria. M. bovis strains all appear to have a point mutation from C to G in the
pncA gene, rendering all M. bovis strains resistant to PZA.
The prodrug PZA is thought to enter the cell through passive diffusion and then
is rapidly converted to POA via the mechanism described above. POA exits the cell
both passively and by an efflux mechanism. In an acid extracellular environment,
POA is further modified (protonated) outside the cell and reenters the tubercle bacil-
lus where it accumulates. The introduction of protons into the cell by POA likely
disrupts both the cell membranes and inhibits the function of vital enzymes.
Resistance to PZA is caused by mutations in the pncA gene, which converts the
prodrug to POA. Apparently, no strain of M. tuberculosis is resistant to POA. The
conversion of the prodrug to its active metabolite is thus the critical step in the
action of the drug, and pncA mutations are the clinically relevant changes that gen-
erate PZA resistance. Most mutations in pncA are missense, although some non-
sense mutations and mutations in the promoter region have been identified. Unlike
the mutations in the short hotspot region of rpoB that cause resistance to rifampin,
mutations in pncA are scattered throughout the gene, although there is some cluster-
ing at three different spots.
As noted above, identification of PZA-resistant strains of M. tuberculosis by
traditional phenotypic drug-susceptibility testing is difficult due to the requirement
for an acidic medium for drug activity. This has made precise estimates of the preva-
lence of clinically relevant PZA resistance difficult to obtain. This is discussed
extensively in an excellent review by Zhang and Mitchison [66]. As mutations in
pncA are now known to cause the vast majority (95%) of cases of clinical PZA
5 Drug Resistance in Tuberculosis 175

resistance, the use of genotypic testing has rapidly increased. pncA is 558 base pairs
long, and mutations can occur all along the gene. Thus determining PZA resistance
is a more difficult diagnostic problem than interrogating the 81-base pair hotspot
region of rpoB that is responsible for rifampin resistance.
A recent review of global PZA resistance pooled a large number of studies to
develop estimates of phenotypic and genotypic resistance [32]. Resistance was seen
in every region of the globe. Allowing for the fact that there is probable ascertain-
ment bias in deciding which strains to test for PZA resistance, results are nonethe-
less striking. By phenotypic testing, 16.2% of isolates from a worldwide collection
were determined to be resistant. Geographically, this ranged from 11.4% of M.
tuberculosis isolates in the European region to 21.9% in the Americas. Among
strains at high risk for MDR, PZA resistance was found in 41.3% of patients, and in
strains collected from people with MDR-TB, PZA resistance was found in a stag-
gering 60.5%. In reviewing genotypic testing results, Whitfield and colleagues
determined that 20.9% of a sample of 8651 reported cases showed resistance. As
seen in previous reports of smaller sample sizes, the pncA mutations were diverse
and widely scattered; 608 unique polymorphisms were noted at 397 positions in the
gene. Although some polymorphisms were more common than others, the 20 most
common accounted for only one-third of all isolates with phenotypic resistance.
These more recent estimates by Whitfield differ somewhat from a slightly older
review published by Chang and colleagues [69]. They found that PZA resistance
was present in 51% of multidrug-resistant strains of M. tuberculosis and in only 5%
of non-MDR strains. The differences between these two reviews can be at least
partially explained by sampling or ascertainment bias, as well as some differences
in methodology of resistance detection. It is likely that in the future, more precise
estimates of the prevalence of PZA resistance will be made as molecular testing
becomes used more widely around the world.

5.7.4 Streptomycin and Other Injectable Drugs

As described above, the development of resistance to streptomycin in the 1948 land-

mark BMRC trial led to the initial recognition of mycobacterial drug resistance [1].
The near euphoria, which must have greeted both patients and physicians when the
painful injections of this novel antibiotic led to marked clinical and radiographic
improvement, was followed by the crushing disappointment of relapses with
streptomycin-­resistant organisms. The relatively rapid introduction of INH and
para-aminosalicylic acid soon led to the development of combination chemotherapy
regimens that greatly reduced the emergence resistance and established the princi-
ple that active tuberculosis could never be satisfactorily treated with a single antibi-
otic. Nor should one antibiotic ever be added to a failing drug regimen, as resistance
to that single new drug would almost invariably emerge. Ultimately, the introduc-
tion of rifampin as the cornerstone of short-course therapy relegated streptomycin
and the other injectables to the status of a second-line agents. However, as multidrug
176 N. W. Schluger

resistance has emerged as a major problem in tuberculosis treatment and control

around the world, these drugs have regained clinical importance.
Streptomycin belongs to the aminoglycoside class of antibiotics, as do two of the
other injectables now commonly used, amikacin and kanamycin. A fourth inject-
able, capreomycin, is often assumed to be a member of this class, but in reality, it is
a polypeptide having a different mechanism of action and of acquisition of
resistance.
Streptomycin is a bactericidal drug that typically has MICs in the range of 2–4
μgm/ml [58]. Its mechanism of action is through the inhibition of protein synthesis.
It binds to the 30S bacterial ribosomal subunit at ribosomal protein S12, thus lead-
ing to misreading of mRNA during translation [70]. Resistance to streptomycin is
caused most often by amino acid substitutions in the S12 protein encoded by the
rpsL gene as well as by mutations in a gene, rrs, encoding a streptomycin-binding
protein on the 16S ribosomal subunit [71–73]. Mutations in rpsL are thought to be
responsible for 50% of streptomycin resistance, and rrs mutations are considered
responsible for 20% of resistance. The mutations in rpsL and rrs are found at a rela-
tively few loci, certainly as compared to the generation of PZA resistance by muta-
tions in the pncA gene. In the 30% of strains of M. tuberculosis resistant to
streptomycin that lack rpsL or rrs mutations, other genes have been implicated. A
mutation in a gene called gidB, which encodes a methyltransferase specific for 16S
rRNA, has been found in as many as 20–30% of streptomycin-resistant cases [74].
Amikacin is a derivative of kanamycin. The mechanism of resistance to these
drugs is the same, and it is related, but not identical, to the mechanism of streptomy-
cin resistance [75, 76]. Amikacin and kanamycin inhibit protein synthesis by modi-
fying the structure at 16S rRNA; mutations at a single rrs position, 1400, cause
high-level resistance to both antibiotics.
As previously stated, capreomycin is a polypeptide, not an aminoglycoside,
though it is often classified together with streptomycin, amikacin, and kanamycin
because it, too, cannot be administered orally. Capreomycin also interferes with
protein synthesis, but the target of its action seems to be an rRNA methyltransferase
encoded by a gene called tlyA [77].
Because of the slightly different action mechanisms of the injectable drugs,
resistance to one does not of necessity imply resistance to all. Isolates that are resis-
tant to kanamycin can be assumed to be resistant to amikacin, but such isolates may
still be susceptible to streptomycin or capreomycin, depending on the exact nature
of the mutation causing resistance. Importantly, resistance to streptomycin does not
necessarily imply resistance to amikacin or kanamycin.
Since streptomycin is the oldest antituberculosis antibiotic in clinical use, it is
perhaps not surprising that resistance to it is fairly common, although isolated strep-
tomycin resistance in the absence of any other drug resistance is of no real clinical
significance in the vast majority of cases. Two recent reports, one from Cameroon
and one from China, provide similar estimates for the prevalence of streptomycin
resistance. Sidze and colleagues reported that in Cameroon, streptomycin resistance
in the 1990s was common, with a prevalence of over 15% of cases [27]. Following
this, the National TB Control Program in the country was reorganized to provide
5 Drug Resistance in Tuberculosis 177

better and more consistent care, and by 2011, among smear-positive patients, only
3.3% had streptomycin monoresistance. INH resistance was also lower than in
many parts of the world, at 4.7%. Kanamycin resistance was very uncommon,
reported in only 0.2% of strains tested, and MDR-TB was seen in only 1.1% of
cases in the most recent survey. A recent report from Hunan province in China
found streptomycin resistance in 20.5% of cases [78, 79]. Consistent with the dis-
cussion of resistance mechanism, resistance to capreomycin, amikacin, and kana-
mycin was found in only 2.3%, 1.2%, and 1.8% of isolates, respectively. Mutations
in the rrs gene associated with streptomycin resistance were all in the 388–1084 bp
region, whereas mutations in the gene associated with capreomycin, amikacin, or
kanamycin resistance were in the 1158–1674 bp region, as expected.

5.7.5 Fluoroquinolones

Among the most important developments in tuberculosis therapeutics in the last

several decades is the emergence of fluoroquinolones (FQ) as effective drugs for
this disease. Because of their demonstrated in vitro efficacy against M. tuberculosis,
widespread availability (perhaps too much so), and generally good safety and toler-
ability record, this class of drugs has rapidly established itself as a major weapon in
the treatment of drug-resistant tuberculosis, often becoming the first choice of all
the so-called second-line drugs [80]. In addition, there has been considerable inter-
est in using fluoroquinolones in the treatment of drug-susceptible tuberculosis to
shorten therapy below the 6-month short-course regimens that are currently the
global standard [81–83]. While the proper formula for doing this has not yet been
defined, there are many ongoing clinical trials that are seeking to determine the
optimal use of fluoroquinolones in the treatment of drug-susceptible disease. In
addition, this class of drugs is often used to substitute for isoniazid in cases of resis-
tance or intolerance, although the evidence supporting the need for this is still
incomplete.
The most commonly used fluoroquinolones for the treatment of tuberculosis
have been ofloxacin, levofloxacin, and gatifloxacin, although most patients cur-
rently are treated with levofloxacin or moxifloxacin, as both are more potent in vitro
than ofloxacin, and there have been concerns about episodes of dysglycemia
expressed with the use of gatifloxacin.
The mechanism of action of fluoroquinolones against M. tuberculosis is the same
as it is for all bacteria and involves enzymes known as DNA topoisomerases [84–
86]. These enzymes maintain chromosome topology by regulating DNA supercoil-
ing and helping to unlink tangles of DNA strands (catenanes) to facilitate replication
and transcription. Fluoroquinolones exert their activity by trapping gyrase, the only
type II topoisomerase in M. tuberculosis, on DNA such that the DNA is broken. The
resulting complexes block DNA replication, and, in the bacterial species studied,
chromosome fragmentation and accumulation of toxic reactive oxygen species
occur. Both are expected to contribute to the lethal action of the fluoroquinolones.
178 N. W. Schluger

DNA gyrase is a tetramer, consisting of 2 A and 2 B subunits (A2B2). The sub-

units have different functions and are encoded by two different genes, gyrA and
gyrB. Mutations in gyrA and gyrB are associated with fluoroquinolone resistance. A
conserved region in both genes, known as the quinolone resistance-determining
region (QRDR), is generally the target region in both the gyrA (a 320 bp segment)
and gyrB (a 375 bp segment) genes [87]. Quinolone resistance-causing mutations
are generally clustered around a few particular codons in the gyrA and gyrB genes,
and it is generally thought that two mutations in gyrA or mutations in both gyrA and
gyrB are needed to develop high-level resistance in M. tuberculosis. Single muta-
tions in gyrB are not generally associated with high-level resistance. Codons 90, 91,
and 94 of gyrA are the most commonly mutated sites associated with resistance.
More than with the other antituberculosis drugs discussed so far, there has been
considerable discussion and debate about whether DNA gyrase mutations are com-
pletely necessary or sufficient to cause phenotypic drug resistance to fluoroquino-
lones. This debate is assuming more and more importance as molecular techniques,
such as whole-genome sequencing, are being increasingly used for detecting resis-
tance. In some studies, the percentage of isolates of M. tuberculosis that are pheno-
typically resistant and also have gyrase mutations is extremely high, but in others it
has been quite low. As Zhang and Yew pointed out in their review, the reasons for
these discrepancies might be multiple [58]: There may be differences in methodol-
ogy of genotyping from study to study; there may be differences in the MICs used
to define drug resistance; or there may be other resistance mechanisms at work.
Heteroresistance is also a common explanation.
A recent paper by Avalos and colleagues reviewed data from 46 studies reported
from 18 countries involving nearly 4000 clinical isolates of M. tuberculosis [88].
Overall, 87% of isolates that were phenotypically resistant to moxifloxacin and 83%
of isolates that were phenotypically resistant to ofloxacin contained gyrA mutations.
That leaves a considerable number of isolates that are phenotypically resistant but
that do not have identifiable gyrA mutations.
Most recently, Farhat and colleagues examined a cohort of 172 patients from
Peru who had been treated with fluoroquinolones [89, 90]. In this study, baseline
drug susceptibility testing was done using the method of proportions on 7H10 solid
medium and, for PZA, in the broth-based BACTEC system. Isolates in which there
was phenotypic fluoroquinolone resistance underwent genotypic analysis—
sequencing of the gyrA and gyrB genes using molecular inversion probes. The
authors found that there were significant discrepancies between phenotypic and
genotypic resistance testing. A large percentage of M. tuberculosis isolates that
were phenotypically resistant lacked specific gyr resistance mutations. In addition,
mutations that were generally thought to be associated with high-level resistance
were occasionally (3–4% of the time) found in isolates that were phenotypically
drug susceptible. Importantly though, a consistent finding of the study was that
mutations in gyrase that are associated with high-level resistance were strongly and
consistently correlated with adverse outcomes, including treatment failure and
death. In general, the correlation between poor outcome and resistance testing was
stronger for genotypic resistance determinations than for phenotypic d­ eterminations.
5 Drug Resistance in Tuberculosis 179

Overall, the bulk of studies in this area indicates that gyrA mutations should almost
always be taken to mean that there is high-level resistance to fluoroquinolones.
It is worth noting that the fluoroquinolones are among the most widely used
antibiotics in the world for a host of bacterial infections, and they are available over
the counter and without prescription in many countries. It is probably for this reason
that the prevalence of fluoroquinolone resistance in M. tuberculosis strains is
extraordinarily high in certain locations. Invariably, some patients with tuberculosis
are being treated (either self-treated or under a physician’s care) with fluoroquino-
lones as single agents for misdiagnosed pneumonia or bronchitis, a practice that is
leading to the generation of FQ-resistant tuberculosis. A very recent multi-country
survey of fluoroquinolone resistance in over 5000 tuberculosis patients found that
the prevalence of resistance ranged from 0.5% to 12.4% for levofloxacin and 0.9%
to 14.6% for moxifloxacin when using a breakpoint definition of resistance of MIC
greater than 0.5 μgm/ml, with the highest rates of resistance occurring in Pakistan
[91]. Using an MIC of 2 μgm/ml as a resistance breakpoint for moxifloxacin was
uncommon, regardless of the country examined.

5.7.6 Bedaquiline

It seems appropriate to comment on the mechanism of action and the development

of drug resistance related to the novel drug bedaquiline, as this agent is being
increasingly studied and used in the treatment of multidrug-resistant tuberculosis.
Moreover, there is interest in its use as part of a so-called novel, universal regimen
that could be used with any case of tuberculosis, regardless of susceptibility testing
results to conventional first-line agents.
Bedaquiline, formerly known as TMC207, belongs to a novel class of antituber-
culosis medications [92, 93]. It is an ATP synthase inhibitor that has little activity
against most Gram-positive and Gram-negative bacteria, although it is quite active
against M. tuberculosis and perhaps other mycobacterial species. This drug inacti-
vates the F1/F0-ATP synthase of mycobacteria but not that of mammalian cells.
Since this synthase is a critical enzyme in oxidative phosphorylation, interrupting
its function leads to critical energy depletion and cell death. Bedaquiline exposure
leads to rapid depletion of ATP in mycobacteria, although killing is delayed for
several days after initial administration of the drug, perhaps because of activation of
salvage pathways for energy sources in the cell. At any rate, increasing evidence
suggests that this drug may be highly potent against M. tuberculosis when used
clinically, particularly in regimens that do not contain rifampin, which actually may
be a marked advantage in the treatment of MDR-TB. Since no other drug with this
mechanism of action has been used in the treatment of tuberculosis, resistance to it
should be rare, occurring only as emergence of resistance rather than as transmis-
sion of resistance. However, since the drug has started to find its way into clinical
use (to date there have been no more than several thousand patients worldwide who
have received it), reports of resistance have emerged [94].
180 N. W. Schluger

Studies of the molecular basis of resistance focus on two genes. One, atpE,
encodes the previously mentioned F1/F0 ATP synthase; mutations in this gene have
been reported in about 30% of M. tuberculosis strains that have phenotypic bedaqui-
line resistance. Resistance through this mechanism seems to be mediated by abnor-
malities in the C subunit of the ATP synthase.
In the majority of reports, resistance to bedaquiline appears to be mediated by a
gene called rv0678, which encodes a protein of the same name. All bedaquiline-­
resistant isolates identified in South Africa to date have had mutations in this gene.
(Of note, mutations in this gene also confer resistance to clofazimine, a second-line
drug that had long been relegated to the end of the therapeutic line, but which
recently regained some prominence for the treatment of multidrug-resistant strains,
particularly as a component of the Bangladesh regimen [95, 96].) The rv0678 pro-
tein encodes a transcriptional repressor of the genes encoding the MmpS5-MmpL5
efflux pump, and there is evidence that resistance via this mechanism can be over-
come in part by the use of efflux pump inhibitors such as verapamil. However, the
feasibility of such an approach in clinical practice is unclear, since high, toxic levels
of verapamil may be needed to achieve this effect. The naturally occurring fre-
quency of resistance to bedaquiline is on the order of 10−8, similar to that of rifampin;
thus resistance should, in general, emerge slowly. However, bedaquiline also has a
very long tissue half-life, which may favor selection of resistant strains of M. tuber-
culosis in clinical use. There has not yet been enough experience with this drug to
make confident statements about the likelihood of resistance emerging as a signifi-
cant clinical issue.

5.7.7 Clofazimine

As noted above, clofazimine has long been relegated to the bottom of the list of
second-line agents, primarily because of its side effect profile, which includes skin
hyperpigmentation, which can be quite marked and bothersome to many patients,
although it is generally said to be a reversible condition. Clofazimine is now a com-
ponent of the 9-month Bangladesh regimen for the treatment of MDR-TB that has
recently been endorsed for use in selected cases by the WHO [96]. Additionally,
recent use of clofazimine in murine models of tuberculosis has encouraged recon-
sideration its potential as a treatment-shortening agent for patients with drug-­
susceptible tuberculosis [97].
Even some 50 years after its identification as a drug useful in the treatment of
tuberculosis, the precise mechanism of clofazimine action remains obscure. There
is some evidence that it exerts its effect through disruption of cellular redox cycling,
and there is also evidence that the drug directly disrupts the cell membrane [95].
Clinical isolates that are resistant to clofazimine seem extremely rare, although
resistant strains have been developed in this laboratory. Most of those laboratory
strains have resistance mutations in the same rv0678 gene that was described above
as causing resistance to bedaquiline [95]. This gene is a transcriptional regulator
5 Drug Resistance in Tuberculosis 181

that represses expression of mmpS5-mmpL5, the gene that encodes an efflux pump
of the same name. Interestingly, the rv0678 locus mmpS5-mmpL5 is absent in M.
leprae, the causative agent of leprosy, and no strain of this mycobacterium has been
found that is resistant to clofazimine. There have been other genes (rv1979c and
rv2535c) that have been associated with clofazimine resistance, but the mechanisms
for their involvement have not been identified. As expected, Rv0678 mutations lead
to cross-resistance with bedaquiline.

5.7.8 Multidrug Resistance Mechanisms

Although the majority of cases of antituberculosis drug resistance occur singly for
each drug that is used to treat infection, in at least some cases, efflux pumps are
associated with resistance to multiple drugs, as seen with cancer chemotherapy and
which seems to occur at least in vitro in cases of cross-resistance to bedaquiline and
clofazimine.
Recently, there has been evidence that efflux pumps may be playing a role in
resistance to multiple drugs used to treat tuberculosis. As described by Almeida da
Silva, there are five superfamilies of efflux pumps: the ATP-binding cassette (ABC),
major facilitator superfamily (MFS), resistance nodulation division (RND), small
multidrug resistance (SMR), and multidrug and toxic-compound extrusion (MATE)
[98]. The ABC superfamily is of particular interest in M. tuberculosis. Roughly
2.5% of all genes in M. tuberculosis encode ABC transporters, and more than 12
efflux pump genes have been identified. Genes of other superfamilies have also
been implicated in tuberculosis drug resistance. Several experiments have shown
that MICs of antituberculosis drugs can be reduced by administration of efflux
pump inhibitors such as verapamil. This suggests strongly that these pumps can
play a role in at least some forms of drug resistance.

5.8 Treatment of Drug-Resistant Tuberculosis

At the outset, it is important to state that there is no well-defined regimen for the
treatment of most forms of drug-resistant tuberculosis, whether monoresistant,
polyresistant, multidrug resistant, pre-extensively drug resistant, or extensively
drug resistant. Treatment recommendations for all of these forms of resistance rely
on expert opinion based on experience, observational cohorts, small case series, and
extrapolations from clinical trials early in the antibiotic era when drugs such as
rifampin were not yet available. Newer drugs that have become available in the last
few years, such as delamanid, bedaquiline, fluoroquinolones, and linezolid, have
also been studied mostly as single agents added to optimized background regimens
rather than as part of novel regimens tested in randomized controlled trials.
182 N. W. Schluger

In cases of isoniazid-monoresistant tuberculosis, most experts recommend that

the standard regimen may be used and that it is unnecessary to add additional drugs,
such as fluoroquinolones, to compensate for the loss of INH. Thus, a regimen of
rifampin, pyrazinamide, and ethambutol could be used for a total of 6 months, with-
out the addition of a quinolone [80].
On the other hand, in most cases of rifampin monoresistance (a much less com-
mon form of resistance), most experts recommend treating the patient as if he/she
had multidrug resistance.
The treatment of multidrug-resistant tuberculosis depends to a great degree of
the availability of drug susceptibility testing and of a steady supply of the second-­
line agents that will be required to construct the best possible regimen. If drug sus-
ceptibility testing results are available, then a regimen should be designed using
those results as a guide, with the inclusion at all times of at least three drugs in the
regimen to which the isolate is susceptible. Two of the three drugs in the regimen
should be from the fluoroquinolone and injectable groups of antibiotics. If however,
drug susceptibility testing is not routinely (or immediately) available, regimens
must be constructed on programmatic grounds, taking into account prevailing drug
susceptibility patterns in the community as determined by periodic surveys using
regional, national, or supernational reference laboratories.
At present, there are two main approaches to the design of a regimen for
multidrug-­resistant tuberculosis: the WHO-recommended approach to designing an
MDR regimen [80] and the so-called Bangladesh regimen [96]. The WHO assigns
drugs to categories based on their mechanism of action, like efficacy and safety, and
tolerability profiles. Drugs are then selected from the different categories until a
regimen is constructed. Ideally, such a regimen will consist of a fluoroquinolone, an
injectable (either amikacin or kanamycin, or capreomycin), pyrazinamide, and eth-
ambutol. Overall, the WHO guidelines state that in patients with MDR-TB (or
rifampin monoresistant TB), a regimen with at least five effective TB medicines
during the intensive phase is recommended, including pyrazinamide and four core
second-line TB medicines—one chosen from group A (levofloxacin, moxifloxacin,
or gatifloxacin), one from group B (amikacin, capreomycin, kanamycin, or strepto-
mycin), and at least two from group C (ethionamide or prothionamide, cycloserine
or terizidone, linezolid or clofazimine). If the minimum number of effective TB
medicines cannot be composed as given above, an agent from group D2 (bedaqui-
line, delamanid) and other agents from group D3 (p-aminosalicylic acid, imipenem-­
cilastatin, meropenem, amoxicillin clavulanate, or thiacetazone) may be added to
bring the total to five. In patients with MDR-TB, it is recommended that the regimen
be further strengthened with high-dose isoniazid and/or ethambutol. The total dura-
tion of treatment with the regimen is 18–24 months.
As the WHO guidelines themselves state, these recommendations are not made
with a high degree of confidence or support from controlled clinical trials. Many
experts would favor using bedaquiline (a WHO group D drug) instead of any of the
drugs in group C if bedaquiline can be obtained because of its reported efficacy and
side effect profile, which seem favorable as compared to the group C drugs.
5 Drug Resistance in Tuberculosis 183

The Bangladesh regimen is a 9-month program that consists of an intensive

phase of 4 months (extended up to a maximum of 6 months in cases where sputum
smear conversion does not occur) of gatifloxacin (or moxifloxacin), kanamycin,
prothionamide, clofazimine, high-dose isoniazid, pyrazinamide, and ethambutol.
The intensive phase is followed by a continuation phase of 5 months with gatifloxa-
cin (or moxifloxacin), clofazimine, pyrazinamide, and ethambutol. This regimen is
restricted in the WHO guidelines to patients who have not been previously treated
and in whom the prospect of resistance to fluoroquinolones and aminoglycosides is
felt to be very low. In addition, the regimen is not recommended for use in pregnant
women with tuberculosis.
Although the Bangladesh regimen is appreciably shorter than the standard WHO
approach to the treatment of MDR-TB, it has uncertainties and limitations. First, as
noted above, the only available data concerning its use and success come from two
reports of cohorts rather than from randomized, controlled trials. Second, the num-
ber of patients who are actually eligible for this regimen is a matter of debate, as
many experts have argued that at most 25% of patients with MDR-TB would meet
the criteria that WHO has set for its use.
Very recently, preliminary results from a clinical trial comparing the Bangladesh
regimen with the standard longer course WHO-recommended regimen were
reported in abstract form [99]. The crude favorable outcome rate with the standard,
longer regimen was 80.6%, and for the Bangladesh, regimen was 78.1%. The trial
was designed as a non-inferiority trial, and the lower bound of the confidence inter-
val for the Bangladesh regimen arm included the non-inferiority margin. Thus, a
claim of non-inferiority for the Bangladesh regimen could not be supported by the
findings (that does not mean that the regimen is inferior). In addition, there was a
trend toward worse outcomes using the Bangladesh regimen in patients with HIV
infection that was not appreciated in patients without HIV infection.
A novel regimen for patients with MDR- or XDR-TB is being studied in the
Nix-TB regimen, sponsored by the Global Alliance for TB Drug Development
[100]. In this open-label, single-arm observational trial, patients are being treated
with bedaquiline, linezolid, and Pa-824, an oxazolidinone that is being developed
by the Global Alliance. Preliminary results from this study have been reported in
abstract form and are promising in terms of sputum culture conversion status,
although adverse effects have required careful patient management.
We note that none of the strategies consider the importance of pharmacokinetic
overlap. For combination therapy to severely restrict the emergence of resistance, at
least two agents must be at concentrations above their MICs. If a regimen contains
an agent with a half-life that is much longer than that of other members of the regi-
men, the long half-life compound will be present at the equivalent of monotherapy.
Resistance to that compound is likely to emerge during long treatment periods.
Conversely, good pharmacokinetic overlap is likely to contribute to little emergence
of resistance. An example of the latter was seen in a very small trial in which resis-
tance emerged more of with rifapentine than with rifampin when in combination
with isoniazid (ref. is Andrew Vernon et al., The Lancet 353: 1843–1847).
184 N. W. Schluger

5.9 Conclusions

The problem of drug-resistant tuberculosis, particularly MDR- and XDR-TB, is one

that threatens to upset and reverse decades of progress in tuberculosis control around
the world [101]. Delays in diagnosis are leading to prolonged periods of infectious-
ness with strains of M. tuberculosis that are difficult to treat under the best of cir-
cumstances and that generally are concentrated in countries and regions that lack
access to the most modern tools for drug susceptibility testing and newer, more
potent antibiotics. In order to reverse this trend, a greater sense of urgency is needed,
and resources must be devoted to strengthening basic TB control program activities,
to implementation of novel diagnostics that allow drug susceptibility testing for
first- and second-line drugs, and to assuring that drugs needed for the treatment of
MDR- and XDR-TB are available to patients who need them. Without that kind of
sustained effort, decades of progress in tuberculosis control will be reversed.

Major Points
• Tuberculosis is a serious, often fatal, airborne disease if untreated or
untreatable.
• Drug resistance emerges so readily within individual patients that tuberculosis
serves as a paradigm for implementing combination therapy.
• Most of the antituberculosis drugs are preferentially active with M. tuberculosis;
thus long treatment times do not contribute to resistance in other pathogens (fluo-
roquinolones are an exception).
• Effective treatment regimens exist, even for MDR-TB; however, the necessary
infrastructure is often lacking in resource-challenged countries, which then
become “breeding grounds” for new drug-resistant strains.
• The increasing prevalence of resistance requires that new antituberculosis agents
be developed and administered with sufficient resources and evidence-based
knowledge to stem what is becoming a global healthcare crisis.

References

1. British Medical Research Council. STREPTOMYCIN treatment of pulmonary tuberculosis.

Br Med J. 1948;2:769–82.
2. Mitchison DA. How drug resistance emerges as a result of poor compliance during short
course chemotherapy for tuberculosis. Int J Tuberc Lung Dis. 1998;2:10–5.
3. British Medical Research Council. Comparative trial of isoniazid alone in low and high dos-
age and isoniazid plus pas in the treatment of acute pulmonary tuberculosis in East Africans.
Tubercle. 1960;41:83–102.
4. A concurrent comparison of isoniazid plus PAS with three regimens of isoniazid alone in the
domiciliary treatment of pulmonary tuberculosis in South India. Bull World Health Organ.
1960;23:535–85.
5. Rate of inactivation of isoniazid in South Indian patients with pulmonary tuberculosis*. Bull
World Health Organ. 1961;25:779–92.
5 Drug Resistance in Tuberculosis 185

6. The prevention and treatment of isoniazid toxicity in the therapy of pulmonary tuberculosis.
Bull World Health Organ. 1963;29:457–81.
7. British Medical Research Council. Rate of inactivation of isoniazid in South Indian patients
with pulmonary tuberculosis*. Bull World Health Organ. 1961;25:765–77.
8. Combs DL, O’Brien RJ, Geiter LJ. USPHS Tuberculosis Short-Course Chemotherapy Trial
21: effectiveness, toxicity, and acceptability. The report of final results. Ann Intern Med.
1990;112:397–406.
9. Centers for Disease Control. Primary resistance to antituberculosis drugs—United States.
MMWR Morb Mortal Wkly Rep. 1983;32:521–3.
10. Doster B, Caras GJ, Snider DE. A continuing survey of primary drug resistance in tubercu-
losis, 1961 to 1968. A U.S. Public Health Service cooperative study. Am Rev Respir Dis.
1976;113:419–25.
11. Hobby GL, Johnson PM, Lenert TF, et al. A continuing study of primary drug resistance
in tuberculosis in a Veteran population within the United States. I. Am Rev Respir Dis.
1964;89:337–49.
12. Kopanoff DE, Kilburn JO, Glassroth JL, Snider DE Jr, Farer LS, Good RC. A continu-
ing survey of tuberculosis primary drug resistance in the United States: March 1975 to
November 1977. A United States Public Health Service cooperative study. Am Rev Respir
Dis. 1978;118:835–42.
13. Frieden TR, Sterling T, Pablos-Mendez A, Kilburn JO, Cauthen GM, Dooley SW. The emer-
gence of drug-resistant tuberculosis in New York City. N Engl J Med. 1993;328:521–6.
14. Frieden TR, Fujiwara PI, Washko RM, Hamburg MA. Tuberculosis in New York City—turn-
ing the tide. N Engl J Med. 1995;333:229–33.
15. Health NYCDo. New York City annual tuberculosis report 1992. 1992.
16. Park MM, Davis AL, Schluger NW, Cohen H, Rom WN. Outcome of MDR-TB patients,
1983–1993. Prolonged survival with appropriate therapy. Am J Respir Crit Care Med.
1996;153:317–24.
17. Pablos-Mendez A, Raviglione MC, Laszlo A, et al. Global surveillance for antituberculosis-­
drug resistance, 1994–1997. World Health Organization-International Union against
Tuberculosis and Lung Disease Working Group on Anti-Tuberculosis Drug Resistance
Surveillance. N Engl J Med. 1998;338:1641–9.
18. Zignol M, Dean AS, Falzon D, et al. Twenty years of global surveillance of antituberculosis-­
drug resistance. N Engl J Med. 2016;375:1081–9.
19. World Health Organization. Global tuberculosis report. Geneva: World Health Organization;
2016.
20. Baez-Saldana R, Delgado-Sanchez G, Garcia-Garcia L, et al. Isoniazid mono-resistant tuber-
culosis: impact on treatment outcome and survival of pulmonary tuberculosis patients in
Southern Mexico 1995–2010. PLoS One. 2016;11:e0168955.
21. Chien JY, Chen YT, Wu SG, Lee JJ, Wang JY, Yu CJ. Treatment outcome of patients with
isoniazid mono-resistant tuberculosis. Clin Microbiol Infect. 2015;21:59–68.
22. Gegia M, Winters N, Benedetti A, van Soolingen D, Menzies D. Treatment of isoniazid-­
resistant tuberculosis with first-line drugs: a systematic review and meta-analysis. Lancet
Infect Dis. 2017;17:223–34.
23. Nolan CM, Goldberg SV. Treatment of isoniazid-resistant tuberculosis with isoniazid,
rifampin, ethambutol, and pyrazinamide for 6 months. Int J Tuberc Lung Dis. 2002;6:952–8.
24. Five-year follow-up of a controlled trial of five 6-month regimens of chemotherapy for pul-
monary tuberculosis. Hong Kong Chest Service/British Medical Research Council. Am Rev
Respir Dis. 1987;136:1339–42.
25. Controlled trial of 2, 4, and 6 months of pyrazinamide in 6-month, three-times-weekly regi-
mens for smear-positive pulmonary tuberculosis, including an assessment of a combined
preparation of isoniazid, rifampin, and pyrazinamide. Results at 30 months. Hong Kong
Chest Service/British Medical Research Council. Am Rev Respir Dis. 1991;143:700–6.
186 N. W. Schluger

26. Menzies D, Benedetti A, Paydar A, et al. Standardized treatment of active tuberculosis in

patients with previous treatment and/or with mono-resistance to isoniazid: a systematic
review and meta-analysis. PLoS Med. 2009;6:e1000150.
27. Sidze LK, Mouafo Tekwu E, Kuaban C, et al. Strong decrease in streptomycin-resistance
and absence of XDR 12 years after the Reorganization of the National Tuberculosis Control
Program in the Central Region of Cameroon. PLoS One. 2014;9:e98374.
28. Espinal MA, Laserson K, Camacho M, et al. Determinants of drug-resistant tuberculosis:
analysis of 11 countries. Int J Tuberc Lung Dis. 2001;5:887–93.
29. Meyssonnier V, Bui TV, Veziris N, Jarlier V, Robert J. Rifampicin mono-resistant tuberculosis
in France: a 2005–2010 retrospective cohort analysis. BMC Infect Dis. 2014;14:18.
30. Ridzon R, Whitney CG, McKenna MT, et al. Risk factors for rifampin mono-resistant tuber-
culosis. Am J Respir Crit Care Med. 1998;157:1881–4.
31. Sandman L, Schluger NW, Davidow AL, Bonk S. Risk factors for rifampin-monoresistant
tuberculosis: a case-control study. Am J Respir Crit Care Med. 1999;159:468–72.
32. Whitfield MG, Soeters HM, Warren RM, et al. A global perspective on pyrazinamide resis-
tance: systematic review and meta-analysis. PLoS One. 2015;10:e0133869.
33. Ahmad Khan F, Salim MAH, du Cros P, et al. Effectiveness and safety of standardised shorter
regimens for multidrug-resistant tuberculosis: individual patient data and aggregate data
meta-analyses. Eur Respir J. 2017;50.
34. Shah NS, Wright A, Bai GH, et al. Worldwide emergence of extensively drug-resistant tuber-
culosis. Emerg Infect Dis. 2007;13:380–7.
35. Kim SJ. Drug-susceptibility testing in tuberculosis: methods and reliability of results. Eur
Respir J. 2005;25:564–9.
36. Abebe G, Paasch F, Apers L, Rigouts L, Colebunders R. Tuberculosis drug resistance testing
by molecular methods: opportunities and challenges in resource limited settings. J Microbiol
Methods. 2011;84:155–60.
37. Dominguez J, Boettger EC, Cirillo D, et al. Clinical implications of molecular drug resis-
tance testing for Mycobacterium tuberculosis: a TBNET/RESIST-TB consensus statement.
Int J Tuberc Lung Dis. 2016;20:24–42.
38. Drobniewski F, Cooke M, Jordan J, et al. Systematic review, meta-analysis and economic
modelling of molecular diagnostic tests for antibiotic resistance in tuberculosis. Health
Technol Assess. 2015;19:1–188, vii–viii.
39. Kalokhe AS, Shafiq M, Lee JC, et al. Multidrug-resistant tuberculosis drug susceptibility and
molecular diagnostic testing. Am J Med Sci. 2013;345:143–8.
40. Lin SY, Desmond EP. Molecular diagnosis of tuberculosis and drug resistance. Clin Lab
Med. 2014;34:297–314.
41. Wilson ML. Rapid diagnosis of Mycobacterium tuberculosis infection and drug susceptibil-
ity testing. Arch Pathol Lab Med. 2013;137:812–9.
42. Morgan M, Kalantri S, Flores L, Pai M. A commercial line probe assay for the rapid detec-
tion of rifampicin resistance in Mycobacterium tuberculosis: a systematic review and meta-­
analysis. BMC Infect Dis. 2005;5:62.
43. Boehme CC, Nicol MP, Nabeta P, et al. Feasibility, diagnostic accuracy, and effectiveness
of decentralised use of the Xpert MTB/RIF test for diagnosis of tuberculosis and multidrug
resistance: a multicentre implementation study. Lancet. 2011;377:1495–505.
44. Piatek AS, Tyagi S, Pol AC, et al. Molecular beacon sequence analysis for detecting drug
resistance in Mycobacterium tuberculosis. Nat Biotechnol. 1998;16:359–63.
45. Xie YL, Chakravorty S, Armstrong DT, et al. Evaluation of a rapid molecular drug-­
susceptibility test for tuberculosis. N Engl J Med. 2017;377:1043–54.
46. Lessells RJ, Cooke GS, McGrath N, Nicol MP, Newell ML, Godfrey-Faussett P. Impact of
Point-of-Care Xpert MTB/RIF on tuberculosis treatment initiation. A cluster-randomized
trial. Am J Respir Crit Care Med. 2017;196:901–10.
47. Mohr E, Daniels J, Muller O, et al. Missed opportunities for earlier diagnosis of rifampicin-­
resistant tuberculosis despite access to Xpert(R) MTB/RIF. Int J Tuberc Lung Dis.
2017;21:1100–5.
5 Drug Resistance in Tuberculosis 187

48. Reddy S, Ntoyanto S, Sakadavan Y, et al. Detecting Mycobacterium tuberculosis using

the loop-mediated isothermal amplification test in South Africa. Int J Tuberc Lung Dis.
2017;21:1154–60.
49. Papaventsis D, Casali N, Kontsevaya I, Drobniewski F, Cirillo DM, Nikolayevskyy V. Whole
genome sequencing of Mycobacterium tuberculosis for detection of drug resistance: a sys-
tematic review. Clin Microbiol Infect. 2017;23:61–8.
50. Winder FG, Collins PB. Inhibition by isoniazid of synthesis of mycolic acids in Mycobacterium
tuberculosis. J Gen Microbiol. 1970;63:41–8.
51. Winder FG, Collins P, Rooney SA. Effects of isoniazid on mycolic acid synthesis in
Mycobacterium tuberculosis and on its cell envelope. Biochem J. 1970;117:27P.
52. Zhang Y, Heym B, Allen B, Young D, Cole S. The catalase-peroxidase gene and isoniazid
resistance of Mycobacterium tuberculosis. Nature. 1992;358:591–3.
53. Hazbon MH, Brimacombe M, Bobadilla del Valle M, et al. Population genetics study of
isoniazid resistance mutations and evolution of multidrug-resistant Mycobacterium tubercu-
losis. Antimicrob Agents Chemother. 2006;50:2640–9.
54. Dookie N, Sturm AW, Moodley P. Mechanisms of first-line antimicrobial resistance in multi-­
drug and extensively drug resistant strains of Mycobacterium tuberculosis in KwaZulu-Natal,
South Africa. BMC Infect Dis. 2016;16:609.
55. Niehaus AJ, Mlisana K, Gandhi NR, Mathema B, Brust JC. High prevalence of inhA pro-
moter mutations among patients with drug-resistant tuberculosis in KwaZulu-Natal, South
Africa. PLoS One. 2015;10:e0135003.
56. Huyen MN, Cobelens FG, Buu TN, et al. Epidemiology of isoniazid resistance muta-
tions and their effect on tuberculosis treatment outcomes. Antimicrob Agents Chemother.
2013;57:3620–7.
57. Siu GK, Yam WC, Zhang Y, Kao RY. An upstream truncation of the furA-katG operon con-
fers high-level isoniazid resistance in a Mycobacterium tuberculosis clinical isolate with no
known resistance-associated mutations. Antimicrob Agents Chemother. 2014;58:6093–100.
58. Zhang Y, Yew WW. Mechanisms of drug resistance in Mycobacterium tuberculosis. Int
J Tuberc Lung Dis. 2009;13:1320–30.
59. Telenti A, Imboden P, Marchesi F, Schmidheini T, Bodmer T. Direct, automated detec-
tion of rifampin-resistant Mycobacterium tuberculosis by polymerase chain reaction
and single-strand conformation polymorphism analysis. Antimicrob Agents Chemother.
1993;37:2054–8.
60. Telenti A, Imboden P, Marchesi F, et al. Detection of rifampicin-resistance mutations in
Mycobacterium tuberculosis. Lancet. 1993;341:647–50.
61. Sirgel FA, Warren RM, Bottger EC, Klopper M, Victor TC, van Helden PD. The rationale
for using rifabutin in the treatment of MDR and XDR tuberculosis outbreaks. PLoS One.
2013;8:e59414.
62. Villegas L, Otero L, Sterling TR, et al. Prevalence, risk factors, and treatment outcomes
of isoniazid- and Rifampicin-Mono-Resistant pulmonary tuberculosis in Lima, Peru. PLoS
One. 2016;11:e0152933.
63. Prach LM, Pascopella L, Barry PM, et al. Rifampin monoresistant tuberculosis and HIV
comorbidity in California, 1993–2008: a retrospective cohort study. AIDS. 2013;27:2615–22.
64. Moore M, Onorato IM, McCray E, Castro KG. Trends in drug-resistant tuberculosis in the
United States, 1993–1996. JAMA. 1997;278:833–7.
65. Mitchison DA. The action of antituberculosis drugs in short-course chemotherapy. Tubercle.
1985;66:219–25.
66. Zhang Y, Mitchison D. The curious characteristics of pyrazinamide: a review. Int J Tuberc
Lung Dis. 2003;7:6–21.
67. Konno K, Feldmann FM, McDermott W. Pyrazinamide susceptibility and amidase activity of
tubercle bacilli. Am Rev Respir Dis. 1967;95:461–9.
68. Sun Z, Zhang Y. Reduced pyrazinamidase activity and the natural resistance of Mycobacterium
kansasii to the antituberculosis drug pyrazinamide. Antimicrob Agents Chemother.
1999;43:537–42.
188 N. W. Schluger

69. Chang KC, Yew WW, Zhang Y. Pyrazinamide susceptibility testing in Mycobacterium
tuberculosis: a systematic review with meta-analyses. Antimicrob Agents Chemother.
2011;55:4499–505.
70. Davies J, Gorini L, Davis BD. Misreading of RNA codewords induced by aminoglycoside
antibiotics. Mol Pharmacol. 1965;1:93–106.
71. Finken M, Kirschner P, Meier A, Wrede A, Bottger EC. Molecular basis of streptomycin
resistance in Mycobacterium tuberculosis: alterations of the ribosomal protein S12 gene
and point mutations within a functional 16S ribosomal RNA pseudoknot. Mol Microbiol.
1993;9:1239–46.
72. Honore N, Cole ST. Streptomycin resistance in mycobacteria. Antimicrob Agents Chemother.
1994;38:238–42.
73. Nair J, Rouse DA, Bai GH, Morris SL. The rpsL gene and streptomycin resistance in
single and multiple drug-resistant strains of Mycobacterium tuberculosis. Mol Microbiol.
1993;10:521–7.
74. Spies FS, da Silva PE, Ribeiro MO, Rossetti ML, Zaha A. Identification of mutations related
to streptomycin resistance in clinical isolates of Mycobacterium tuberculosis and possible
involvement of efflux mechanism. Antimicrob Agents Chemother. 2008;52:2947–9.
75. Alangaden GJ, Kreiswirth BN, Aouad A, et al. Mechanism of resistance to amikacin and
kanamycin in Mycobacterium tuberculosis. Antimicrob Agents Chemother. 1998;42:1295–7.
76. Suzuki Y, Katsukawa C, Tamaru A, et al. Detection of kanamycin-resistant Mycobacterium
tuberculosis by identifying mutations in the 16S rRNA gene. J Clin Microbiol. 1998;36:1220–5.
77. Maus CE, Plikaytis BB, Shinnick TM. Mutation of tlyA confers capreomycin resistance in
Mycobacterium tuberculosis. Antimicrob Agents Chemother. 2005;49:571–7.
78. Zhao LL, Liu HC, Sun Q, et al. Identification of mutations conferring streptomycin resistance
in multidrug-resistant tuberculosis of China. Diagn Microbiol Infect Dis. 2015;83:150–3.
79. Zhao LL, Chen Y, Chen ZN, et al. Prevalence and molecular characteristics of drug-­
resistant Mycobacterium tuberculosis in Hunan, China. Antimicrob Agents Chemother.
2014;58:3475–80.
80. WHO treatment guidelines for drug-resistant tuberculsis (2016 update). Geneva: World
Health Organization; 2016.
81. Gillespie SH, Crook AM, McHugh TD, et al. Four-month moxifloxacin-based regimens for
drug-sensitive tuberculosis. N Engl J Med. 2014;371:1577–87.
82. Jindani A, Harrison TS, Nunn AJ, et al. High-dose rifapentine with moxifloxacin for pulmo-
nary tuberculosis. N Engl J Med. 2014;371:1599–608.
83. Merle CS, Fielding K, Sow OB, et al. A four-month gatifloxacin-containing regimen for treat-
ing tuberculosis. N Engl J Med. 2014;371:1588–98.
84. Drlica K, Malik M. Fluoroquinolones: action and resistance. Curr Top Med Chem.
2003;3:249–82.
85. Alangaden GJ, Manavathu EK, Vakulenko SB, Zvonok NM, Lerner SA. Characterization
of fluoroquinolone-resistant mutant strains of Mycobacterium tuberculosis selected in the
laboratory and isolated from patients. Antimicrob Agents Chemother. 1995;39:1700–3.
86. Takiff HE, Salazar L, Guerrero C, et al. Cloning and nucleotide sequence of Mycobacterium
tuberculosis gyrA and gyrB genes and detection of quinolone resistance mutations.
Antimicrob Agents Chemother. 1994;38:773–80.
87. Zhou J, Dong Y, Zhao X, et al. Selection of antibiotic-resistant bacterial mutants: allelic
diversity among fluoroquinolone-resistant mutations. J Infect Dis. 2000;182:517–25.
88. Avalos E, Catanzaro D, Catanzaro A, et al. Frequency and geographic distribution of gyrA
and gyrB mutations associated with fluoroquinolone resistance in clinical Mycobacterium
tuberculosis isolates: a systematic review. PLoS One. 2015;10:e0120470.
89. Farhat MR, Jacobson KR, Franke MF, et al. Gyrase mutations are associated with vari-
able levels of fluoroquinolone resistance in Mycobacterium tuberculosis. J Clin Microbiol.
2016;54:727–33.
90. Farhat MR, Mitnick CD, Franke MF, et al. Concordance of Mycobacterium tuberculo-
sis fluoroquinolone resistance testing: implications for treatment. Int J Tuberc Lung Dis.
2015;19:339–41.
5 Drug Resistance in Tuberculosis 189

91. Zignol M, Dean AS, Alikhanova N, et al. Population-based resistance of Mycobacterium

tuberculosis isolates to pyrazinamide and fluoroquinolones: results from a multicountry sur-
veillance project. Lancet Infect Dis. 2016;16:1185–92.
92. Andries K, Verhasselt P, Guillemont J, et al. A diarylquinoline drug active on the ATP syn-
thase of Mycobacterium tuberculosis. Science. 2005;307:223–7.
93. Diacon AH, Pym A, Grobusch MP, et al. Multidrug-resistant tuberculosis and culture conver-
sion with bedaquiline. N Engl J Med. 2014;371:723–32.
94. Andries K, Villellas C, Coeck N, et al. Acquired resistance of Mycobacterium tuberculosis to
bedaquiline. PLoS One. 2014;9:e102135.
95. Cholo MC, Mothiba MT, Fourie B, Anderson R. Mechanisms of action and therapeutic effi-
cacies of the lipophilic antimycobacterial agents clofazimine and bedaquiline. J Antimicrob
Chemother. 2017;72:338–53.
96. Aung KJ, Van Deun A, Declercq E, et al. Successful ‘9-month Bangladesh regimen’ for
multidrug-­resistant tuberculosis among over 500 consecutive patients. Int J Tuberc Lung Dis.
2014;18:1180–7.
97. Converse PJ, Tyagi S, Xing Y, et al. Efficacy of rifampin plus clofazimine in a murine model
of Mycobacterium ulcerans disease. PLoS Negl Trop Dis. 2015;9:e0003823.
98. da Silva PE, Von Groll A, Martin A, Palomino JC. Efflux as a mechanism for drug resistance
in Mycobacterium tuberculosis. FEMS Immunol Med Microbiol. 2011;63:1–9.
99. Results from the STREAM stage 1 randomised controlled trial for treatment of MDR-TB
using a standardised shorter regimen. At http://www.who.int/tb/features_archive/
stream_stage1_MDRTB/en/.
100. A Phase 3 study assessing the safety and efficacy of Bedaquiline Plus PA-824 Plus linezolid
in subjects With drug resistant pulmonary tuberculosis. At https://clinicaltrials.gov/ct2/show/
NCT02333799.
101. Dheda K, Gumbo T, Maartens G, et al. The epidemiology, pathogenesis, transmission, diag-
nosis, and management of multidrug-resistant, extensively drug-resistant, and incurable
tuberculosis. Lancet Respir Med. 2017;5:291–360.
Chapter 6
Anaerobic Bacteria: Antimicrobial
Susceptibility Testing and Resistance
Patterns

Audrey N. Schuetz

6.1 Introduction

Anaerobic bacterial pathogens cause serious infections, particularly in the hospital

environment. Among the better known anaerobic bacteria are Bacteroides fragilis
associated with intra-abdominal abscesses, clostridia recovered from skin and soft
tissue infections, Fusobacterium necrophorum leading to Lemierre disease, and
Clostridium difficile causing disease in critically ill patients. Antibiotics are a
mainstay of treatment protocols, and, as with aerobic pathogens, resistance is
becoming a problem [1]. Anaerobic pathogens are generally more difficult to culture
than aerobic bacteria, and most antimicrobials have been developed using aerobic
assays. Thus, our ability to manage populations of anaerobic microbes lags far
behind control methods for aerobic bacteria. To understand some of the problems, it
is necessary to define terms and methods that are often glossed over in discussions
of aerobic pathogens.
A suitable starting point is with the terms “susceptible (S),” “intermediate (I),”
and “resistant (R).” These are “interpretive categories” used to explain the results
obtained from in vitro antimicrobial susceptibility testing of microorganisms
isolated from infected patients. In a general sense, they predict a patient’s response
to treatment with a particular antimicrobial agent. An isolate that tests resistant to an
antimicrobial is highly unlikely to respond to treatment with that antimicrobial,
while an isolate that tests susceptible is likely to respond clinically. However, even
when an isolate is “susceptible” to an antimicrobial agent in vitro, the ultimate
outcome in the patient is very dependent on the condition of the patient and the site
of the infection.

A. N. Schuetz (*)
Mayo Clinic, Rochester, MN, USA
e-mail: [email protected]

© Springer International Publishing AG, part of Springer Nature 2018 191

I. W. Fong et al. (eds.), Antimicrobial Resistance in the 21st Century, Emerging Infectious
Diseases of the 21st Century, https://doi.org/10.1007/978-3-319-78538-7_6
192 A. N. Schuetz

Interpretive categories are interrelated with minimal inhibitory concentrations

(MICs) and “breakpoints” (also known as clinical breakpoints). An MIC is the low-
est concentration of an antimicrobial agent that prevents the growth of a microor-
ganism as measured with in vitro susceptibility tests. “Breakpoints” are the MIC
values that categorize an isolate as S, I, or R to a specific antimicrobial agent.
The process used to establish breakpoints involves examination of three types of
data:
MIC distribution data, pharmacokinetic/pharmacodynamic (PK/PD) data, and
clinical outcome data. For MIC distribution data, the MICs determined for a large
number of randomly selected isolates (i.e., at least 100 isolates) are plotted as a
distribution curve or graph (Fig. 6.1). The MIC distribution curve may demonstrate
a unimodal or bimodal pattern depending on the antimicrobial and organism. In
general, isolates with lower MICs are less likely to carry microbial resistance
mechanisms, and isolates with higher MICs (e.g., on the right of the curve) are more
likely to express resistance mechanisms.
In addition to the MIC distribution curve, PK/PD data must be obtained by exam-
ining the levels of antimicrobial agent attained in the patient and the manner in
which the antimicrobial agent acts on the infecting organisms. Finally, clinical
outcome data are based on how well the patient does when treated for an infection
due to a specific organism having a specific MIC (Fig. 6.2). All these data are
gathered and used to set breakpoints. S and R interpretive categories are fairly
straightforward; however, the meaning of the intermediate category is complex.
This category includes isolates having MICs between the susceptible and resistant
populations such that the MIC is usually below attainable blood and tissue levels but
for which good clinical response rates may be lower than for isolates with MICs in
the susceptible range.

Fig. 6.1 Minimal inhibitory concentration (MIC) distribution curve of meropenem for 976
Bacteroides fragilis isolates with Clinical and Laboratory Standards Institute (CLSI) breakpoints
noted
6 Anaerobic Bacteria: Antimicrobial Susceptibility Testing and Resistance Patterns 193

Fig. 6.2 Graphic display of the general process by which breakpoints are determined

Committees of experts who are members of breakpoint-setting organizations

examine available data and determine breakpoints. In the USA, the Clinical and
Laboratory Standards Institute (CLSI) sets breakpoints. In Europe, the European
Committee for Antimicrobial Susceptibility Testing (EUCAST) sets breakpoints.
These committees do not always agree on the breakpoints, especially when the
distribution of MIC values is indistinct or when the clinical treatments (e.g., treat-
ment dosages) differ between the patient populations examined.
Information collected from clinical laboratories can be used to determine the
prevalence of resistance for a particular pathogen, antimicrobial agent, and region
(resistance is fundamentally a local phenomenon). Data from such resistance sur-
veys (e.g., antibiograms) guide clinicians in their choices of antimicrobial for
empirical therapy (i.e., treatment either without culturing the pathogen or when
the organism identification is known but the MIC has not yet been measured).
When surveys are carried out in the same location over a period of years, they can
provide insight into how quickly resistance is increasing or whether changes in
antimicrobial exposure are lowering the prevalence of resistance. Many assump-
tions underlie such longitudinal studies, including unchanging breakpoints, stable
patient populations, and comparable sampling from 1 year to the next.
Consequently, the methods underlying antimicrobial susceptibility testing (AST),
the determination of breakpoints, and whether the breakpoint committees agree
are crucial to interpreting survey data. These factors are sometimes problematic
with anaerobic pathogens. Nevertheless, important statements have emerged con-
cerning these bacteria. I begin this chapter with a discussion of AST. A discussion
of antimicrobial resistance mechanisms of major groups of anaerobic pathogens
then follows.
194 A. N. Schuetz

6.2 Antimicrobial Susceptibility Testing of Anaerobes

With the increasing use of MALDI-TOF MS (matrix-assisted laser desorption/ion-

ization-time of flight mass spectrometry) by clinical laboratories to identify anaero-
bic bacteria, accurate species identification is becoming increasingly available. As
compared to the limited identifications of anaerobes in the past by phenotypically or
biochemically based methods, MALDI-TOF MS offers faster and more specific
identification. An expected result is that increasing attention is being given to anaer-
obes in their role in human infections. Although the data are only anecdotal, it is
believed by some that AST is being requested more often by clinicians. In the past,
anaerobic pathogens were often reported only with broad terms such as “Gram-
positive anaerobe,” while now they are often being reported using genus and species
names. Thus, anaerobe susceptibility testing and reporting not only continue to
remain relevant, but they are also likely to increase in demand as clinical microbiol-
ogy laboratories improve technologically.
Empirical information regarding antimicrobial susceptibility can be particularly
helpful for clinicians treating a patient with an infection due to an anaerobe. In con-
trast to many aerobic bacteria, anaerobes tend to be slow-growing, and pathogen
identification is often obtained well before antimicrobial susceptibility test results.
In addition, anaerobes are often present in polymicrobial aerobic/anaerobic infec-
tions or mixed with other anaerobes. Thus, it may be difficult to obtain pure cultures
for AST. In such cases, it is helpful to know the predicted anaerobic susceptibility
patterns based on information about strains circulating in the local region.
Individual hospitals and healthcare systems perform periodic surveillance of
anaerobes and publish the results as susceptibility rates (antibiograms). CLSI
guidance M39 Analysis and Presentation of Cumulative Antimicrobial Susceptibility
Test Data, which is currently under revision, provides recommendations for the
proper collection, analysis, and presentation of antimicrobial susceptibility test data
[2]. If a clinical laboratory cannot perform AST for anaerobes, it may be helpful for
the laboratory and local clinicians to refer to the anaerobe antibiogram data
published in the CLSI M100 and CLSI M11 documents [3, 4]. Changing trends in
antimicrobial resistance over time, based on the antibiogram data, have been
published by members of the CLSI [5]. It is important to recognize that the data
presented in the anaerobe antibiogram from CLSI are derived from isolates obtained
globally and may not be applicable to isolates within a given hospital or region.
Analysis and comparison of published reports of anaerobic antimicrobial suscep-
tibility patterns is complex for several reasons. First, many different methods are
used for susceptibility testing of anaerobes; different testing methods can lead not
only to different MICs but also to different interpretive categories (e.g., susceptible
or resistant). There are two reference methods for anaerobic susceptibility testing
accepted by breakpoint-setting organizations: broth microdilution for the B. fragilis
group of bacteria and agar dilution for all anaerobes, including the B. fragilis group.
In the broth microdilution assay, a series of wells in a microtiter plate, containing
dilutions of the test antimicrobial and bacterial culture, are incubated to determine
6 Anaerobic Bacteria: Antimicrobial Susceptibility Testing and Resistance Patterns 195

the lowest drug concentration that blocks visible bacterial growth. In the agar dilu-
tion assay, the antimicrobial is placed in agar on which a dilute bacterial culture at a
standardized concentration is applied (a set of plates is prepared in which the anti-
microbial concentration is varied). The drug concentration that inhibits colony for-
mation after suitable incubation is taken as the MIC. Another method that is used by
clinical laboratories is called agar gradient diffusion. Examples of this commer-
cially available method include the Etest (bioMérieux, Durham, NC) and the MTS
strip (Liofilchem, Italy). Plastic or paper strips are impregnated on one face with
concentrations of an antimicrobial in a gradient fashion. When the strip is placed on
an agar plate that has been inoculated with a standardized number of bacteria, the
antimicrobial agent diffuses onto the agar and inhibits bacterial growth where con-
centrations are above the MIC. The MIC is read from a scale printed on the strip –
the point at which the ellipse of bacterial growth meets the strip is taken as
MIC. Results from agar gradient diffusion assays may not always correlate well
with the gold standards of broth microdilution and agar dilution, but these commer-
cial strips are widely used due to their ease of use [6]. Although appropriate for use
with aerobic organisms, disk diffusion testing (placement of antibiotic-impregnated
disks onto an agar plate in which a bacterial culture grows) is not suggested by the
CLSI as a test method for anaerobes due to inaccurate results and poor correlation
with the agar dilution method [3]. However, some research groups are currently try-
ing to develop EUCAST disk diffusion breakpoints for testing of the B. fragilis
group organisms [7]. Unfortunately, MIC results from these various AST methods
do not always correlate well with each other. Thus, knowledge of the testing method
used to obtain a particular data set is important when assessing the literature or
reviewing the MIC of a particular isolate. Moreover, the assays can be difficult to
adapt for routine testing by clinical microbiology laboratories. The important point
is that published reports using anaerobic AST may vary in resistance rates simply
from differences among the methods used to determine the MIC.
AST for anaerobes is still an unsettled situation. For example, CLSI recommends
agar dilution as the testing method for anaerobes, but few clinical laboratories use
this methodology [3]. Use of broth microdilution is limited to the B. fragilis group.
CLSI is now undertaking studies to reevaluate whether broth microdilution can be
performed for anaerobes other than the B. fragilis group, but results for other
anaerobes, such as C. difficile, do not look promising [8]. Moreover, broth
microdilution itself can be difficult to perform, and restricting it to only one group
of anaerobic organisms limits its utility for clinical laboratories where many types
of anaerobes other than solely members of the B. fragilis group are isolated.
When considering the published literature, careful attention must be placed on
the breakpoints applied to the datasets, because breakpoints can differ between
breakpoint-setting organizations (see Table 6.1 below for examples of differing
breakpoints for CLSI and EUCAST). Thus, when comparing resistance prevalence
between regions, one may expect a higher prevalence of resistance reported for
ertapenem if the EUCAST resistant breakpoint is applied rather than the CLSI
breakpoint, which for this drug is three doubling dilutions higher. The reason for the
196 A. N. Schuetz

Table 6.1 Examples of current resistant breakpoint differences for EUCAST and CLSI
EUCAST resistant CLSI resistant
Antimicrobial Organism(s) breakpoint breakpoint
Penicillin Gram-positive and Gram-negative >0.5 μg/mL ≥2 μg/mL
anaerobes
Piperacillin-­ Gram-positive and Gram-negative >16/4 μg/mL ≥128/4 μg/mL
tazobactam anaerobes
Ertapenem Gram-positive and Gram-negative >1 μg/mL ≥16 μg/mL
anaerobes
Metronidazole Gram-positive and Gram-negative >4 μg/mL ≥32 μg/mL
anaerobes, except C. difficile

different breakpoints is due in part to the different testing methodologies used in the
data gathered for setting breakpoints and to differences in interpretation of break-
point data.
Updates in bacterial taxonomy have also contributed to confusion in interpreta-
tion of previous reports. Some anaerobes that had previously been grouped within a
particular species are now grouped in a different genus (e.g., Peptoniphilus harei
was previously listed within the Peptostreptococcus genus). Other anaerobes have
been newly described (e.g., Murdochiella asaccharolytica), and it is uncertain
where they would have fallen in prior classification schemes [9, 10]. Taxonomic
refinements are ongoing for anaerobes and will continue to require close attention
until new data sets are accumulated.
Finally, the increasing use of nucleic acid and proteomic methods to identify
anaerobic bacteria has resulted in more accurate differentiation and characterization
of isolates. Thus, more anaerobes are identified to the genus or species level as
compared to previous practices that occasionally reported the organism according
to a morphologic group (e.g., anaerobic Gram-positive cocci). As a result, more
accurate AST results are published on anaerobes when attributed to particular
bacterial species.
In summary, surveys of resistance trends among anaerobes vary in their quality
and size. Resistance patterns differ by geographic region, and some resistance rates
have changed significantly over time. It is unclear how much of the differences
among regions is due to true geographic variation, to methodologic testing variation,
or to application of different breakpoints. Patterns of resistance that deserve close
attention due to rising rates include:
• Resistance to the β-lactam-β-lactamase inhibitor combinations among the B. fra-
gilis group.
• Increasing clindamycin resistance among all anaerobes.
• Metronidazole resistance is no longer limited to the B. fragilis group, as it now
includes Gram-positive cocci and bacilli.
• Resistance of Clostridium to vancomycin.
Moreover, many anaerobes are frequently isolated from polymicrobial infections;
sometimes as many as eight or more different organisms are cultured from the site
6 Anaerobic Bacteria: Antimicrobial Susceptibility Testing and Resistance Patterns 197

of infection. The importance of performing AST on anaerobes from mixed infec-

tions can be questionable.
Below I consider several groups of anaerobes and the types of antimicrobial
resistance they display. As with aerobic bacteria, anaerobes have developed multiple
mechanisms that lead to antimicrobial resistance. These resistance factors include
genes located in the bacterial chromosome or within mobile genetic elements
(plasmids or transposons) leading to production of enzymes that degrade
antimicrobials, as well as to alterations in antimicrobial targets.

6.3 Anaerobic Cocci: Gram-Positive and Gram-Negative

6.3.1 Clinical Disease and Taxonomic Changes

Gram-positive anaerobic cocci comprise approximately one-fourth of all anaerobic

isolates cultivated from human infections [11]. This group of anaerobes is part of
the normal flora of the human mouth and upper respiratory tract, but it also com-
prises normal flora of the gastrointestinal tract, female genitourinary tract, and skin
[11]. Oral infections due to Gram-positive anaerobic cocci are usually polymicro-
bial in nature and often are associated with a breach in the mucosal barrier or an
immunocompromised state. These anaerobes are often implicated in cases of peri-
odontitis, gingivitis, and abscesses in and around the oropharynx, mouth, and neck.
Gram-positive anaerobic cocci are also associated with anaerobic infections of the
lower respiratory tract, such as empyema or aspiration pneumonia [11]. Cases of
gynecological and obstetrical sepsis may be due to Gram-positive anaerobic cocci
[12]. Skin and soft tissue abscesses due to Gram-positive anaerobic cocci are usu-
ally located in the neck or perigenital area due to the preponderance of these anaer-
obes as normal flora in these anatomic sites. Finally, Gram-positive anaerobic cocci
are highly associated with diabetic foot infections [11]. They are less commonly
causes of bloodstream infections or intra-abdominal infections as compared to
B. fragilis or other Gram-negative anaerobic bacilli.
Anaerobic Gram-negative cocci include Veillonella and Megasphaera, among
others. Veillonella is the most commonly recovered anaerobic Gram-negative
coccus in the clinical microbiology laboratory. It is found as normal flora in the
upper respiratory and gastrointestinal tracts and may cause disease in these anatomic
sites when defenses are compromised. Veillonella is often recovered as part of a
mixed culture; its virulence capabilities are poorly understood [10].
More than any other group of anaerobes, the anaerobic Gram-positive cocci have
undergone major taxonomic updates within the past 20 years [10]. For example,
there are now over 13 different genera of anaerobic bacteria that were previously
classified within the genus Peptostreptococcus, which formerly was one of the
largest anaerobic groupings [9, 10]. Now only four species remain in the
Peptostreptococcus genus: P. anaerobius, P. canis, P. russellii, and P. stomatis.
198 A. N. Schuetz

Taxonomic updates and the increased ability to identify organisms more accurately
over time must be taken into account when interpreting AST resistance trends and
patterns over different geographic regions.

6.3.2 Resistance Patterns

Anaerobic Gram-positive cocci are generally considered susceptible to most anaer-

obic antimicrobials including penicillin, piperacillin-tazobactam, meropenem, and
metronidazole. In most cases, infections with these anaerobic cocci are polymicro-
bial: other anaerobes and aerobes are often present. Thus, providing AST results to
clinicians for anaerobic Gram-positive cocci may not be as critical in guiding the
choice of antimicrobial therapy as for the more resistant anaerobic Gram-negative
bacteria, such as B. fragilis. For polymicrobial anaerobic and aerobic infections, the
empiric antimicrobial chosen by clinicians tends to be broad-­spectrum, thereby cov-
ering both aerobes and anaerobes. Although it is likely that broad-spectrum antimi-
crobials do cover anaerobic Gram-positive cocci, this is not always the case. If an
anaerobic coccus is isolated in pure culture from a sterile source and is thought to
be contributing to infection, performance of AST should be considered.
The most common Gram-positive anaerobic cocci associated with human infec-
tions include Finegoldia magna, Parvimonas micra, Peptoniphilus harei, and
Peptoniphilus asaccharolyticus [13–15]. In general, many antimicrobials are
considered to be effective against these bacteria. Strains of P. asaccharolyticus,
F. magna, and P. micra are usually susceptible to the penicillins. A resistance survey
by Brazier and colleagues reported approximately 7% resistance to penicillin and
clindamycin for Gram-positive anaerobic cocci and 3.5% resistance to amoxicillin-­
clavulanate [16]. β-Lactam-β-lactamase inhibitor combinations and cephalosporins
are usually active [16, 17]. Carbapenems and metronidazole are highly active [17].
In contrast, the prevalence of clindamycin resistance among the Gram-positive
cocci approaches 30% in some surveys [18]. In another resistance survey that
included 113 Gram-positive anaerobic cocci, the highest rates of resistance for all
tested antimicrobials was to tetracycline (42%) [16].
Among the Gram-positive anaerobic cocci, F. magna stands out for its high rates
of resistance to penicillin and clindamycin. The sites of infection most commonly
affected by F. magna include skin and soft issue as well as bone and joints [10]. F.
magna also carries virulence factors that appear to be involved in bacterial
pathogenesis, among which are host cell adherence factors [19]. Resistance rates of
F. magna range between 10% and 20% for antimicrobials such as penicillin and
clindamycin [13, 16, 20, 21]. For example, Wren reported 16% and 8% resistance
to penicillin for F. magna and P. micra, respectively [14]. In another survey, the rate
of clindamycin resistance was 13% for 98 isolates of F. magna as compared to 1%
for Peptostreptococcus anaerobius (n = 92 isolates tested) and <1% for Parvimonas
micra (n = 146) [22]. Although much lower resistance rates to metronidazole are
seen for F. magna relative to clindamycin, significant metronidazole resistance rates
6 Anaerobic Bacteria: Antimicrobial Susceptibility Testing and Resistance Patterns 199

of 3% have been noted in some large surveys [22]. Although penicillin susceptibility
rates for Peptoniphilus asaccharolyticus and Anaerococcus prevotii are similar to
those of other anaerobic Gram-positive cocci, high MIC90 (e.g., MIC of 90% of the
isolates) values for clindamycin of >32 μg/mL have been noted with these species
[17, 23].
Peptostreptococcus anaerobius and P. stomatis are taxonomically distinct from
other anaerobic Gram-positive cocci. These species are generally associated with
female genital tract and intra-abdominal infections [10]. Resistance of P. anaerobius
to penicillin was notably high (7%) as compared to <1% for the other anaerobic
Gram-positive cocci assessed in one resistance survey [16]. Antimicrobial
susceptibility comparisons between P. anaerobius and P. stomatis demonstrate
higher resistance rates of the former to amoxicillin-clavulanate and consistently
higher MIC50 and MIC90 values for many other antimicrobials [24]. Thus, species
identification can be an important guide for therapy.
Peptoniphilus harei is another Gram-positive anaerobic coccus isolated from
infections of skin and soft tissue. It generally demonstrates low MICs to a variety of
antimicrobials [25]. This species can be difficult to identify phenotypically, since it
resembles P. asaccharolyticus biochemically, and in the past it has likely been
misidentified as P. asaccharolyticus [25]. Thus, past reports of AST results of P.
asaccharolyticus may have been limited by misidentification.
In general, Veillonella spp. are highly susceptible to several of the anti-anaerobic
agents, such as penicillins, β-lactam-β-lactamase inhibitor combinations,
carbapenems, clindamycin, and metronidazole. Indeed, no resistance to amoxicillin,
amoxicillin-clavulanate, clindamycin, or metronidazole was found in a recent
survey of Veillonella in the Netherlands [26].

6.3.3 Mechanisms of Resistance

While mechanisms of antimicrobial resistance among the anaerobic cocci have not
been extensively studied, Reig and colleagues postulated that penicillin-binding
proteins may account for the elevated MICs to penicillin that are occasionally seen
in Veillonella spp., because β-lactamases were not detected in penicillin-resistant
Veillonella isolates [27]. Likewise, penicillin resistance in Gram-positive cocci is
thought to be due to penicillin-binding proteins [28]. Resistance of anaerobes to
clindamycin is due to methylation of 23S rRNA at the site of drug action in the 50S
subunit of ribosomes [29]. The nitroimidazole nimB gene has been detected in a
large number of anaerobic Gram-positive cocci, but it is not always associated with
metronidazole resistance [30]. In a study by Theron and colleagues, the nimB gene
was present in 19/21 metronidazole-susceptible strains of Gram-positive cocci.
Thus, the presence of the nimB gene is not sufficient in and of itself for expression
of metronidazole resistance.
200 A. N. Schuetz

6.4 Gram-Positive, Non-Spore-Forming Bacilli

6.4.1 Clinical Disease and Taxonomic Changes

The “Eubacterium” group, Actinomyces, Cutibacterium, Propionibacterium,

Lactobacillus, Eggerthella lenta, and Bifidobacterium are included in this group of
bacteria. These bacteria are commensals of the skin and the mucocutaneous surfaces
of the oral cavity, the gastrointestinal tract, and the urogenital tract. Anaerobic
Gram-positive, non-spore-forming bacilli typically are associated with infections of
the head and neck that originate from the oral cavity [31]. Infections are frequently
polymicrobial in origin. Intra-abdominal and urogenital infections may develop due
to a break in the mucosal barrier. Actinomyces are particularly common oral
colonizers; infections may result from untreated dental caries, thus leading to
cervicofacial lesions and abscesses. Pulmonary actinomycosis may result from
aspiration of oral contents into the lungs. Pelvic actinomycosis has been associated
with the use of intrauterine contraception devices (IUDs). Cutibacterium species
cause a variety of infections, such as endocarditis, central nervous system (CNS)
infections (notably CNS shunt infections), and osteomyelitis [32]. Specifically, C.
acnes is associated with foreign-body infections of joints (particularly the shoulder
joint due to the large number of C. acnes normally present around the axillary area)
[31]. Clinical infections most commonly reported with Lactobacillus spp. include
bacteremia and endocarditis. The mouth is believed to be the route of entry of
lactobacilli into the bloodstream in patients with dental caries [33]. Reports of
Lactobacillus bacteremia associated with ingestion of probiotics containing
lactobacilli have been published [34]. Eubacterium spp. are present as normal flora
in the oral mucosa and can cause oral infections. Eggerthella lenta causes disease in
the intra-abdominal cavity [35]. Bifidobacterium spp. colonize the gut, but the most
common clinical disease due to this genus is dental caries. Other organisms included
in this group of Gram-positive, non-spore-forming anaerobic bacilli are Mobiluncus
and Atopobium; they are not frequently recovered in the clinical microbiology
laboratory.
The propionibacteria have recently undergone significant taxonomic changes.
Traditionally, species within the genus Propionibacterium have been informally
classified as either cutaneous or classic propionibacteria [36]. The “cutaneous
group” included P. acnes, P. avidum, and P. granulosum. A new member of the
cutaneous group, tentatively named P. humerusii, has been reported, but the work
has not yet been formally published [37]. The “classic group” of propionibacteria
include P. freudenreichii and P. propionicum, as well as other species, such as those
isolated from dairy products. In 2016, in an effort to address the issue of the
taxonomically challenging P. propionicum, Scholz and Kilian regrouped the
propionibacteria into three novel genera, Acidipropionibacterium, Cutibacterium,
and Pseudopropionibacterium, in addition to an amended description of the
genus Propionibacterium [38]. The new genus Cutibacterium now contains the
cutaneous species formerly known as P. acnes, P. avidum, P. granulosum, and
6 Anaerobic Bacteria: Antimicrobial Susceptibility Testing and Resistance Patterns 201

P. humerusii. Pseudopropionibacterium propionicum is the sole member of its

genus. The former “classic propionibacteria” group has been divided into the genera
Acidipropionibacterium and Propionibacterium. These name changes must be
taken into account when assessing reports regarding changes in susceptibility
patterns of the propionibacteria over time.

6.4.2 Resistance Patterns

The anaerobic non-spore-forming bacilli are usually susceptible to the β-lactams,

including the penicillins, cephalosporins, cephamycins, carbapenems, and β-lactam-­
β-lactamase inhibitor combinations [39–41]. Eggerthella lenta (formerly
Eubacterium lentum) is an exception: MICs of third-generation cephalosporins are
elevated with this species. Currently, no CLSI clinical breakpoints exist for
vancomycin with anaerobes; however, this agent shows good in vitro activity against
the “Eubacterium” group, Cutibacterium, Propionibacterium, and Actinomyces
[42]. Many of the non-spore-forming Gram-positive anaerobic bacilli, including
Actinomyces, Bifidobacterium, Cutibacterium, and some Lactobacillus species, are
intrinsically resistant to metronidazole. Occasionally, non-spore-forming Gram-­
positive bacilli demonstrate resistance to clindamycin [41]. For example, surveys
have shown approximately 7% resistance of C. acnes to clindamycin [5]. Telavancin
and moxifloxacin have shown good activity against many of the non-spore-forming
Gram-positive anaerobic bacilli [43, 44]. In summary, resistance to metronidazole
is typical for the anaerobic non-spore-forming Gram-positive bacilli, with most
isolates being susceptible to the β-lactams.
Most lactobacilli that grow well in ambient air are intrinsically vancomycin
resistant. Vancomycin-resistant lactobacilli include Lactobacillus casei, L.
rhamnosus, L. plantarum, L. salivarius, and L. fermentum. In contrast, the majority
of lactobacilli that grow only anaerobically demonstrate low MICs to vancomycin.
Among these are bacteria such as the L. acidophilus group, L. crispatus, L. gasseri,
L. johnsonii, and L. jensenii. In general, lactobacilli that demonstrate aerobic growth
should be tested for susceptibility aerobically as outlined in the CLSI M45 document
guidelines for infrequently isolated or fastidious bacteria [45]. Lactobacillus spp.
are generally reported as susceptible to clindamycin but with AST results that may
vary according to species (e.g., L. fermentum modal imipenem MIC is ≤0.03 μg/mL
[8/12 isolates]; L. rhamnosus imipenem MICs ranged from 0.25 to 4 μg/mL [modal
MIC 2 μg/mL]) [46]. Likewise, AST results differ according to the cephalosporin
tested (e.g., modal ceftriaxone MIC for L. rhamnosus isolates was ≥256 μg/mL
[17/22 isolates], while modal cefuroxime MIC for the same strains was 4 μg/mL
[15/22 isolates]). Notably, most surveys of Lactobacillus susceptibility have either
not addressed incubation conditions (e.g., aerobic vs. anaerobic) or have used
microaerobic atmospheres. Data on resistance mechanisms among anaerobic
lactobacilli are rare.
202 A. N. Schuetz

6.5 Gram-Positive, Spore-Forming Bacilli

6.5.1 Clostridioides difficile

6.5.1.1 Clinical Disease and Taxonomic Changes

Clostridium difficile has recently been renamed as Clostridioides difficile [47].

C. difficile is currently the most commonly reported pathogen causing healthcare-­
associated infections in US hospitals. C. difficile is also an important agent of
diarrheal illness in outpatients. Gastrointestinal disease associated with C. difficile
usually presents with a range of clinical findings, from diarrhea to pseudomembranous
colitis or even toxic megacolon. Disease is due to the ingestion of spores from the
environment, as the spores are resistant to alcohol-based gels and to many
disinfectants used in hospitals. When spores are transmitted by the fecal-oral route,
they can then germinate in the intestinal tract and produce toxins. The most
significant risk factor for acquisition of C. difficile infection (CDI) is antibiotic
exposure, presumably due to reduction of competing organisms. Although
clindamycin and broad-spectrum β-lactam antimicrobials are most often implicated
in CDI, any antimicrobial may lead to acquisition of CDI.

6.5.1.2 C. difficile Resistance Patterns

The most common antimicrobials used to treat CDI are metronidazole and vanco-
mycin. These agents are effective for most cases of CDI, although isolates with
elevated MICs to these antimicrobials have been reported [48]. It is important to
note, however, that the existing clinical breakpoints for C. difficile are based on
systemic infections that utilize systemic rather than intra-luminal (intraintestinal)
pharmacokinetics and pharmacodynamics of the antimicrobials. For example, the
levels of vancomycin orally administered to treat CDI are very high intraluminally
in the intestine (e.g., higher than 1000 μg/mL in feces), which is well above the MIC
resistant breakpoint of >2 μg/mL [49].
Antimicrobial resistance of C. difficile is important to monitor, not only for
patient treatment but also for the epidemiologic associations with certain ribotypes.
For example, the emergence of the hypervirulent C. difficile strain 027/BI/NAP has
been associated with consumption of fluoroquinolones [50]. Even though
fluoroquinolones are not used clinically to treat C. difficile infection, antimicrobial
pressure from widespread use of these agents is believed to have allowed the spread
of this fluoroquinolone-resistant, hypervirulent strain.
Resistance of C. difficile to metronidazole is reported to be relatively infrequent
in most areas of the world, but this conclusion depends upon which breakpoints are
used to interpret the data [51]. CLSI metronidazole breakpoints for C. difficile are
susceptible ≤8 μg/mL, intermediate 16 μg/mL, and resistant ≥32 μg/mL. EUCAST
6 Anaerobic Bacteria: Antimicrobial Susceptibility Testing and Resistance Patterns 203

breakpoints are lower: susceptible ≤2 μg/mL and resistant >2 μg/mL. In a US-based
survey of 925 isolates obtained in 2011 and 2012, 2.6% displayed resistance to
metronidazole based on the EUCAST breakpoint, while no isolate was resistant
based on the higher CLSI breakpoint [52].
Elevated MICs of C. difficile to vancomycin have also been reported. Most stud-
ies have applied the EUCAST vancomycin breakpoint (susceptible ≤2 μg/mL;
resistant >2 μg/mL), since CLSI has not set a clinical breakpoint for vancomycin
with anaerobes. Surveys have generally reported vancomycin resistance below 5%,
but higher percentages have been reported from smaller studies [52, 53]. Elevated
MICs to rifampin (>16 μg/mL) have also been reported, ranging from 8% to over
50% in different regions [53, 54]. There are no EUCAST or CLSI breakpoints for
rifampin. Fidaxomicin is also a common alternative agent used to treat antibiotic-­
refractory CDI or recurrent CDI. Decreased susceptibility to fidaxomicin is currently
rare, although one C. difficile isolate has been reported to have an MIC of 16 μg/
mL. This isolate was discovered in a clinical trial of fidaxomicin from a case of
recurrent CDI [55].
C. difficile has shown resistance to a variety of antimicrobials that include
clindamycin (8–100%), cephalosporins (51%), erythromycin (13–100%), and
fluoroquinolones (47%), based on 30 AST studies published between 2012 and
2015 [51]. Among the cephamycins, resistance to cefoxitin and cefotetan is
common, with at least 80% resistance in tested isolates [51]. In North America,
ribotype 027 is resistant to multiple antimicrobials, including rifampin, clindamycin,
and moxifloxacin [54]. Another hypervirulent strain, ribotype 078, also shows
resistance to ciprofloxacin, moxifloxacin, erythromycin, and imipenem [56]. The
most relevant statement is that elevated MICs to both metronidazole and vancomycin
have been reported among C. difficile, while elevated MICs to fidaxomicin are
relatively rare.

6.5.1.3 C. difficile Resistance Mechanisms

Resistance in C. difficile is associated with mobile genetic elements and chromo-

somal mutations [57]. Mobile genetic elements comprise a large proportion of the
C. difficile genome – approximately 11% [57]. Clindamycin and erythromycin
resistance is mediated by transposons, which are associated with the ermB gene and
MLSB-inducible resistance [51]. Resistance to the β-lactam antibiotics is due to
both penicillin-binding proteins and β-lactamase production [51]. Cfr gene
expression leads to linezolid resistance, while tet(M) expression leads to tetracycline
resistance [58]. Resistance to metronidazole and vancomycin are believed to be due
to alterations in targets or changes in the metabolic pathways of the organism [51].
Mutations in the rpoB gene are associated with reduced susceptibility to fidaxomicin
and rifampin [59, 60].
204 A. N. Schuetz

6.5.2 Clostridia Other Than C. difficile

6.5.2.1 Clinical Disease

Clostridia are commonly found in the environment in nature. Disease due to

Clostridium spp. other than C. difficile may be exogenous or endogenous in origin.
Exogenous clostridial infections include diseases such as tetanus due to C. tetani,
foodborne botulism due to C. botulinum, and wound infections leading to gangrene
or myonecrosis due to a variety of clostridia. Endogenous infections associated with
clostridia are usually due to breaches in the mucosal barrier (e.g., mucosa of the
intestine or respiratory tract) that lead to abscesses, lung empyemas, aspiration
pneumonia, or intra-abdominal infections. Risk factors for clostridial disease
include surgical procedures, dirty wounds, malignancy, and diabetes mellitus [61].
Most clostridia isolated in the clinical microbiology laboratory are obtained from
bloodstream infections, wounds, or intra-abdominal sources. C. septicum and
C. perfringens are two of the most common clostridia that cause bacteremia. In a
Canadian study of Clostridium-associated bacteremia, the most common species
identified was C. perfringens (58/138, 42%), followed by C. septicum (19/138;
14%) [62]. Since C. septicum is commonly associated with occult gastrointestinal
malignancy, the isolation of this bacterium from blood should prompt a study of the
gastrointestinal tract. Other clostridial species isolated in cases of bacteremia
include C. tertium and C. sordellii. The most common clostridia associated with
skin and soft tissue infections include C. perfringens, C. septicum, C. histolyticum,
and C. sordellii. Spontaneous myonecrosis is often associated with C. septicum,
while traumatic gas gangrene is usually associated with C. perfringens. C. sordellii
is known for its historical association with gynecologic infections, especially medi-
cally induced abortions. Antimicrobial susceptibility patterns of these species are
listed below.

6.5.2.2 Resistance Patterns

Many of the most common species of non-C. difficile clostridia isolated from human
infections are highly susceptible to antimicrobials having anti-anaerobic activity
(e.g., metronidazole, clindamycin, carbapenems, and piperacillin-tazobactam).
Among the non-C. difficile clostridia species, the “RIC” group – namely, C. ramo-
sum, C. innocuum, and C. clostridioforme – demonstrates the highest resistance
rates. C. clostridioforme produces β-lactamase and thus is resistant to several
β-lactam antibiotics but is susceptible to vancomycin [63, 64]. C. innocuum is resis-
tant to cefoxitin and cefotetan, and it displays high MICs (8–32 μg/mL) to vanco-
mycin [65]. Fortunately, C. innocuum is susceptible to metronidazole. Finally,
C. ramosum also demonstrates high vancomycin MICs. Many species of clostridia
have shown resistance to clindamycin, including C. perfringens, C. ramosum,
C. tertium, and C. sporogenes [66]. In general, clindamycin resistance among clos-
tridia ranges from 10% to 20%, but it varies by species [4]. Metronidazole,
6 Anaerobic Bacteria: Antimicrobial Susceptibility Testing and Resistance Patterns 205

carbapenems, and β-lactam-β-lactamase inhibitor combinations have maintained

activity against most clostridial species [66, 67]. Resistance to metronidazole is
likewise low, while resistance to moxifloxacin averages around 20% in some sur-
veys [5].
C. perfringens isolates are highly susceptible to penicillin and clindamycin.
However, recent surveys, such as that by Hastey et al., demonstrate rising resistance
of C. perfringens over time to clindamycin, from no resistance noted in 2007–2009
to 7% resistance noted in 2010–2012 [5]. In another clostridia resistance survey of
blood isolates from 2000 to 2006, 8/58 (14%) C. perfringens isolates were resistant
to clindamycin [62]. All C. perfringens isolates in that study were susceptible to
penicillin and metronidazole.

6.5.2.3 Mechanisms of Resistance

The “RIC” group of clostridia, as well as C. butyricum, produces β-lactamases and

rarely penicillin-binding proteins that confer resistance to the β-lactams, including
the cephamycins and cephalosporins [29, 66]. The β-lactamases are chromosomally
encoded and are released extracellularly [29]. Resistance to clindamycin is caused
by methylation of the 23S ribosomal RNA subunit, leading to modification of the
site of drug action [29].

6.6 Gram-Negative Anaerobic Bacilli

6.6.1 Bacteroides fragilis Group

6.6.1.1 Clinical Disease and Taxonomic Changes

The B. fragilis group has historically been comprised of a variety of Bacteroides

spp. including B. fragilis, B. thetaiotaomicron, B. ovatus, and B. vulgatus. The B.
fragilis group was originally conceptualized as the particular bile-resistant
Bacteroides species. However, this grouping is “informal,” in that it is not formally
recognized as a taxonomic group. Therefore, some authors refer to the B. fragilis
group as the “former B. fragilis group” and prefer to refer to the specific species of
Bacteroides instead of aggregating together the various species. There are other
Bacteroides species, such as B. caccae, that are not grouped with the B. fragilis
group. Likewise, Parabacteroides distasonis, previously known as Bacteroides
distasonis, had been placed within the B. fragilis group. Therefore, it is discussed
with the B. fragilis group below.
Bacteroides spp. are some of the most frequently encountered anaerobes in the
clinical microbiology laboratory and are among the most virulent. Bacteroides spp.
are also typically more resistant than other anaerobes to antimicrobial agents. They
are the most common anaerobes isolated from blood, abscesses (particularly intra-­
206 A. N. Schuetz

abdominal), bone, and skin lesions [68]. Although usually recovered from abscesses
as mixtures with other anaerobes and aerobes, Bacteroides spp. are usually recovered
from blood as the only pathogen present.

6.6.1.2 Resistance Patterns

Of the Bacteroides species, B. fragilis is generally one of the most susceptible to

antimicrobials. A susceptibility survey of various B. fragilis group isolates collected
from ten US medical centers from 1997 to 2004 demonstrated differences in the
prevalence of susceptibility across various species [69]. Compared to other members
of the B. fragilis group, B. ovatus and B. thetaiotaomicron displayed higher MICs to
ertapenem [69]. B. vulgatus, P. distasonis, and B. thetaiotaomicron also displayed
higher MICs than B. fragilis to piperacillin-tazobactam. P. distasonis demonstrates
high MICs to many of the β-lactams, excluding carbapenems. B. thetaiotaomicron
is also known for higher rates of resistance to several antimicrobials as compared to
other members of the former B. fragilis group. Decreasing susceptibility of the B.
fragilis group to these frequently used antimicrobials is concerning.
β-Lactam-β-lactamase inhibitor combinations are highly active against many
strains of the former B. fragilis group, with the exception of P. distasonis [4]. Recent
worldwide studies have reported increasing resistance to ampicillin-sulbactam and
piperacillin-tazobactam in Bacteroides species [70]. One survey of 1580 isolates
demonstrated a significant increase in resistance to piperacillin-tazobactam from
2% in 2007–2009 to 7% in 2010–2012 [5]. During this time period, susceptibility to
ampicillin-sulbactam also significantly decreased, from 86% in 2007–2009 to 82%
in 2010–2012.
Cefoxitin and cefotetan are generally active against members of the former
B. fragilis group. Although some surveys demonstrate slight decreases in resistance
of the B. fragilis group to cefoxitin, other surveys show that rates of resistance to
cefoxitin and cefotetan are rising in some areas of the world [5, 71], implying that
susceptibility testing should be performed if these antimicrobials are to be used for
therapy. Broad-spectrum cephalosporins show poor activity against the B. fragilis
group organisms [72].
Carbapenems are also generally active against members of the B. fragilis group.
Most studies report susceptibility rates of ≥95% and usually well above 99%
susceptibility [69]. However, carbapenem resistance is increasing, albeit slowly,
from zero resistance noted in 2007–2009 to 1% resistance in 2010–2012 [5]. Seifert
et al. reported 95% carbapenem susceptibility for B. fragilis group isolates associated
with intra-abdominal infections in a multicenter German study [73].
Resistance to clindamycin and moxifloxacin appears to be increasing.
Clindamycin resistance is increasing worldwide and has reached approximately
40% for Bacteroides group isolates [4]. Rates of susceptibility to moxifloxacin are
variable among species, with approximately 31% and 43% of B. fragilis and B.
ovatus isolates demonstrating resistance, respectively [69]. In another study,
Snydman et al. reported an overall resistance rate of >80% for moxifloxacin for
6 Anaerobic Bacteria: Antimicrobial Susceptibility Testing and Resistance Patterns 207

B. fragilis group isolates [74]. Bacteroides isolates appear to be acquiring resistance

to moxifloxacin, as demonstrated by a multicenter European surveillance study in
which susceptibility rates to moxifloxacin decreased between 2003 and 2009 from
91 to 86.4% [71].
Metronidazole-resistant strains of B. fragilis group isolates are rare, but they
have been reported. In 2011, Sherwood and colleagues published a case report
concerning a multidrug resistant strain of B. fragilis from a leg wound [75]. The
isolate possessed the nim gene, which was located on a plasmid and conferred
resistance to metronidazole. The isolate did, however, display low MICs to
moxifloxacin and linezolid. Overall, approximately 95–98% of B. fragilis isolates
are susceptible to metronidazole [5].

6.6.1.3 Mechanisms of Resistance

Close to 97% of all Bacteroides isolates are resistant to penicillin and ampicillin,
principally due to β-lactamase production. However, penicillin-binding proteins can
also be expressed by strains [76]. Penicillin and cephalosporin resistance is mediated
primarily by cepA or cfxA [77]. A chromosomal cephalosporinase is encoded by
cepA, which leads to cephalosporin and aminopenicillin resistance. However,
isolates expressing cepA may remain susceptible to piperacillin and β-lactam-β-­
lactamase inhibitor combinations. Porin losses can also lead to increases in β-lactam
MICs, including MICs to the β-lactam-β-lactamase inhibitor combinations, as
demonstrated by one group which reported amoxicillin-clavulanate resistance in a
B. thetaiotaomicron strain [78]. The cfxA gene encodes a broad-spectrum
β-lactamase that is responsible for loss of susceptibility to cefoxitin and several
other β-lactam drugs [79, 80]. Carbapenem resistance is often mediated by a zinc
metallo-β-lactamase, encoded by cfiA, that confers resistance to β-lactams and
β-lactam-β-lactamase inhibitor combinations [75]. cfiA in Bacteroides is silent
unless an insertion sequence element activates it [81]. A report from Turkey noted
10% resistance to carbapenems in B. fragilis group isolates, many of which were
confirmed to carry cfiA [82]. Resistance to clindamycin is mediated by erm, which
is located on transferable plasmids [83]. Metronidazole resistance is linked to nim
(nitroimidazole reductase). Expression of this gene leads to reduction of the nitrate
moiety of metronidazole to an amino derivative, which decreases effectiveness of
the antibiotic. However, the mere presence of nim is not predictive of metronidazole
resistance, as nim can be found in metronidazole-susceptible strains. Of 206 B.
fragilis isolates in one study, nim was detected in 24%, and metronidazole MICs of
the nim-positive isolates ranged from 1.5 (susceptible) to >256 (resistant) μg/mL
[84]. Although many nim-positive Bacteroides strains do not show a loss of
susceptibility, they can be induced to express high MICs to metronidazole by
subinhibitory concentrations of the antibiotic [85]. Resistance to moxifloxacin is
mediated by gyrA mutations and efflux pumps. In summary, there are many different
types of resistance genes that may be carried and expressed by B. fragilis group
isolates.
208 A. N. Schuetz

6.6.2  ram-Negative Anaerobic Bacilli Other Than the B.

G
fragilis Group
6.6.2.1 Clinical Disease

Gram-negative anaerobic bacilli other than the B. fragilis group include

Porphyromonas, Prevotella, Fusobacterium, and Bilophila wadsworthia, among
others. Porphyromonas spp. are known components of oral flora and may cause
significant oral and periodontal disease. Porphyromonas may also be isolated from
intra-abdominal sites and the female genital tract. Prevotella spp. are also
predominant in the oral cavity. Prevotella may be implicated in human bite wounds
and periodontal infections. The two most commonly recovered fusobacteria include
Fusobacterium nucleatum and F. necrophorum. Lemierre syndrome (e.g., septic
thrombophlebitis of the internal jugular vein) is most often caused by F. necrophorum,
but it may be caused by other anaerobes or aerobes. Fusobacterium spp. are also
recovered from blood in cases of sepsis, and they are associated with abscesses in
other sites of the body. B. wadsworthia is a significant cause of polymicrobial intra-­
abdominal infections.

6.6.2.2 Resistance Patterns

Prevotella and Porphyromonas are more susceptible to antimicrobials used to treat

anaerobic infections than the B. fragilis group. Prevotella tends to be less susceptible
to a variety of antimicrobials than Porphyromonas. Resistance to penicillin and
ampicillin is found in approximately three-fourths of Prevotella isolates [86]. In a
New Zealand surveillance study covering 1999 to 2003, susceptibility rates among
45 isolates of Prevotella spp. showed only 18% susceptibility to penicillin but 96%
susceptibility to clindamycin [86]. Fewer than 5% of Prevotella are resistant to
cefoxitin. Liu and colleagues reported a small decrease in susceptibility of Prevotella
spp. over time to imipenem; their isolates demonstrated 100% susceptibility in 2002
but decreased to 94% in 2006 [87].
Antimicrobial surveillance studies of clinically significant Porphyromonas spp.
are rare. Authors of one study used the Etest method to assess susceptibility of a
variety of anaerobes and reported 94% susceptibility for 45 Porphyromonas isolates
to penicillin and 97% susceptibility to cefoxitin [88]. In addition, Porphyromonas
demonstrated 100% susceptibility to ampicillin-sulbactam, imipenem, clindamycin,
and metronidazole. Production of β-lactamase in Porphyromonas is rare. Prevotella
and Porphyromonas are almost uniformly susceptible to carbapenems,
metronidazole, and tigecycline.
Resistance among fusobacteria is unusual, and susceptibility rates may vary by
species. Penicillin resistance among Fusobacterium spp. is relatively rare. In a
Taiwanese survey including 36 Fusobacterium isolates, only 4 (11%) were positive
for β-lactamase [87]. All Fusobacterium spp. were susceptible to metronidazole,
6 Anaerobic Bacteria: Antimicrobial Susceptibility Testing and Resistance Patterns 209

piperacillin-tazobactam, and ertapenem, while resistance to moxifloxacin was

demonstrated in approximately 20% of isolates. While F. mortiferum may be
resistant to cephalosporins, over 90% of F. nucleatum and F. necrophorum isolates
are susceptible to cephalosporins and cephamycins [22].
Campylobacter gracilis is susceptible to most antimicrobial agents, including
β-lactam-β-lactamase inhibitor combinations, cefoxitin, and clindamycin [89].
Sutterella wadsworthensis may be resistant to clindamycin, piperacillin, or
metronidazole [42]. Bilophila wadsworthia is usually resistant to penicillin, but it is
highly susceptible to other antibiotics, such as amoxicillin-clavulanate, clindamycin,
carbapenems, and metronidazole, that are used to treat anaerobic infections.

6.6.2.3 Mechanisms of Resistance

Most Prevotella spp. are resistant to penicillin and ampicillin due to β-lactamase
production. When penicillin resistance is present in Porphyromonas, Fusobacterium,
or B. wadsworthia, it is usually as a result of β-lactamases [90]. Most β-lactamases
in the Gram-negative anaerobic bacilli other than the B. fragilis group are penicil-
linases, although cephalosporinases have been described in Prevotella spp. [29].

6.7 Concluding Remarks

Although anaerobic bacteria cause serious diseases, they have not received as much
attention as their aerobic counterparts, in part because they are more difficult to
cultivate. In addition, anaerobes are often present in polymicrobial aerobic/anaerobic
infections or mixed with other anaerobes; therefore, their exact role in contributing
toward infection can sometimes be difficult to ascertain. Some antimicrobials are
effective against both aerobic and anaerobic bacteria, but certain antimicrobials,
such as metronidazole, are only effective against anaerobes. Since many of the
anaerobes tend to grow slowly in culture, susceptibility testing is slow. For
anaerobes, pathogen identification is often obtained well before antimicrobial
susceptibility test results are available. Thus, the anaerobe antibiogram can be
particularly useful in guiding appropriate empirical choice of antimicrobial. It has
been shown that AST results vary according to the type of assay used; thus, future
anaerobic antimicrobial surveys should clearly present the methodologic method of
testing and the breakpoint cutoff applied. Clear delineation of data will aid data
interpretation. Ideally, authors should also present AST data on different species
separately, rather than lumping together groups of morphologically similar
organisms (e.g., anaerobic cocci).
In addition to the technical issues discussed above, aerobic and anaerobic patho-
gens may differ in virulence. Such activity is not reflected in susceptibility (MIC)
measurements due to the nature of the assays, but it is likely to be clinically impor-
210 A. N. Schuetz

tant for clearing infection. It has been proposed that with aerobic bacteria, reactive
oxygen species contribute to the lethal activity of aerobes (see Chap. 20). These
toxic molecules are not expected to be present in anaerobes. Thus, as resistance
emerges to carbapenems and metronidazole, for example, it may be necessary to
develop more agents that are specifically lethal for anaerobes. Such agents could
have additional utility against both aerobic and anaerobic organisms.
Major Points
• With the increasing use of MALDI-TOF MS to identify anaerobic bacteria in
clinical microbiology laboratories, more accurate species identification and anti-
microbial susceptibility results are becoming available. More accurate differen-
tiation and characterization of strains provide more meaningful AST results.
• When reviewing anaerobic AST data, it is important to be aware of the method
of AST utilized in generating the data and the specific breakpoints applied, since
various methods and breakpoints affect interpretation of data.
• A unified AST testing approach for clinical laboratories needs to be developed
for anaerobes. For example, broth microdilution testing is recommended only for
the B. fragilis group. Agar dilution testing can be performed on all anaerobes,
including the B. fragilis group.
• It is important to recognize that the data presented in the anaerobe antibiogram
from CLSI M100 document [4] are derived from isolates obtained globally and
may not be applicable to isolates within a given hospital or region.
• Taxonomic name changes must be taken into account when assessing antimicro-
bial susceptibility testing results of many anaerobes including the anaerobic
Gram-­positive cocci, the propionibacteria, and the clostridia.
• Bacteroides spp. are some of the most frequently encountered anaerobes in the
clinical microbiology laboratory and are also among the most virulent. They are
typically more resistant than other anaerobes to antimicrobial agents.
• Certain patterns of resistance in anaerobes that are on the rise include resistance
to the β-lactam-β-lactamase inhibitor combinations among the B. fragilis group,
increasing clindamycin resistance among all anaerobes, and resistance of
Clostridium to vancomycin. In addition, metronidazole resistance is no longer
limited to the B. fragilis group, as it now includes Gram-positive cocci and
bacilli. Although the prevalence of resistance to certain antimicrobials is rising,
effective antimicrobials are generally available.

References

1. Brook I, Wexler HM, Goldstein EJ. Antianaerobic antimicrobials: spectrum and susceptibility
testing. Clin Microbiol Rev. 2013;26(3):526–46.
2. CLSI. Analysis and presentation of cumulative antimicrobial susceptibility test data; approved
guideline, 4th ed. CLSI M39-A4. Wayne, PA: Clinical and Laboratory Standards Institute;
2014.
6 Anaerobic Bacteria: Antimicrobial Susceptibility Testing and Resistance Patterns 211

3. CLSI. Methods for antimicrobial susceptibility testing of anaerobic bacteria; approved stan-
dard, 8th ed. CLSI M11-A8. Wayne, PA: Clinical and Laboratory Standards Institute; 2012.
4. CLSI. Performance standards for antimicrobial susceptibility testing. 27th ed. CLSI supple-
ment M100. Wayne, PA: Clinical and Laboratory Standards Institute; 2017.
5. Hastey CJ, Boyd H, Schuetz AN, Anderson K, Citron DM, Dzink-Fox J, et al. Changes in the
antibiotic susceptibility of anaerobic bacteria from 2007–2009 to 2010–2012 based on the
CLSI methodology. Anaerobe. 2016;42:27–30.
6. Poulet PP, Duffaut D, Lodter JP. Evaluation of the Etest for determining the in-vitro sus-
ceptibilities of Prevotella intermedia isolates to metronidazole. J Antimicrob Chemother.
1999;43(4):610–1.
7. Nagy E, Justesen US, Eitel Z, Urban E, Infection ESGoA. Development of EUCAST disk dif-
fusion method for susceptibility testing of the Bacteroides fragilis group isolates. Anaerobe.
2015;31:65–71.
8. Hastey CJ, Dale SE, Nary J, Citron D, Law JH, Roe-Carpenter DE, et al. Comparison of
Clostridium difficile minimum inhibitory concentrations obtained using agar dilution vs broth
microdilution methods. Anaerobe. 2017;44:73–7.
9. Munson E, Carroll KC. What’s in a name? New bacterial species and changes to taxonomic
status from 2012 through 2015. J Clin Microbiol. 2017;55(1):24–42.
10. Song Y, Finegold SM. Peptostreptococcus, Finegoldia, Anaerococcus, Peptoniphilus,
Veillonella, and other anaerobic cocci. In: Jorgensen JH, Pfaller MA, editors. Manual of clini-
cal microbiology. 11th ed. Washington, DC: ASM Press; 2015. p. 909–39.
11. Murdoch DA. Gram-positive anaerobic cocci. Clin Microbiol Rev. 1998;11(1):81–120.
12. Brook I. Recovery of anaerobic bacteria from clinical specimens in 12 years at two military
hospitals. J Clin Microbiol. 1988;26(6):1181–8.
13. Brazier J, Chmelar D, Dubreuil L, Feierl G, Hedberg M, Kalenic S, et al. European surveil-
lance study on antimicrobial susceptibility of Gram-positive anaerobic cocci. Int J Antimicrob
Agents. 2008;31(4):316–20.
14. Wren MW. Anaerobic cocci of clinical importance. Br J Biomed Sci. 1996;53(4):294–301.
15. Wildeboer-Veloo AC, Harmsen HJ, Welling GW, Degener JE. Development of 16S rRNA-­
based probes for the identification of Gram-positive anaerobic cocci isolated from human
clinical specimens. Clin Microbiol Infect. 2007;13(10):985–92.
16. Brazier JS, Hall V, Morris TE, Gal M, Duerden BI. Antibiotic susceptibilities of Gram-positive
anaerobic cocci: results of a sentinel study in England and Wales. J Antimicrob Chemother.
2003;52(2):224–8.
17. Goldstein EJ, Citron DM, Warren YA, Tyrrell KL, Merriam CV, Fernandez HT. In vitro activities
of dalbavancin and 12 other agents against 329 aerobic and anaerobic Gram-positive isolates
recovered from diabetic foot infections. Antimicrob Agents Chemother. 2006;50(8):2875–9.
18. Mory F, Lozniewski A, Bland S, Sedallian A, Grollier G, Girard-Pipau F, et al. Survey of
anaerobic susceptibility patterns: a French multicentre study. Int J Antimicrob Agents.
1998;10(3):229–36.
19. Goto T, Yamashita A, Hirakawa H, Matsutani M, Todo K, Ohshima K, et al. Complete
genome sequence of Finegoldia magna, an anaerobic opportunistic pathogen. DNA Res.
2008;15(1):39–47.
20. Aldridge KE, Ashcraft D, Cambre K, Pierson CL, Jenkins SG, Rosenblatt JE. Multicenter
survey of the changing in vitro antimicrobial susceptibilities of clinical isolates of Bacteroides
fragilis group, Prevotella, Fusobacterium, Porphyromonas, and Peptostreptococcus species.
Antimicrob Agents Chemother. 2001;45(4):1238–43.
21. Hawser SP. Activity of tigecycline and comparators against recent clinical isolates of
Finegoldia magna from Europe. Eur J Clin Microbiol Infect Dis. 2010;29(8):1011–3.
22. Koeth LM, Good CE, Appelbaum PC, Goldstein EJ, Rodloff AC, Claros M, et al.
Surveillance of susceptibility patterns in 1297 European and US anaerobic and capnophilic
isolates to co-amoxiclav and five other antimicrobial agents. J Antimicrob Chemother.
2004;53(6):1039–44.
212 A. N. Schuetz

23. Citron DM, Kwok YY, Appleman MD. In vitro activity of oritavancin (LY333328), vanco-
mycin, clindamycin, and metronidazole against Clostridium perfringens, Propionibacterium
acnes, and anaerobic Gram-positive cocci. Anaerobe. 2005;11(1–2):93–5.
24. Kononen E, Bryk A, Niemi P, Kanervo-Nordstrom A. Antimicrobial susceptibilities of
Peptostreptococcus anaerobius and the newly described Peptostreptococcus stomatis isolated
from various human sources. Antimicrob Agents Chemother. 2007;51(6):2205–7.
25. Veloo AC, Welling GW, Degener JE. Antimicrobial susceptibility of clinically relevant Gram-­
positive anaerobic cocci collected over a three-year period in the Netherlands. Antimicrob
Agents Chemother. 2011;55(3):1199–203.
26. Veloo AC, van Winkelhoff AJ. Antibiotic susceptibility profiles of anaerobic pathogens in the
Netherlands. Anaerobe. 2015;31:19–24.
27. Reig M, Mir N, Baquero F. Penicillin resistance in Veillonella. Antimicrob Agents Chemother.
1997;41(5):1210.
28. Reig M, Baquero F. Antibacterial activity of clavulanate and tazobactam on Peptostreptococcus
spp. J Antimicrob Chemother. 1994;33(2):358–9.
29. Garcia-Rodriguez JA, Garcia-Sanchez JE, Munoz-Bellido JL. Antimicrobial resistance in
anaerobic bacteria: current situation. Anaerobe. 1995;1(2):69–80.
30. Theron MM, Janse Van Rensburg MN, Chalkley LJ. Nitroimidazole resistance genes (nimB) in
anaerobic Gram-positive cocci (previously Peptostreptococcus spp.). J Antimicrob Chemother.
2004;54(1):240–2.
31. Hall V, Copsey SD. Propionibacterium, Lactobacillus, Actinomyces, and other non-spore-­
forming anaerobic Gram-positive rods. In: Jorgensen JH, Pfaller MA, editors. Manual of clini-
cal microbiology. 11th ed. Washington, DC: ASM Press; 2015. p. 920–39.
32. Siqueira JF Jr, Rocas IN. Polymerase chain reaction detection of Propionibacterium propioni-
cus and Actinomyces radicidentis in primary and persistent endodontic infections. Oral Surg
Oral Med Oral Pathol Oral Radiol Endod. 2003;96(2):215–22.
33. Cannon JP, Lee TA, Bolanos JT, Danziger LH. Pathogenic relevance of Lactobacillus: a retro-
spective review of over 200 cases. Eur J Clin Microbiol Infect Dis. 2005;24(1):31–40.
34. Land MH, Rouster-Stevens K, Woods CR, Cannon ML, Cnota J, Shetty AK. Lactobacillus
sepsis associated with probiotic therapy. Pediatrics. 2005;115(1):178–81.
35. Rautio M, Saxen H, Siitonen A, Nikku R, Jousimies-Somer H. Bacteriology of histopathologi-
cally defined appendicitis in children. Pediatr Infect Dis J. 2000;19(11):1078–83.
36. Stackebrandt E, Cummins CS, Johnson JL. Family Propionibacteriaceae: the genus
Propionibacterium. In: Dworkin M, Falkow S, Rosenberg E, Schleifer K, Stackebrandt E, edi-
tors. The prokaryotes. New York: Springer; 2006. p. 400–18.
37. Butler-Wu SM, Sengupta DJ, Kittichotirat W, Matsen FA 3rd, Bumgarner RE. Genome
sequence of a novel species, Propionibacterium humerusii. J Bacteriol. 2011;193(14):3678.37.
38. Scholz CF, Kilian M. The natural history of cutaneous propionibacteria, and reclassifica-
tion of selected species within the genus Propionibacterium to the proposed novel genera
Acidipropionibacterium gen. nov., Cutibacterium gen. nov. and Pseudopropionibacterium gen.
nov. Int J Syst Evol Microbiol. 2016;66(11):4422–32.
39. Oprica C, Nord CE, Bacteria ESGoARiA. European surveillance study on the antibiotic sus-
ceptibility of Propionibacterium acnes. Clin Microbiol Infect. 2005;11(3):204–13.
40. Tyrrell KL, Citron DM, Warren YA, Fernandez HT, Merriam CV, Goldstein EJ. In vitro
activities of daptomycin, vancomycin, and penicillin against Clostridium difficile, C.
perfringens, Finegoldia magna, and Propionibacterium acnes. Antimicrob Agents Chemother.
2006;50(8):2728–31.
41. Smith AJ, Hall V, Thakker B, Gemmell CG. Antimicrobial susceptibility testing of Actinomyces
species with 12 antimicrobial agents. J Antimicrob Chemother. 2005;56(2):407–9.
42. Goldstein EJC, Citron DM. Resistance trends in antimicrobial susceptibility of anaerobic bac-
teria, part I. Clin Microbiol Newsl. 2011;33(1):1–8.
43. Goldstein EJ, Citron DM, Merriam CV, Warren YA, Tyrrell KL, Fernandez HT. In vitro
activities of the new semisynthetic glycopeptide telavancin (TD-6424), vancomycin, dapto-
6 Anaerobic Bacteria: Antimicrobial Susceptibility Testing and Resistance Patterns 213

mycin, linezolid, and four comparator agents against anaerobic Gram-positive species and
Corynebacterium spp. Antimicrob Agents Chemother. 2004;48(6):2149–52.
44. Edmiston CE, Krepel CJ, Seabrook GR, Somberg LR, Nakeeb A, Cambria RA, et al. In
vitro activities of moxifloxacin against 900 aerobic and anaerobic surgical isolates from
patients with intra-abdominal and diabetic foot infections. Antimicrob Agents Chemother.
2004;48(3):1012–6.44.
45. CLSI, editor. Methods for antimicrobial dilution and disk susceptibility testing of infre-
quently isolated or fastidious bacteria. 3rd ed. CLSI guideline M45. Wayne, PA: Clinical and
Laboratory Standards Institute; 2015.
46. Salminen MK, Rautelin H, Tynkkynen S, Poussa T, Saxelin M, Valtonen V, et al. Lactobacillus
bacteremia, species identification, and antimicrobial susceptibility of 85 blood isolates. Clin
Infect Dis. 2006;42(5):e35–44.
47. Lawson PA, Citron DM, Tyrrell KL, Finegold SM. Reclassification of Clostridium difficile as
Clostridioides difficile (Hall and O'Toole 1935) Prévot 1938. Anaerobe. 2016 Aug;40:95–9.
48. Goudarzi M, Goudarzi H, Alebouyeh M, Azimi Rad M, Shayegan Mehr FS, Zali MR, et al.
Antimicrobial susceptibility of clostridium difficile clinical isolates in Iran. Iran Red Crescent
Med J. 2013;15(8):704–11.
49. Baines SD, Wilcox MH. Antimicrobial resistance and reduced susceptibility in Clostridium
difficile: potential consequences for induction, treatment, and recurrence of C. difficile infec-
tion. Antibiotics (Basel). 2015;4(3):267–98.
50. He M, Miyajima F, Roberts P, Ellison L, Pickard DJ, Martin MJ, et al. Emergence and global
spread of epidemic healthcare-associated Clostridium difficile. Nat Genet. 2013;45(1):109–13.
51. Spigaglia P. Recent advances in the understanding of antibiotic resistance in Clostridium dif-
ficile infection. Ther Adv Infect Dis. 2016;3(1):23–42.
52. Snydman DR, McDermott LA, Jacobus NV, Thorpe C, Stone S, Jenkins SG, et al. U.S.-Based
National Sentinel Surveillance Study for the epidemiology of Clostridium difficile-associated
diarrheal isolates and their susceptibility to fidaxomicin. Antimicrob Agents Chemother.
2015;59(10):6437–43.
53. Freeman J, Vernon J, Morris K, Nicholson S, Todhunter S, Longshaw C, et al. Pan-European
longitudinal surveillance of antibiotic resistance among prevalent Clostridium difficile ribo-
types. Clin Microbiol Infect. 2015;21(3):248 e9–e16.
54. Tenover FC, Tickler IA, Persing DH. Antimicrobial-resistant strains of Clostridium difficile
from North America. Antimicrob Agents Chemother. 2012;56(6):2929–32.
55. Goldstein EJ, Babakhani F, Citron DM. Antimicrobial activities of fidaxomicin. Clin Infect
Dis. 2012;55(Suppl 2):S143–8.
56. Keessen EC, Hensgens MP, Spigaglia P, Barbanti F, Sanders IM, Kuijper EJ, et al. Antimicrobial
susceptibility profiles of human and piglet Clostridium difficile PCR-ribotype 078. Antimicrob
Resist Infect Control. 2013;2:14.
57. Peng Z, Jin D, Kim HB, Stratton CW, Wu B, Tang YW, et al. Update on antimicrobial resis-
tance in Clostridium difficile: resistance mechanisms and antimicrobial susceptibility testing.
J Clin Microbiol. 2017;55(7):1998–2008.
58. Shen J, Wang Y, Schwarz S. Presence and dissemination of the multiresistance gene cfr in
Gram-positive and Gram-negative bacteria. J Antimicrob Chemother. 2013;68(8):1697–706.
59. O'Connor JR, Galang MA, Sambol SP, Hecht DW, Vedantam G, Gerding DN, et al. Rifampin
and rifaximin resistance in clinical isolates of Clostridium difficile. Antimicrob Agents
Chemother. 2008;52(8):2813–7.
60. Leeds JA, Sachdeva M, Mullin S, Barnes SW, Ruzin A. In vitro selection, via serial passage,
of Clostridium difficile mutants with reduced susceptibility to fidaxomicin or vancomycin.
J Antimicrob Chemother. 2014;69(1):41–4.
61. Stevens DL, Bryant AE, Carroll KC. Clostridium. In: Jorgensen JH, Pfaller MA, editors.
Manual of clinical microbiology. 11th ed. Washington, DC: ASM Press; 2015. p. 940–66.
62. Leal J, Gregson DB, Ross T, Church DL, Laupland KB. Epidemiology of Clostridium species
bacteremia in Calgary, Canada, 2000-2006. J Infect. 2008;57(3):198–203.
214 A. N. Schuetz

63. Dubreuil L, Odou MF. Anaerobic bacteria and antibiotics: what kind of unexpected resistance
could I find in my laboratory tomorrow? Anaerobe. 2010;16(6):555–9.
64. Finegold SM, Song Y, Liu C, Hecht DW, Summanen P, Kononen E, et al. Clostridium clos-
tridioforme: a mixture of three clinically important species. Eur J Clin Microbiol Infect Dis.
2005;24(5):319–24.
65. Finegold SM, Molitoris D, Vaisanen ML. Study of the in vitro activities of rifaximin and com-
parator agents against 536 anaerobic intestinal bacteria from the perspective of potential utility
in pathology involving bowel flora. Antimicrob Agents Chemother. 2009;53(1):281–6.
66. Winn WC, Allen SD, Janda WM, Koneman EW, Procop G, Schrechenberger PC, et al.
Koneman’s color atlas and textbook of diagnostic microbiology. Philadelphia: Lippincott
Williiams & Wilkins; 2005.
67. Ackermann G, Schaumann R, Pless B, Claros MC, Goldstein EJ, Rodloff AC. Comparative
activity of moxifloxacin in vitro against obligately anaerobic bacteria. Eur J Clin Microbiol
Infect Dis. 2000;19(3):228–32.
68. Wexler HM. Bacteroides: the good, the bad, and the nitty-gritty. Clin Microbiol Rev.
2007;20(4):593–621.
69. Snydman DR, Jacobus NV, McDermott LA, Golan Y, Hecht DW, Goldstein EJ, et al. Lessons
learned from the anaerobe survey: historical perspective and review of the most recent data
(2005-2007). Clin Infect Dis. 2010;50(Suppl 1):S26–33.
70. Trevino M, Areses P, Penalver MD, Cortizo S, Pardo F, del Molino ML, et al. Susceptibility
trends of Bacteroides fragilis group and characterisation of carbapenemase-producing strains
by automated REP-PCR and MALDI TOF. Anaerobe. 2012;18(1):37–43.
71. Nagy E, Urban E, Nord CE, Bacteria ESGoARiA. Antimicrobial susceptibility of
Bacteroides fragilis group isolates in Europe: 20 years of experience. Clin Microbiol Infect.
2011;17(3):371–9.
72. Snydman DR, Jacobus NV, McDermott LA, Supran S, Cuchural GJ Jr, Finegold S, et al.
Multicenter study of in vitro susceptibility of the Bacteroides fragilis group, 1995 to 1996,
with comparison of resistance trends from 1990 to 1996. Antimicrob Agents Chemother.
1999;43(10):2417–22.
73. Seifert H, Dalhoff A, Group PS. German multicentre survey of the antibiotic susceptibility of
Bacteroides fragilis group and Prevotella species isolated from intra-abdominal infections:
results from the PRISMA study. J Antimicrob Chemother. 2010;65(11):2405–10.
74. Snydman DR, Jacobus NV, McDermott LA, Golan Y, Goldstein EJ, Harrell L, et al. Update on
resistance of Bacteroides fragilis group and related species with special attention to carbapen-
ems 2006-2009. Anaerobe. 2011;17(4):147–51.
75. Sherwood JE, Fraser S, Citron DM, Wexler H, Blakely G, Jobling K, et al. Multi-drug resistant
Bacteroides fragilis recovered from blood and severe leg wounds caused by an improvised
explosive device (IED) in Afghanistan. Anaerobe. 2011;17(4):152–5.
76. Piriz S, Vadillo S, Quesada A, Criado J, Cerrato R, Ayala J. Relationship between penicillin-­
binding protein patterns and beta-lactamases in clinical isolates of Bacteroides fragilis with
different susceptibility to beta-lactam antibiotics. J Med Microbiol. 2004;53(Pt 3):213–21.
77. Aldridge KE, Ashcraft D, O'Brien M, Sanders CV. Bacteremia due to Bacteroides fragilis
group: distribution of species, beta-lactamase production, and antimicrobial susceptibility pat-
terns. Antimicrob Agents Chemother. 2003;47(1):148–53.
78. Behra-Miellet J, Calvet L, Dubreuil L. A Bacteroides thetaiotamicron porin that could take
part in resistance to beta-lactams. Int J Antimicrob Agents. 2004;24(2):135–43.
79. Boente RF, Ferreira LQ, Falcao LS, Miranda KR, Guimaraes PL, Santos-Filho J, et al.
Detection of resistance genes and susceptibility patterns in Bacteroides and Parabacteroides
strains. Anaerobe. 2010;16(3):190–4.
80. Garcia N, Gutierrez G, Lorenzo M, Garcia JE, Piriz S, Quesada A. Genetic determinants
for cfxA expression in Bacteroides strains isolated from human infections. J Antimicrob
Chemother. 2008;62(5):942–7.
6 Anaerobic Bacteria: Antimicrobial Susceptibility Testing and Resistance Patterns 215

81. Soki J, Edwards R, Hedberg M, Fang H, Nagy E, Nord CE, et al. Examination of cfiA-medi-
ated carbapenem resistance in Bacteroides fragilis strains from a European antibiotic suscep-
tibility survey. Int J Antimicrob Agents. 2006;28(6):497–502.
82. Toprak NU, Uzunkaya OD, Soki J, Soyletir G. Susceptibility profiles and resistance genes
for carbapenems (cfiA) and metronidazole (nim) among Bacteroides species in a Turkish
University Hospital. Anaerobe. 2012;18(1):169–71.
83. Roberts MC, Sutcliffe J, Courvalin P, Jensen LB, Rood J, Seppala H. Nomenclature for macro-
lide and macrolide-lincosamide-streptogramin B resistance determinants. Antimicrob Agents
Chemother. 1999;43(12):2823–30.
84. Gal M, Brazier JS. Metronidazole resistance in Bacteroides spp. carrying nim genes and
the selection of slow-growing metronidazole-resistant mutants. J Antimicrob Chemother.
2004;54(1):109–16.
85. Lofmark S, Fang H, Hedberg M, Edlund C. Inducible metronidazole resistance and nim
genes in clinical Bacteroides fragilis group isolates. Antimicrob Agents Chemother.
2005;49(3):1253–6.
86. Roberts SA, Shore KP, Paviour SD, Holland D, Morris AJ. Antimicrobial susceptibility of
anaerobic bacteria in New Zealand: 1999-2003. J Antimicrob Chemother. 2006;57(5):992–8.
87. Liu CY, Huang YT, Liao CH, Yen LC, Lin HY, Hsueh PR. Increasing trends in antimicro-
bial resistance among clinically important anaerobes and Bacteroides fragilis isolates causing
nosocomial infections: emerging resistance to carbapenems. Antimicrob Agents Chemother.
2008;52(9):3161–8.
88. Bahar H, Torun MM, Demirci M, Kocazeybek B. Antimicrobial resistance and beta-lacta-
mase production of clinical isolates of prevotella and porphyromonas species. Chemotherapy.
2005;51(1):9–14.
89. Molitoris E, Wexler HM, Finegold SM. Sources and antimicrobial susceptibilities of
Campylobacter gracilis and Sutterella wadsworthensis. Clin Infect Dis. 1997;25(Suppl
2):S264–5.
90. King A, Downes J, Nord CE, Phillips I, European Study G. Antimicrobial susceptibility of
non-Bacteroides fragilis group anaerobic Gram-negative bacilli in Europe. Clin Microbiol
Infect. 1999;5(7):404–16.
Chapter 7
Clinical Significance and Biologic Basis
of HIV Drug Resistance

Rodger D. MacArthur

7.1 Introduction

Antiretroviral resistance limits the initial and subsequent activity of antiretrovirals,

as well as limits the duration of their usefulness. Worse, some antiretroviral-selected
drug mutations limit the activity of one or more other antiretrovirals within the same
class, or even all antiretrovirals within a particular class. An understanding of the
topic is still important, despite the overall decline in circulating drug resistance-­
associated mutations (DRAMs). For instance, some drugs and classes are substan-
tially less likely to be limited by the development of resistance or retain activity in
an environment of previously selected DRAMs. Thus, even the current Department
of Health and Human Services (DHHS) guidelines [1] recommend different classes
of antiretrovirals for initial use in the setting of uncertain medication adherence,
chaotic lifestyles, etc.
Knowledge accumulated after three decades of studying antiretroviral resistance,
at least arguably, has been primarily responsible for the decline in circulating
DRAMs over the last 15 years. In 2000, 15–20% of HIV isolates from various North
American and European cohorts showed resistance to at least one class of antiretro-
virals; that figure is around 5–10% today. Unfortunately, circulating DRAMS to key
antiretrovirals in low- and middle-income countries are substantially higher, typi-
cally exceeding 15–20% [2, 3]. This chapter will summarize much of the relevant
antiretroviral resistance data, especially information that is still relevant today. As
such, it will serve as a useful resource for clinicians caring for HIV-infected persons
today, as well as provide an historical framework to the field.

R. D. MacArthur (*)
Medical College of Augusta at Augusta University, Augusta, GA, USA
e-mail: [email protected]

© Springer International Publishing AG, part of Springer Nature 2018 217

I. W. Fong et al. (eds.), Antimicrobial Resistance in the 21st Century, Emerging Infectious
Diseases of the 21st Century, https://doi.org/10.1007/978-3-319-78538-7_7
218 R. D. MacArthur

7.2 General Principals

Mutations in various structural HIV proteins are random events. They occur as long
as the virus is replicating. These two statements result in Key Concept 1: the only
way to prevent HIV viral resistance from occurring is to shut down viral replication
completely. This concept and other key concepts are summarized in Table 7.1. There
are two corollaries to this concept: (a) If the virus is not replicating, it is not mutat-
ing; and (b) if the virus is replicating, whether in the presence or absence of antiret-
roviral therapy (ART), it is mutating. The rate of mutation development in various
structural HIV proteins is impressive. For instance, it is estimated that in the absence
of antiretroviral therapy, every single possible mutation in every single position of,
say, the 99 amino acid protease, occurs once each day. Of course, many of these
mutations are “dead-end” mutations that are not compatible with viral existence.
Other mutations result in a viral “quasispecies” that is replication-competent but has
a lower replicative capacity (i.e., is less “fit”) than “wild-type” virus. Due to limita-
tions in typically used commercial resistance assays, these less-fit subpopulations
are not detected until their numbers exceed 10–20% of the total viral pool. However,
in the presence of drug, they can become the dominant population, as Fig. 7.1
shows. In addition, when using more sensitive assays that can detect these mutated
quasispecies down to a level of 1%, it has been shown that, when present, they
adversely affect treatment options [4]. Even very low levels of viral replication,
occurring in the presence of drug, will result in the accumulation of drug-limiting
resistance mutations.
Factors that affect the likelihood of selecting for drug-limiting resistance muta-
tions include (a) adherence to the antiretroviral regimen, (b) the potency of the anti-
retroviral drugs, (c) the drug’s intrinsic barrier to resistance, and (d) the duration of
time that antiretroviral drug levels exceed zero but are lower than the level needed
to achieve maximal viral suppression. Intermediate rates of adherence (e.g., 20–80%
of drug taken) typically are associated with the greatest risk for selecting drug-­
limiting resistance mutations. Lower adherence rates (e.g., below 20%) usually
result in drug levels too low to exert any selective pressure on the virus, and higher

Table 7.1 Key concepts concerning HIV resistance and antiretrovirals

The only way to prevent HIV viral resistance from occurring is to shut down viral replication
completely
If a drug, or combination of drugs, is unable to completely shut down viral replication, the
higher the drug levels, the greater the selective pressure on the virus to mutate
If a drug has no activity, there will be no selective pressure on the virus and no further mutations
will occur
The longer drug levels exceed zero but are lower than the level needed to completely shut down
viral replication, the greater is the selective pressure on the virus to mutate
Failure to detect mutations on resistance testing does not rule out their presence (at low levels)
in previously antiretroviral-treated persons
7 Clinical Significance and Biologic Basis of HIV Drug Resistance 219

Fig. 7.1 Evolution of a new dominant quasispecies

Fig. 7.2 In vitro

measurement of drug
potency

adherence rates usually result in drug levels high enough to maximally suppress
viral replication. On the other hand, factors such as poor absorption of the drugs, or
poor bioavailability, will change the dynamics such that higher levels of adherence,
resulting in higher but suboptimal drug levels, will lead to substantial selective pres-
sure on the virus to mutate. Similarly, lower-potency antiretrovirals, or lower-­
potency combinations of antiretrovirals, have the potential to exert substantial
selective pressure on the virus if maximal suppression of replication does not occur.
In vitro, potency of an antiretroviral drug is expressed as percent inhibition of HIV
for any given drug concentration (Fig. 7.2). Clinically, potency of an antiretroviral
drug is expressed as the log reduction in HIV RNA level when that drug is used
220 R. D. MacArthur

Fig. 7.3 A tenfold

increase in resistance as a
result of mutation(s)

alone. Different drugs and different classes have different intrinsic potencies. Log
reductions in HIV RNA level for each drug contribute to the potency of the entire
regimen in an additive fashion: the anticipated potency of any particular a­ ntiretroviral
regimen is the sum of the potencies (in log HIV RNA reduction) of the individual
components. Several key concepts follow: Key Concept 2 – if a drug or combination
of drugs is unable to completely shut down viral replication, the higher the drug
levels, the greater the selective pressure on the virus to mutate. Key Concept 3: if a
drug has no activity, there will be no selective pressure on the virus, and no further
mutations will occur. Key Concept 4: the longer drug levels exceed zero but are
lower than the level needed to completely shut down viral replication, the greater is
the selective pressure on the virus to mutate.
Mutations that occur as a result of selective pressure of a drug have the effect of
increasing the amount of drug necessary to achieve the same degree of killing
(Fig. 7.3). That is, graphically, they shift the “kill curve” to the right. If the shift
results, for instance, in ten times more drug necessary to get the same amount of
killing, this effect often is referred to as a tenfold increase in resistance to a particu-
lar drug. These mutations may have no effect on other antiretrovirals of the same
class, or they may have the same effect, a greater effect, or a lesser effect. In some
situations (discussed later), these mutations may even shift the kill curve to the left
for another drug in the same class, effectively resulting in increased activity of that
drug. This phenomenon is referred to as “hypersusceptibility” of HIV to the drug
with the left-shifted curve.
If the initial mutation that occurs as a result of selective pressure of a particular
drug is not sufficient to eliminate all activity of that drug, additional mutations
will accumulate in the presence of that drug until that drug (and typically many
others in the same antiretroviral class) no longer has any activity against the virus.
In some circumstances, these subsequent mutations will not decrease the activity
of the offending drug but rather restore fitness (replicative capacity) to the virus.
7 Clinical Significance and Biologic Basis of HIV Drug Resistance 221

In these cases, the mutations are referred to as “secondary” or “compensatory”

mutations. Note that not all mutations result in a viral quasispecies with lower fit-
ness than wild type. In that case, these mutated quasispecies will co-circulate with
wild type and not be “overgrown” (outcompeted) by wild-type virus. Even when
there is ­overgrowth of wild-type virus, such that a mutated quasispecies is not
detected on resistance testing after discontinuation of the offending drug, the
mutated quasispecies typically will “reappear” on reintroduction of the drug. In
other words, the mutated quasispecies has been “archived” but not eliminated
from the total viral pool. Key Concept 5 follows: failure to detect mutations on
resistance testing does not rule out their presence (at low levels) in previously
antiretroviral-treated persons.

7.3 The Language of Resistance

There are three HIV enzymes that are the main targets for antiretrovirals: reverse
transcriptase, protease, and integrase. Reverse transcriptase is a 560 amino acid
(AA) heterodimer comprised of p51 and p66 subunits. The p51 subunit contains
the first 450 amino acids; the p66 subunit contains the rest [5]. Most resistance
assays stop sequencing around AA 250 or so, as most of the relevant mutations
occur from AA 41 through AA 219. Protease is composed of two non-covalently
associated, structurally identical monomers of 99 amino acids each [5]. Integrase is
a 288 AA dimer comprised of 3 domains: the N-terminal domain, the catalytic core
domain, and the C-terminal domain. The catalytic core contains the triad of AAs
comprising the D,D-35-E motif, made up of aspartic acid at positions 64 and 116
and glutamine at position 152. Mutations at any of these positions essentially abol-
ish all catalytic activity of integrase [6]. As such, these mutations occur uncom-
monly. Most of the relevant mutations selected by exposure to the integrase strand
transfer inhibitor class of antiretrovirals occur in the catalytic core domain, from
AA 74 through AA 230.
By convention, mutations at any position in these enzymes are listed as the one-­
letter AA code that is expected at a particular numbered position, followed by the
one-letter AA code of the mutated quasispecies. So, for instance, the very com-
monly occurring valine-for-methionine mutation at position 184 of reverse tran-
scriptase is listed as M184V. Often, multiple mutations can be seen (detected) at
each position, including position 184 of reverse transcriptase. The convention is to
list these mutations separated by a slash (e.g., M184V/I). Not every detectable
mutation at a given position results in the same loss of activity to a particular drug.
In addition, because three nitrogenous bases encode each amino acid, and multiple
triads can code for the same amino acid, reverting (or back reverting) to wild type
from a mutated quasispecies often involves going through an intermediate form [7].
So, for instance, the zidovudine-selected mutation, T215Y, tends to mutate to T215S
prior to reverting to wild type. Typically, the back revertants (e.g., T215S) tend to be
fitter than the original mutated version. Finally, mutations such as T215Y that
222 R. D. MacArthur

Table 7.2 One-letter amino Alanine A GCT, GCC, GCA, GCG

acid codes and corresponding
Arginine R CGT, CGC, CGA, CGG, AGA, AGG
DNA codons
Asparagine N AAT, AAC
Aspartic acid D GAT, GAC
Cysteine C TGT, TGC
Glutamic acid E GAA, GAG
Glutamine Q CAA, CAG
Glycine G GGT, GGC, GGA, GGG
Histidine H CAT, CAC
Isoleucine I ATT, ATC, ATA
Leucine L CTT, CTC, CTA, CTG, TTA, TTG
Lysine K AAA, AAG
Methionine M ATG
Phenylalanine F TTT, TTC
Proline P CCT, CCC, CCA, CCG
Serine S TCT, TCC, TCA, TCG, AGT, AGC
Threonine T ACT, ACC, ACA, ACG
Tryptophan W TGG
Tyrosine Y TAT, TAC
Valine V GTT, GTC, GTA, GTG

require double-nucleotide mutations (e.g., 215T(ACC) to 215Y(TAC)) often are slower

to develop than mutations that require only a single nucleotide change [7]. Table 7.2
lists the one-letter amino acid code and the three nitrogenous base sequences for
each amino acid.

7.4 Key Mutations in Reverse Transcriptase

Two classes of antiretrovirals, the nucleoside reverse transcriptase inhibitors

(NRTIs) and the non-nucleoside reverse transcriptase inhibitors (NNRTIs), target
reverse transcriptase. The NRTIs, also known as nucleoside and nucleotide ana-
logues, are analogues of thymidine (zidovudine and stavudine), adenosine (didano-
sine, tenofovir), cytidine (lamivudine, zalcitabine, emtricitabine), and guanosine
(abacavir). These drugs serve as “chain terminators” when they are incorporated
into the viral DNA during HIV synthesis in newly infected cells. No other base can
be added to viral DNA once an analogue has been incorporated. These antiretrovi-
rals are considered “1-log” drugs, as they each lower HIV RNA levels by that
amount. Mutations selected by these drugs typically are drug-specific, as indicated
in Table 7.3. With monotherapy, some of these mutations develop over many months
(e.g., T215Y/F), others over a few weeks (M184V). Some of the listed mutations in
7 Clinical Significance and Biologic Basis of HIV Drug Resistance 223

Table 7.3 Mutations Zidovudine M41L, D67N, K70R, L210W,

selected by nucleoside and T215Y/F, K219Q/E
nucleotide reverse
Stavudine M41L, D67N, K70R, L210W,
transcriptase inhibitors
T215Y/F, K219Q/E
Didanosine K65R, L74V*
Tenofovir K65R
Zalcitabine K65R, T69D*, L74V, M184V
Lamivudine M184V
Emtricitabine M184V
Abacavir K65R*, L74V, Y115F, M184V
Note: The most common mutations selected by
some drugs are indicated by *

Table 7.3 are never selected during combination therapy. For instance, the M184V
mutation is not selected when abacavir is used in combination with zidovudine,
likely due to the hypersusceptibility that M184V imparts to zidovudine. The extent
of resistance conferred to a particular drug by a particular mutation often depends
on both the drug and the mutation. So, for instance, the M184V mutation confers
high-level resistance to lamivudine and emtricitabine, hypersusceptibility to zid-
ovudine, tenofovir, and stavudine, and only partial resistance to abacavir. On the
other hand, the mutations that are selected by both zidovudine and stavudine,
referred to as “thymidine analogue mutations (TAMs),” confer some degree of resis-
tance to all of the NRTIs.
There are some mutations that are associated with properties not limited to one
or a few of the NRTIs. Foremost among these is the M184V mutation. This muta-
tion, while having the effect on individual drugs described above, also results in a
viral quasispecies that has substantial loss in fitness. This phenomenon, first
described by Mark Wainberg and colleagues in Montreal [8], results clinically in a
sustained 0.5 log reduction in HIV RNA; at the same time the virus has developed
up to 1000-fold resistance to either lamivudine or emtricitabine [9]. In addition, this
mutation results in a viral quasispecies that has greater “fidelity,” meaning that it is
more difficult for the virus to revert to wild type [10].
Other mutations of note that are selected by exposure to the NRTIs are (1)
Q151M (along with accessory mutations at positions 62, 75, 77, and 116) which
conveys high-level resistance to all NRTIs; (2) 69 insertion complex, which
refers to amino acid insertions between codons 67 and 70 and conveys resistance
to all NRTIs; (3) 70 deletion complex, which refers to amino acid deletions
between codons 67 and 70 and conveys resistance to all NRTIs; and (4) E40F,
E44D/A, and V118I are referred to as “accessory” mutations that are seen occa-
sionally in association with the TAMs and contribute to the extent of resistance
conferred by the TAMs.
Mutations, in general, impart some degree of structural change to the relevant
enzyme. In the case of NRTI-limiting mutations, the resulting structural changes to
224 R. D. MacArthur

Fig. 7.4 Mechanisms of resistance to NRTIs

reverse transcriptase have one of two effects: they prevent the nucleoside analogue
from binding, due to steric hindrance, or they facilitate excision of the already incor-
porated analogue (Fig. 7.4). The excision activity of HIV reverse transcriptase cata-
lyzes the 3′-terminal residue of the analogue to an acceptor substrate, most likely
ATP [11]. Some mutations, such as K65R, have the effect of decreasing binding/
incorporation for several of the NRTIs, while at the same time having a variable
effect on the stability of the incorporated drug. So, for instance, the effect of K65R
is to make the virus hypersusceptible to zidovudine, susceptible to stavudine and
abacavir, and more resistant to didanosine and tenofovir.
The NNRTI class of antiretrovirals also inhibit HIV reverse transcriptase,
although the mechanism is not as straightforward as with the NRTIs. Unlike the
NRTIs, the NNRTIs are structurally quite diverse. However, all of them bind allo-
sterically to reverse transcriptase in the same general region, albeit to different
amino acids. The NNRTI-binding pocket of reverse transcriptase is hydrophobic,
located about 10 angstroms from the catalytic site in the “palm” region of the p66
subunit [12]. Binding of the NNRTIs induces conformational changes that inhibit
the catalytic activities of reverse transcriptase. Drugs in this class are more potent
than the NRTIs, typically able to decrease HIV RNA levels by 2–4 logs or more.
However, single-point mutations at positions 103 and 181 (K103N, Y181C) devel-
oped quickly in the presence of the “first generation” of NNRTIs, nevirapine, dela-
virdine, and efavirenz [13]. These mutations alter the size, shape, and polarity of the
NNRTI-­binding pocket, thereby reducing access to the pocket by those NNRTIs
[12]. Interestingly, although not clinically relevant due to the lack of current use of
zidovudine, it was shown that the addition of zidovudine to nevirapine changed the
mutation selection pattern from predominantly Y181C to K103N [14]. The “second-­
generation” NNRTIs, etravirine and rilpivirine, retain good activity against HIV in
7 Clinical Significance and Biologic Basis of HIV Drug Resistance 225

Table 7.4 Other important key mutations

K103N Most common mutation selected by efavirenz. Confers high-level resistance to
efavirenz, nevirapine, and delavirdine. Does not limit the activity of etravirine
or rilpivirine
Y181C More commonly selected by nevirapine and delavirdine than efavirenz [13].
Confers high-level resistance to efavirenz, nevirapine, and delavirdine and
intermediate-level resistance to etravirine and rilpivirine
V90I, Selected by, and limits the activity of, etravirine and rilpivirine
E138A/K,
M230L
L33F, I84V Multi-protease inhibitor mutations, resulting in substantial reductions in, or
elimination of, activity of all of the protease inhibitors
Q148H/R/K, Major mutations for both raltegravir and elvitegravir
N155H

the presence of the K103N mutation but have reduced activity in the presence of the
Y181C/I/V mutation. Other mutations that limit the activity of etravirine and rilpi-
virine include V90I, E138A/K, and M230L. The effect of these mutations on the
NNRTIs is summarized in Table 7.4.

7.5 Key Mutations in Protease

The protease inhibitor class of antiretrovirals, when “boosted” with the cytochrome
p450 inhibitors ritonavir or cobicistat, is the most potent of all of the classes. The
drugs in this class inhibit a key enzyme (protease) in the replicative cycle of
HIV. This enzyme cleaves unnecessary pieces of viral protein just prior to maturity
of the virus, thereby allowing the virus to fold into its compact, infectious, shape/
form. The class is the least susceptible to the development of resistance mutations,
when combined with one of the boosters. Even monotherapy (with a booster) results
in sustained HIV RNA decreases of 4–6 logs. It is the only class for which mono-
therapy can be considered, although it is not recommended to do so routinely. It is
extraordinarily rare for any mutations to develop in previously antiretroviral-naïve
individuals when on a standard multidrug combination. Mutations have developed,
rarely, in persons on lopinavir and atazanavir, for instance, when each is boosted
with ritonavir. Much of our knowledge of the virologic consequences of the devel-
opment of mutations in this class comes from 20 years ago before the use of a
booster was common. For instance, unboosted indinavir selected for mutations
slowly and retained antiviral activity even in the presence of several mutations. On
the other hand, by the time additional mutations developed, HIV protease typically
was resistant to all of the drugs in the class (unboosted). From a historical perspec-
tive, mutations that developed at key binding locations in protease resulted in
lengthening of the bond between the enzyme and the inhibitor, as well as other
structural changes. For instance, the L33F multidrug resistance mutation [15] has
226 R. D. MacArthur

been shown to reduce flexibility of protease in the 30s and 80s loops [16]. As a
result of early mutational changes in protease, the enzyme typically loses a substan-
tial degree of fitness. Over time, additional mutations develop away from the bind-
ing site that have the effect of restoring fitness by “tightening” the structure of the
enzyme
Certain other unboosted protease inhibitor-selected mutations are now of passing
interest only, as these drugs are not used either at all or in the absence of a booster.
So, for instance, the PI nelfinavir selected for the D30N mutation, which substan-
tially limited the activity of that drug but had no effect on other PIs. The atazanavir-­
selected I50L mutation resulted in high-level resistance to atazanavir but had no
effect on the activity of amprenavir. Similarly, the I50V amprenavir-selected muta-
tion resulted in high-level resistance to amprenavir but had no effect on the activity
of atazanavir. As with the NRTI and NNRTI classes, there were a number of other
“signature” mutations (e.g., G48V and L90M for saquinavir) that often would reli-
ably indicate the previous drug exposure history of the virus.

7.6 Key Mutations in Integrase

The integrase strand transfer inhibitor (INSTI) class also is extremely potent, result-
ing in sustained reductions in HIV RNA of 4–5 logs. However, the class is more
susceptible to the development of drug-limiting resistance mutations than is the PI
class. For instance, in the setting of prior virologic failure on an NRTI-based regi-
men, a switch from a boosted PI regimen (lopinavir-ritonavir) to a raltegravir-based
regimen was associated with a substantial virologic failure rate in the raltegravir
arm, despite participants having an HIV RNA level < 400 copies/ml at the time of
the switch [17].
In addition to raltegravir, there are two other currently available integrase strand
transfer inhibitors: elvitegravir, which must be boosted (and coformulated) with
cobicistat, and dolutegravir. All of these drugs work at the last step of integration,
by inhibiting insertion of viral DNA into host chromosomal DNA (strand transfer
step). There is substantial overlap in the mutations that are selected by, and limit the
activity of, raltegravir and elvitegravir. These include T66A/I/K, E92Q, E138K/A/T,
G140S/A/C, Y143R/C/H (raltegravir only), and, especially, Q148H/R/K and
N155H. These mutations are located near the catalytic core domain of integrase,
which is where the INSTIs bind [18]. These mutations have the effect of excluding
binding of the INSTIs but at the expense of reduced enzymatic fitness. As might be
anticipated, a number of accessory mutations often are seen in association with the
major mutations. These secondary mutations have the effect of restoring fitness to
the virus.
Dolutegravir, the third INSTI currently available, is the most resilient and least
susceptible to the development of resistance of the three. In fact, mutations which
substantially limit the activity of both raltegravir and elvitegravir typically have
little impact on the activity of dolutegravir. However, mutations do develop, espe-
7 Clinical Significance and Biologic Basis of HIV Drug Resistance 227

cially when the drug is used as monotherapy in an “off-label,” not recommended,

fashion. In addition, combinations of mutations at positions 138, 140, 148, and 155
result in substantial loss of activity of the drug, which may be overcome by doubling
the dose.

7.7 Mutations from Other Classes

Enfuvirtide is the only drug in the fusion inhibitor class of antiretrovirals. It also is
the only antiretroviral which must be administered by injection. Relatively little is
known about the development of resistance to this drug and even less about the
impact of specific mutations. Furthermore, routine resistance testing is not avail-
able, and the use of the drug is limited to the heavily treatment-experienced group
of patients, those defined by mutations in multiple other classes, and comprising at
most 5% of the total HIV-infected population. Enfuvirtide binds to the HR1 domain
of GP41, after the envelope protein has bound to the CD4 receptor of the cell [19].
Binding of enfuvirtide prevents the conformational changes that are needed for
membrane fusion. While mutations in the envelope gene at positions 36–43 have
been detected in viruses from patients on enfuvirtide, it is not clear what impact any
single or multiple mutations has on drug activity. In fact, most clinical trials using
resistance as an entry criterion default to “ever use” of enfuvirtide to arbitrarily
declare that HIV-1 likely is resistant to that drug.
Maraviroc is the only antiretroviral that targets a human protein, the CCR5 co-­
receptor. The drug has good activity in both B- and non-B-subtypes of HIV-1, as
long as the individual is CCR5-tropic and not dual- (CCR5, CXCR4) or mixed-­
tropic. The main resistance mechanism involves utilization of maraviroc-bound
CCR5 co-receptors for entry, as a result of multiple mutations in the V3 loop of
GP120 [20]. Another way HIV can enter in the presence of maraviroc is by co-­
receptor switching (i.e., utilizing CXCR4). Co-receptor switching occurs infre-
quently, as it results in substantially reduced fitness and efficiency of entry, as a
result of multiple mutations in GP160 [20]. There is no cross-resistance with enfu-
virtide. Like enfuvirtide, maraviroc is used infrequently at the present time. Early
assays used to detect minority variants that were dual- or mixed-tropic were not
sensitive enough to exclude this population of patients from entry into clinical
trials.

7.8 Mutation Pathways

Mutation “pathways” have been well-described for the thymidine analogue drugs
zidovudine and stavudine, as well as for the INSTI raltegravir. Pathways refer to the
accumulation of specific mutations based on the previous development of other
mutations. Little is known about the factors that “force” the virus down one pathway
228 R. D. MacArthur

Fig. 7.5 Zidovudine pathways

or another, but fitness likely plays a role. The two thymidine analogue pathways are
shown in Fig. 7.5. While the mutations seen at the full expression of each pathway
overlap, the process by which they accumulate differs between the pathways. In
addition, it has been observed that the mutation L210W is always found in associa-
tion with T215Y, a mutation that is not present in the “second” TAM pathway. Note
that the first TAM pathway, involving the mutations L210W and T215Y, is the more
common of the two and results in a virus more resistant to all of the NRTIs.
Even less is known about the 2–3 raltegravir pathways, the N155H pathway, the
Q148H/R/K pathway, and the Y143R/C/H pathway. The Q148H/R/K pathway is
associated with the least reduction in viral fitness. Minority variants that exist at the
time of drug initiation may play a role in determining the specific pathway the virus
takes to accumulate mutations.

7.9 Resistance Tests and Algorithms

There are two main types of HIV resistance tests: genotypic tests, based on sequenc-
ing and determining the specific amino acids present at each position in key
enzymes, and phenotypic tests, which test for the degree of killing of the isolated
virus in the presence of different antiretrovirals. Genotypic tests are an indirect mea-
sure of resistance, based primarily on assigning a “likelihood of failure” score to a
particular mutation or set of mutations. Phenotyping is a direct measure of a drug’s
activity in killing the mutated or unmutated virus. Genotyping is quicker, more
readily available (e.g., not limited to commercial laboratories), and less expensive.
7 Clinical Significance and Biologic Basis of HIV Drug Resistance 229

Fig. 7.6 Key mutations predicting lopinavir/ritonavir resistance

However, a certain amount of “faith” needs to be placed in the algorithm-based

interpretation of mutations inherent in this approach.
The earliest algorithms that purported to predict resistance to antiretrovirals were
relatively primitive, often relying on the total number of mutations detected. In
other words, the more mutations present, the more likely was the virus to be deemed
resistant to a particular drug. Even though it was known relatively early that, for
instance, the M184V mutation in reverse transcriptase resulted in high-level resis-
tance to lamivudine and emtricitabine, and the K103N mutation had the same effect
on the first generation of NNRTIs, a simple count predominated the early efforts at
predicting resistance to the PIs. An early refinement to this approach involved pat-
tern recognition and the identification of “key” mutations. Predicting resistance to
the ritonavir-boosted PI lopinavir is an excellent example of how this approach
progressed. Figure 7.6 illustrates the identified key mutations from three different
algorithms [21–23]. All were reasonably effective at predicting lack of efficacy, but
the three had only about 1/3 of their mutations in common. Call these the “really
key” mutations, but it wasn’t until “weights” were assigned to all mutations in each
algorithm that the algorithms became substantially predictive of drug activity [24].
A similar approach for the NNRTI etravirine is shown in Fig. 7.7. Around that same
time (circa 2007), multiple algorithms were “competed” to see which was best at
predicting resistance to particular PIs [25]. As a result of these competitions, it
became apparent that large databases had a huge advantage over much smaller
databases.
The Virco™ Vircotype (or virtual phenotype) report (no longer available) took
this approach one step further, by correlating over 350,000 genotype specimens
230 R. D. MacArthur

Fig. 7.7 Algorithm-based determinations

with over 90,000 paired phenotype specimens and assigning weights to each singlet
and pair of mutations associated with different antiretrovirals [26]. By doing so,
Virco™ was able to show that a single mutation (i.e., I84V) that substantially
reduced the activity of, for instance, tipranavir, behaved substantially differently in
combination with the L10F mutation. Today, as the cost of phenotyping has dropped
in North America and Europe, genotype and phenotype tests often are run together.
When genotype tests are run without a phenotype, the reported results typically are
substantiated by a large database of isolates.

7.10 Clinical Issues of Relevance

The history of treating HIV-infected persons with antiretrovirals has benefitted from
three decades of research on resistance issues specific to these drugs. Treatment
guidelines [1] have evolved as we have acquired a better understanding of the devel-
opment and prevention of resistance to antiretrovirals. For instance, prior to 2005,
the DHHS guidelines recommended baseline resistance testing only in the acutely
or recently infected individual. This recommendation was based on the belief that
drug-selected resistance mutations would be overgrown by (revert to) wild-type
virus in the absence of continuous selective pressure of antiretroviral therapy. That
7 Clinical Significance and Biologic Basis of HIV Drug Resistance 231

belief was shown to be incorrect, by an article documenting the persistence of muta-

tions in multiple classes in individuals likely infected years earlier [27].
Unfortunately, the HIV RNA and resistance assays that we take for granted in North
America and Europe are not available in much of the rest of the world. As a result,
resistance continues to develop at an alarming rate in individuals still clinically well
who remain on failing antiretroviral regimens [28].
Finally, it is worth noting two excellent sources of current information about the
effect of mutations on various antiretrovirals: the Stanford University HIV database
[29] and the annually updated IAS-USA list of mutations by class [30].

Major Points
• Understand basic principles of antiretroviral resistance
• Learn key drug- and class-related resistance mutations
• Apply basic science-related principles of drug resitance to clinical applications

References

1. Panel on Antiretroviral Guidelines for Adults and Adolescents. Guidelines for the Use
of Antiretroviral Agents in Adults and Adolescents Living with HIV. Department of
Health and Human Services. Available at http://aidsinfo.nih.gov/contentfiles/lvguidelines/
AdultandAdolescentGL.pdf. Accessed 15 Nov 2017.
2. World Health Organization. HIV drug resistance report 2017. Available at www.who.int/hiv/
pub/drugresistance/hivdr-report-2017/en. Accessed 17 Nov 2017.
3. Beyrer C, Pozniak A. HIV drug resistance – an emerging threat to epidemic control. N Engl
J Med. 2017;377:1605–7.
4. Simen BB, Simons JF, Hullsiek KH, et al. Low-abundance drug-resistant viral variants in
chronically HIV-infected, antiretroviral treatment-naive patients significantly impact treatment
outcomes. J Infect Dis. 2009;199:693–701.
5. Shafer RW, Jung DR, Betts BJ, et al. Human immunodeficiency virus reverse transcriptase and
protease sequence database. Nucleic Acids Res. 2000;28:346–8.
6. Craigie R. HIV integrase, a brief overview from chemistry to therapeutics. J Biol Chem.
2001;276:23213–6.
7. Kuritzkes DR. A fossil record of zidovudine resistance in transmitted isolates of HIV-1. PNAS.
2001;98:13485–7.
8. Quan Y, Brenner BG, Oliveira M, Wainberg MA. Lamivudine can exert a modest antivi-
ral effect against human immunodeficiency virus type 1 containing the M184V mutation.
Antimicrob Agents Chemother. 2003;47:747–54.
9. Eron J, Benoit S, Jemsek J, et al. The safety and efficacy of lamivudine (3TC)/zidovudine com-
bination therapy vs. zidovudine monotherapy and lamivudine monotherapy in therapy-naive,
HIV-positive patients with CD4 cell counts of 200- 500/mm3: a randomized, double-blind,
multicenter, controlled trial. N Engl J Med. 1995;333:1662–9.
10. Wainberg MA, Drosopoulos WC, Salomon H. Enhanced fidelity of 3TC-selected mutant
HIV-1 reverse transcriptase. Science. 1996;271:1282–5.
11. Acosta-Hoyos A, Scott WA. The role of nucleotide excision by reverse transcriptase in HIV
drug resistance. Virus. 2010;2:372–94.
12. deBéthune MP. Non-nucleoside reverse transcriptase inhibitors (NNRTIs), their discovery,
development, and use in the treatment of HIV-1 infection: a review of the last 20 years (1989 –
2009). Antivir Res. 2010;85:75–90.
232 R. D. MacArthur

13. van den Berg-Wolf M, Hullsiek KH, Peng G, Kozal MJ, Novak RM, Chen L, Crane LR,
MacArthur RD, CPCRA 058 Study Team, the Terry Beirn Community Programs for Clinical
Research on AIDS (CPCRA), and The International Network for Strategic Initiative in Global
HIV Trials (INSIGHT). Virologic, immunologic, clinical, safety, and resistance outcomes
from a long-term comparison of efavirenz-based versus nevirapine- based antiretroviral regi-
mens as initial therapy in HIV-1-infected persons. HIV Clin Trials. 2008;9:324–36.
14. Richman DD, Havlir D, Corbeil J, et al. Nevirapine resistance mutations of human immunode-
ficiency virus type 1 selected during therapy. J Virol. 1994;68:1660–6.
15. Kozal MJ, Hullsiek KH, Leduc R, Novak RM, MacArthur RD, Lawrence J, Baxter JD. Terry
Beirn Community Programs for Clinical Research on AIDS (CPCRA). Prevalence and impact
of HIV-1 protease codon 33 mutations and polymorphisms in treatment-naive and treatment-­
experienced patients. Antivir Ther. 2006;11:457–63.
16. Kuiper BD, Keusch BJ, Dewdney TG, et al. The L33F darunavir resistance mutation acts as a
molecular anchor reducing the flexibility of the HIV-1 protease 30s and 80s loops. Biochem
Biophys Rep. 2015;2:160–5.
17. Bianco JL, Gonzalez-Cordón A, Llibre JM, et al. Impact of prior virologic failure and nucelos(t)
ide genotypic resistance mutations on the efficacy of switching from ritonavir-boosted prote-
ase inhibitors to raltegravir. Antivir Ther. 2015;20:487–92.
18. Mbisa JL, Martin SA, Cane PA. Patterns of resistance development with integrase inhibitors in
HIV. Infect Drug Resist. 2011;4:65–76.
19. Reeves JD, Lee F-H, Miamidian JL, et al. Enfuvirtide resistance mutations: impact on human
immunodeficiency virus envelope function, entry inhibitor sensitivity, and virus neutrilazation.
J Virol. 2005;79:4991–9.
20. MacArthur RD, Novak RM. Reviews of anti-infective agents: maraviroc: the first of a new
class of antiretroviral agents. Clin Infect Dis. 2008;47:236–41.
21. Kempf DJ, Isaacson JD, King MS, et al. Identification of genotypic changes in human immu-
nodeficiency virus protease that correlate with reduced susceptibility to the protease inhibitor
lopinavir among protease inhibitor-experienced patients. J Virol. 2001;75:7462–9.
22. Parkin N, Chappey C, Petropoulos C. Improving lopinavir genotype algorithm through
phenotype correlations: novel mutation patterns and amprenavir cross-resistance. AIDS.
2003;17:955–61.
23. Grant P, Wong EC, Rode R, et al. Virologic response to lopinavir-ritonavir-based antiretroviral
regimens in a multicenter international cohort: comparison of genotypic interpretation scores.
Antimicrob Agents Chemother. 2008;52:4050–6.
24. King MS, Rode R, Cohen-Codar I, et al. Predictive genotypic algorithm for virologic
response to lopinavir-ritonavir in protease inhibitor-experienced patients. Antimicrob Agents
Chemother. 2007;51:3076–4.
25. Schapiro JM, Scherer J, Vinisko R, et al. Genotypic tipranavir scores as predictors of response.
11th European AIDS conference; Madrid, Spain; 24–27 October 2007; P3.4/16.
26. Van Houtte M, Picchio G, Van Der Borght K, et al. A comparison of HIV-1 drug susceptibility
as provided by conventional phenotyping and by a phenotype prediction tool based on viral
genotype. J Med Virol. 2009;81:1702–9.
27. Novak RM, Chen L, MacArthur RD, Baxter J, Huppler Hullsiek K, Peng G, Xiang Y, Henley C,
Schmetter B, Uy J, van den Berg-Wolf M, Kozal M, and the Terry Beirn Community Programs
for Clinical Research on AIDS 058 Study Team. Prevalence of antiretroviral drug resistance
mutations in chronically human immunodeficiency virus- infected, treatment-naïve patients:
implications for routine resistance screening before initiation of antiretroviral therapy. Clin
Infect Dis. 2005;40:468–74.
28. Sungkanuparth S, Manosuth W, Kiertiburanakul S, et al. Options for a second-line antiretrovi-
ral regimen for HIV type 1-infected patients whose intial regimen of a fixed-dose combination
of stavudine, lamivudine, and nevirapine fails. Clin Infect Dis. 2007;44:447–52.
29. Stanford University HIV Drug Resistance Database. Available at: https://hivdb.stanford.edu.
Last accessed 19 Nov 2017.
30. Wensing AM, Calvez V, Günthard HF, et al. 2017 update of the drug resistance mutations in
HIV-1. Top Antivir Med. 2017;24:132–41.
Chapter 8
Resistance of Herpesviruses to Antiviral
Agents

William L. Drew, Jocelyne Piret, and Guy Boivin

8.1 Antiviral Agents for Herpesvirus Infection

Three antiviral agents and a prodrug are currently available for the systemic treat-
ment of human cytomegalovirus (HCMV) infections. Ganciclovir (GCV, Cytovene®,
Roche) is a deoxyguanosine analogue and was the first drug to be approved in 1988.
Since then, it has remained the first-line treatment for HCMV infections in immu-
nocompromised patients. Upon entry in HCMV-infected cells, GCV is selectively
phosphorylated by a viral phosphotransferase (the product of the UL97 gene,
pUL97). Subsequently, cellular kinases convert GCV-monophosphate into GCV-­
triphosphate, which acts as a potent inhibitor of the HCMV DNA polymerase (pol)
by competing with deoxyguanosine triphosphate on the enzyme binding site
(Fig. 8.1). Ganciclovir is also incorporated into the viral DNA where it slows down
and eventually stops chain elongation [15, 22, 213]. Ganciclovir formulations are
available for intravenous (IV) or oral administration as well as intravitreal injections
for the treatment of HCMV diseases in immunocompromised patients. Due to its
poor bioavailability (~6%), efforts were made to develop a prodrug of
GCV. Valganciclovir (VGCV, Valcyte®, Roche) is a L-valyl ester formulation of
GCV exhibiting about ten times the bioavailability of GCV following oral adminis-
tration [175].
The other two compounds approved for systemic treatment of HCMV infections
are also potent inhibitors of the viral DNA pol. However, due to their toxicity

W. L. Drew
Departments of Pathology and Laboratory Medicine, University of California,
San Francisco, CA, USA
J. Piret · G. Boivin (*)
Research Center in Infectious Diseases of the CHU of Quebec- Laval University,
Quebec City, Canada
e-mail: [email protected]

© Springer International Publishing AG, part of Springer Nature 2018 233

I. W. Fong et al. (eds.), Antimicrobial Resistance in the 21st Century, Emerging Infectious
Diseases of the 21st Century, https://doi.org/10.1007/978-3-319-78538-7_8
234 W. L. Drew et al.

Fig. 8.1 Mechanisms of action of the different classes of antiviral agents. The nucleoside ana-
logues such as ganciclovir (GCV), acyclovir (ACV), and penciclovir (PCV) must be first phos-
phorylated by the UL97 protein kinase or viral thymidine kinase (TK) and then by cellular kinases
to be converted into their active forms. The cyclic nucleoside analogue brivudine (BVDU) is con-
verted into monophosphate and diphosphate forms by the viral TK followed by an additional
phosphorylation by cellular kinase. The acyclic nucleoside phosphonate derivatives such as cido-
fovir (CDV) must be phosphorylated by cellular kinases only to be active. The resulting triphos-
phate forms compete with deoxynucleotide triphosphates (dNTPs) to inhibit viral replication. The
pyrophosphate analogue foscarnet (FOS) directly inhibits the activity of the viral DNA poly-
merase. Key: Ⓟ represents one phosphate group

p­ rofiles and absence of oral formulation, they are usually reserved for patients fail-
ing or not tolerating GCV therapy. Cidofovir (CDV, Vistide®, Gilead Sciences) is a
nucleotide analogue of cytidine (also called acyclic nucleoside phosphonate) that
only requires activation (phosphorylation) by cellular but not viral enzymes to exert
its antiviral activity [61]. The diphosphate form of CDV is a competitive inhibitor of
HCMV DNA pol, and it may act as a DNA chain terminator when two consecutive
incorporations into DNA occur (Fig. 8.1). The IV formulation of CDV is indicated
for the treatment of HCMV retinitis in patients with AIDS and is also occasionally
used in transplant recipients. Foscarnet (FOS, Foscavir®, AstraZeneca), a pyrophos-
phate analogue, differs from the two previous antivirals both by its mechanism of
action and by the fact that it does not require any activation step to exert its antiviral
activity. Foscarnet binds to and blocks the pyrophosphate binding site on the viral
polymerase, thus preventing incorporation of incoming deoxynucleotide triphos-
phate (dNTP) into viral DNA (Fig. 8.1) [59]. The IV formulation of FOS is indi-
cated for the treatment of HCMV retinitis in individuals with AIDS and for
GCV-resistant HCMV infections in immunocompromised patients.
In addition to the treatment of established HCMV disease, antivirals have also
been used to prevent symptomatic episodes, especially in transplant recipients. The
8 Resistance of Herpesviruses to Antiviral Agents 235

first strategy, defined as prophylaxis, consists of administering an antiviral to

patients during the first 3–6 months after transplantation. This strategy is employed
in recipients of solid organ transplants (SOT) but not in hematopoietic stem cell
transplants (HSCT) due to the marrow toxicity of ganciclovir. However, the occur-
rence of late-onset HCMV disease which is associated with high rates of graft loss
[7] and mortality [146] is an important issue after discontinuing prophylaxis. The
second strategy, referred to as “preemptive therapy,” consists of using short courses
of antivirals only for high-risk patients based on evidence of active viral replication
(e.g., detection of early HCMV antigens such as the pp65 protein or sufficient
amounts of viral DNA/mRNA in the blood) [24]. The advantages of preemptive
therapy include a lower rate of delayed occurrence of HCMV disease and less toxic-
ity [203]. However, patients are more prone to recurrent episodes of DNAemia, and
the indirect effects of HCMV infection on graft and patient survival may not be
prevented. These preventive strategies have shown efficacy in preventing HCMV
disease in both SOT and HSCT patients [25, 95, 118, 151, 171–173].
Antiviral agents currently licensed for the treatment of herpes simplex virus
(HSV) and varicella-zoster virus (VZV) infections include acyclovir (ACV,
Zovirax®, GlaxoSmithKline) and its L-valyl-ester prodrug valacyclovir (VACV,
Valtrex®, GlaxoSmithKline), famciclovir (FCV, Famvir®, Novartis) which is the
L-valyl-ester prodrug of penciclovir (PCV), and FOS. Acyclovir and PCV are deox-
yguanosine analogues that must be phosphorylated by the thymidine kinase (TK) of
HSV and VZV and then by cellular kinases to exert their antiviral activity [22, 98].
The triphosphate forms are competitive inhibitors of the viral DNA pol of HSV and
VZV (Fig. 8.1). In addition, incorporation of ACV-triphosphate into the replicating
viral DNA chain stops synthesis [15]. Oral ACV, VACV, and FCV are used for short-­
term therapy of primary and recurrent HSV infections (particularly genital herpes),
long-term suppressive therapy of recurrent genital herpes, as well as treatment of
herpes zoster. The IV formulation of ACV is indicated for the management of severe
HSV (including encephalitis and neonatal herpes) and VZV infections. Topical for-
mulations of ACV and PCV (Denavir®, Novartis) are used for the treatment of her-
pes labialis and keratitis. Brivudine or bromovinyldeoxyuridine (BVDU) is an
analogue of thymidine that is converted into monophosphate and diphosphate
metabolites by the TK of HSV-1 and VZV and then in its triphosphate form by cel-
lular kinase. BVDU triphosphate competes with dNTPs and also acts by terminat-
ing DNA chain elongation. Brivudine was also shown to inhibit the activity of
thymidylate synthase of VZV [113]. In some European countries, BVDU is
approved for the treatment of herpes zoster in immunocompetent adults [71]. The
pyrophosphate analogue FOS is usually indicated for ACV- or PCV-resistant HSV
or VZV infections [185–187]. Topical and IV formulations of CDV may be used
“off label” in the treatment of nucleoside analogue- and/or FOS-resistant HSV
infections [225].
236 W. L. Drew et al.

8.2 Human Cytomegalovirus Resistance

8.2.1  henotypic and Genotypic Assays to Evaluate HCMV

P
Drug Susceptibility

Two different albeit complementary approaches have been developed to assess

HCMV drug resistance. In the phenotypic method, the virus is grown in the pres-
ence of various concentrations of an antiviral in order to determine the concentra-
tion of drug that will inhibit a percentage (most commonly 50%) of viral growth in
cell culture. In this assay, a standardized viral inoculum is inoculated in different
wells. The virus is then allowed to grow for a few days (typically 7–10 days) in the
presence of serial drug dilutions before staining the cells. The number of viral
plaques per concentration is first determined. Then, the percentage of viral growth,
as compared to a control well without antiviral, is plotted against drug concentra-
tions to determine the concentration that will inhibit the growth of 50% of viral
plaques (50% effective concentration or EC50). Even though recent efforts have
been made to standardize this assay [139], the inter-assay and interlaboratory vari-
ability is still problematic. In addition to the relative subjectivity of this method,
there are some differences in the cutoff values defining drug resistance depending
on the laboratory. Similar assays, either based on detection of HCMV DNA by
hybridization [68] or quantitative PCR [200] or detection of specific HCMV anti-
gens by ELISA [217], flow cytometry [131, 158], immunofluorescence [219] or
immunoperoxidase [100], or real-time cell analysis [180], have also been devel-
oped. EC50 cutoff values defining resistance are still a matter of debate although a
value of 6 μM is most often used for GCV. Altogether, phenotypic assays are time-­
consuming, unstandardized, and subject to possible selection bias introduced during
viral growth of mixed viral populations in cell culture [104, 111] and may lack
sensitivity to detect low-level resistance or minor resistant subpopulations [54,
104].
In contrast to phenotypic assays, which directly measure drug susceptibility of
viral isolates, genotypic assays detect the presence of viral mutations known to be
associated with drug resistance. Those assays are based on DNA sequencing of viral
genes (UL97 and UL54) that are the sites of HCMV resistance to antivirals. One of
the advantages of these assays is that they can be performed directly on clinical
specimens [28, 234, 235], thus reducing considerably the time required to obtain
results. By omitting the need to grow the virus, such methods also minimize the
risks of introducing a selection bias. The limited number of UL97 mutations respon-
sible for GCV resistance has allowed the development of rapid assays to detect their
presence in clinical samples [49, 112]. Indeed, approximately 70% of GCV-resistant
clinical isolates contain mutations in one of three UL97 codons (460, 594, and 595)
[29]. However, due to reports of resistance mutations at other codons, DNA sequence
of the entire UL97 region involved in GCV resistance should be determined for a
comprehensive analysis, and this approach is the current standard (Fig. 8.2).
8 Resistance of Herpesviruses to Antiviral Agents 237

Fig. 8.2 Confirmed mutations associated with resistance to ganciclovir in the UL97 gene and to
ganciclovir (GCVR), foscarnet (FOSR), and/or cidofovir (CDVR) in the UL54 gene of clinical
HCMV isolates. In UL97 gene, the ATP-binding site, the phosphate transfer (P-transfer) domain,
the nucleoside-binding site (NBS), and some regions conserved among the protein kinase family
(i.e., I, II, III, VIB, VII, VIII, and IX) are represented by the black boxes. In UL54 gene, conserved
regions among the Herpesviridae DNA polymerases are represented by the black boxes. The
roman numbers (I to VII) and δ-region C corresponding to each of these regions are indicated
above the boxes. Conserved motifs (Exo I, Exo II, and Exo III) in the exonuclease domain are also
indicated above the boxes. Scale represents nucleotide positions in each gene. Bars (│) indicate
amino acid substitutions associated with antiviral drug resistance. ashaded area corresponds to the
codon 590–603 region in UL97 gene where different amino acid deletions conferring ganciclovir
resistance were identified (i.e., deletions 591–594, 591–607, 595, 595–603, 600, and 601–603).
b
amino acid deletion 981–982 in UL54 gene that confers resistance to all three antivirals

For genotypic analysis of UL54 DNA polymerase mutations, DNA sequencing

is required due to the large number of mutations reported within all conserved
regions of this gene [103] (Fig. 8.2). Genotypic approaches are fast, but their inter-
pretation is not always straightforward, i.e., discriminating between mutations asso-
ciated with natural polymorphisms [31, 52, 94, 152, 154] from those related to drug
resistance. In order to prove that a new mutation is associated with drug resistance,
recombinant viruses need to be generated using either overlapping cosmid/plasmid
inserts [62] or by marker transfer experiments of the mutated gene in a wild-type
[10, 11, 50] or genetically engineered [54, 55] virus prior to testing of this mutant
virus in phenotypic assays. The introduction of a reporter gene (luciferase) in a
permissive cell line [105] or directly in a recombinant virus [56, 79] should acceler-
ate the phenotypic testing of new mutants by allowing an automated and more
objective evaluation of viral replication.
238 W. L. Drew et al.

8.2.2  linical Significance, Incidence, and Risk Factors

C
for Drug-Resistant HCMV Infections

Drug-resistant strains first emerged as a significant problem in patients with AIDS

being treated for HCMV retinitis. Numerous studies have documented the emer-
gence of drug-resistant HCMV strains (detected by phenotypic or genotypic meth-
ods) and their correlation with progressive or recurrent HCMV disease (mainly
retinitis) during therapy [10, 50, 51, 75, 204, 206, 234, 235]. The first study to evalu-
ate the prevalence of GCV resistance in AIDS patients was conducted by evaluating
the excretion of GCV-resistant strains in the urine of 31 patients with AIDS treated
with IV GCV for HCMV retinitis.
In this study, no resistant isolates were recovered in patients treated for
≤3 months, whereas 38% of those excreting the virus in their urine after >3 months
of GCV had a resistant isolate, representing 8% of the entire cohort of patients [73].
Since then, larger studies have evaluated the temporal emergence of GCV-resistant
strains using either phenotypic [120] or genotypic [29] assays. In all studies, GCV
resistance (defined by an EC50 value ≥6 μM) at the initiation of treatment was a rare
event (≤2.7% of tested strains). Phenotypic evaluation of blood or urine isolates
from 95 patients treated with GCV (mostly intravenous) for HCMV retinitis revealed
that 7, 12, 27, and 27% of patients excreted a GCV-resistant strain after, respec-
tively, 3, 6, 9, and 12 months of drug exposure [120]. In a more recent study of 148
AIDS patients treated for HCMV retinitis with oral VGCV, the presence of GCV
resistance mutations was detected in 2, 7, 9, and 13% of patients after 3, 6, 9, and
12 months of therapy, respectively [29]. The lower incidence of GCV resistance in
the latter study despite the use of sensitive genotypic methods might be explained
by differences in the treatment of the underlying HIV disease in the study popula-
tion, notably improved antiretroviral therapy. Due to their less frequent use in clinic,
fewer data have been reported on the temporal emergence of FOS- and CDV-­
resistant HCMV strains in HIV-infected individuals. One small study found an inci-
dence of phenotypic resistance to FOS of 9, 26, 37, and 37% after 3, 6, 9, and
12 months of therapy using an EC50 cutoff value of 400 μM [121], whereas another
one reported rates of 13, 24, and 37% after 6, 9, and 12 months using an EC50 cutoff
value of 600 μM [231]. The data on CDV resistance (EC50 value ≥2–4 μM) are even
more limited, but they seem to indicate a resistance rate similar to what has been
observed with GCV and FOS [121].
Proposed risk factors for the development of HCMV resistance in this patient
population include inadequate tissue drug concentrations due to poor tissue penetra-
tion (e.g., the eyes) or poor bioavailability (e.g., oral GCV), a sustained and pro-
found immunosuppression status (CD4 counts <50 cells/μl), frequent discontinuation
of treatment due to toxicity, and a high pre-therapy HCMV load [77, 168].
HCMV resistance to GCV appears to be an emerging problem in SOT recipients
and has been associated with an increased number of asymptomatic and symptom-
atic viremic episodes, earlier onset of HCMV disease, graft loss, and an increased
risk of death [21, 144]. Due to the different HCMV preventive strategies and
8 Resistance of Herpesviruses to Antiviral Agents 239

i­ mmunosuppressive regimens in use at different centers and considering the hetero-

geneity of the transplant populations, it has been difficult to precisely evaluate the
temporal emergence of HCMV resistance in that setting. Lung transplant recipients
appear to have the highest incidence of HCMV resistance development with rates of
3.6–9% after median cumulative GCV exposures ranging from 79 to 100 days [137,
145, 153]. In two of those studies, the incidence of resistance increased from 15.8%
to 27% in D+/R– (HCMV antibody-positive donor with HCMV-negative recipient)
patients [145, 153] and occurred as a late complication, i.e., a median of 4.4 months
after transplantation [145]. As opposed to what has been reported in lung transplant
recipients, the incidence of GCV resistance in other SOT populations has been
much lower and almost exclusively seen in D+/R– patients [144, 153]. More specifi-
cally, Lurain and colleagues studied two cohorts of SOT patients including heart,
liver, and kidney recipients at two US centers [153]. Phenotypic evaluation for
HCMV resistance prompted by either clinical suspicion or positive blood cultures
indicated that rates of resistance were generally low (e.g., <0.5%) at one center and
varied from 2.2% to 5.6% at another center depending on the transplanted organ.
Another retrospective study by Limaye and colleagues evaluated 240 SOT patients
including 67 D+/R– patients but excluded lung transplant recipients [144]. In their
cohort, GCV-resistant HCMV disease developed only in D+/R– patients, with resis-
tance rates of 7% in these patients. HCMV resistance was more frequently seen
among recipients of kidney/pancreas or pancreas alone (21%) than among kidney
(5%) or liver (0%) recipients. Of note, cases of GCV-resistant HCMV infections
occurred at a median of 10 months after transplantation with a median total drug
exposure of 194 days (129 days of oral GCV) including two to three treatment
courses for HCMV disease per patient. Importantly, GCV-resistant HCMV infec-
tions accounted for 20% of HCMV diseases that occurred during the first year after
transplantation [144]. In a retrospective analysis published in 2008, Eid and col-
leagues reported a similar rate of GCV resistance in D+/R- recipients of SOT (other
than lung) who received VGCV prophylaxis [82].
Boivin and colleagues reported the first prospective study evaluating the emer-
gence of GCV resistance in SOT recipients [30]. In this study, molecular methods
were used to assess the emergence of UL97 and UL54 mutations associated with
GCV resistance in D+/R– patients (175 liver, 120 kidney, 56 heart, 11 kidney/pan-
creas, and 2 liver/kidney recipients) receiving HCMV prophylaxis with either oral
GCV (1 g TID) or oral VGCV (900 mg OD). Among 301 evaluable patients, the
incidence of GCV resistance at the end of the prophylactic period (day 100 post-
transplant) was very low in both arms (0% and 1.9% for the VGCV and oral GCV
arms, respectively). During the first year following transplantation, GCV resistance-­
associated mutations were found in none compared to 6.1% of patients at the time
of suspected HCMV disease after receiving VGCV and oral GCV prophylaxis,
respectively. Of note, however, no lung transplant and a small number of kidney/
pancreas recipients were included in this trial, which might explain at least partly
the low emergence of GCV resistance in this study as compared to previous ones.
Interestingly, several studies have shown that detection of known GCV resistance
mutations is not always associated with adverse clinical consequences in non-lung
240 W. L. Drew et al.

transplants in contrast to more immunosuppressed lung transplants [30, 32].

Documented risk factors for the emergence of GCV resistance in SOT patients
include the lack of HCMV-specific immunity (as encountered in the D+/R– group)
[14, 17], lung or kidney/pancreas transplantation, longer drug exposure (prophy-
laxis > preemptive therapy), suboptimal plasma or tissue drug concentrations (as
seen with oral GCV), potent immunosuppressive regimens, a high HCMV viral
load, and frequent episodes of HCMV disease [21, 144, 145].
Limited data from small-scale studies suggest that the incidence of GCV resis-
tance in the bone marrow transplant (BMT)/HSCT population might not be as high
as observed in SOT recipients and AIDS patients, perhaps because of the more
limited immunosuppression exposure and the greater use of preemptive antiviral
strategies. In a study published by Gilbert et al., molecular methods were used to
detect the presence of the most common UL97 mutations associated with GCV
resistance in blood samples of HSCT patients selected on the basis of having a posi-
tive HCMV PCR despite ≥14 days of preemptive IV GCV or a second viremic
episode within the first 98 days after transplantation [102]. No UL97 mutations
associated with GCV resistance were detected in this cohort of 50 patients (10 of
them fulfilling the above criteria for genotypic testing) [102]. However, this was a
small study, and resistance would be unlikely after just a short period of preemptive
treatment. In another study designed to evaluate risk factors and outcomes associ-
ated with rising HCMV antigenemia levels during the first 2–4 weeks of preemptive
therapy, Nichols and colleagues prospectively evaluated 119 HSCT patients receiv-
ing preemptive GCV or FOS therapy following a positive pp65 antigenemia test
[167]. Among these subjects, 47 (39%) exhibited a significant rise in antigenemia
levels despite antiviral administration, and 15 had at least one isolate available for
susceptibility testing. Only one GCV-resistant isolate was identified in a patient
who received 4 weeks of GCV therapy [167]. In contrast, Erice et al. reported geno-
typic or phenotypic evidences of infection with a GCV-resistant HCMV strain in
two of five selected patients who had received GCV for a median of 58 days [87].
However, all five patients had also received ACV prophylaxis for a median of
47 days which could have predisposed to the selection of a GCV-resistant HCMV
strain [160]. Of note, the impact of prior ACV in selecting for GCV resistance has
not been confirmed by another group [74]. Springer et al. also reported two HSCT
patients who developed persistent and severe drug-resistant HCMV infections,
including one virus with a DNA polymerase mutation conferring multidrug resis-
tance [209]. Even though short courses of GCV therapy are not usually complicated
by emerging resistance in adult BMT patients, Eckle et al. reported that the situation
might differ in pediatric patients receiving T-cell-depleted unrelated transplants
[81]. In their study of 42 such patients, 3 showed genotypic evidences of GCV resis-
tance, followed by the excretion of a resistant strain after 30–93 days of GCV expo-
sure. Of note, in the same study, none of the 37 patients who underwent a similar
procedure, but who received their transplant from a mismatched related donor,
developed GCV resistance [81]. Rapid emergence of GCV resistance was also doc-
umented in four of five children with congenital immunodeficiency disorders who
underwent T-cell-depleted BMT [236]. In those patients, genotypic evidence of
8 Resistance of Herpesviruses to Antiviral Agents 241

GCV resistance was demonstrated after only 7–24 days (median 10 days) of cumu-
lative GCV therapy. Finally, the emergence of GCV-resistant strains has been
recently associated with previously uncommon central nervous system HCMV dis-
ease and retinitis occurring late after HSCT [111, 237].

8.2.3  ole of HCMV UL97 and UL54 Mutations in Drug-­

R
Resistant Clinical Strains

Ganciclovir resistance is mediated by mutations in one or both of the following

genes – UL97 and UL54. UL97 is responsible for phosphorylating ganciclovir, and
UL54 is the DNA polymerase gene. When resistance develops, it is usually initiated
by mutations in UL97 and later mutation(s) in UL54 ensue. Cumulative results
obtained in three studies that have documented the emergence of UL97 mutations in
clinical isolates [123, 153] or in blood samples [29] from 61 AIDS and SOT patients
are in general agreement with the proposed frequency of UL97 mutations based on
characterization of 76 independent UL97 mutants gathered in a single laboratory
over years [54]. Those data suggest that mutations A594V (30–34.5%), L595S (20–
24%), M460V (11.5–14.5%), and H520Q (5–11.5%) represent the most frequent
UL97 mutations present in GCV-resistant mutants (Fig. 8.2) [76, 106]. Other fre-
quent UL97 mutations associated with resistance include C592G and C603W. Based
on marker transfer experiments, mutations M460V (7× increase in resistance) [49]
and C603W (8×) [50], deletion of codons 595–603 (8.4×) [53], H520Q (10×) [112],
L595S (4.9–11.5×) [49, 54], A594V (10.7×) [49], and C607Y (12.5×) [12], and
deletion of codon 595 (13.3×) [10] appear to be associated with the highest increase
in GCV resistance over the parental strain, whereas mutations C592G, A594T, and
E596G and deletion of codon 600 seem to confer only modest decreases in suscep-
tibility [54]. Interestingly, analysis of GCV-phosphorylating activity of mutated
UL97 genes expressed in a recombinant vaccinia virus expression system would
have predicted mutations H520Q and M460V to confer the greatest decrease in
GCV susceptibility [13].
Among the most frequent DNA pol mutations associated with drug resistance,
there are V715M, V781I, and L802M conferring FOS resistance and F412C,
L501I/F, and P522S conferring GCV/CDV resistance (Fig. 8.2) [76, 106]. Mutation
A809V conferring GCV/FOS resistance has also been reported with some fre-
quency. Importantly, some mutations (E756K, V812L, and del981–982) have been
associated with resistance to all three antivirals, but these mutations are rarely
encountered [55, 62]. With regard to the levels of resistance, mutations L501I and
K513N and deletion of codons 981–982 have been associated with a six- to eight-
fold decrease in GCV susceptibility [55, 62, 63] and mutations F412C/V, K513N,
and A987G with a 10–18-fold decrease in CDV susceptibility [50, 62, 63], whereas
mutations D588N, V715M, E756K, L802M, and T821I seem to confer a 5.5–21-­
fold increased resistance to FOS [11, 50, 55, 62, 164]. A few UL54 mutations have
242 W. L. Drew et al.

been studied in marker transfer experiments for their effect on viral fitness. Among
those, mutations T700A and V715M (conserved region II) [11], K513N (δ-region
C) [63], and D301N (Exo I motif) [55] were shown to significantly reduce the yield
of progeny virus in cell culture supernatants, whereas some others (D413E, T503I,
L516R, and E756K/D) were only associated with a modest attenuation of viral rep-
lication [55]. In the case of HCMV DNA pol mutants selected during GCV therapy,
it should be noted that UL97 mutations have been generally shown to emerge first
and to confer a low level of resistance (EC50 < 30 μM), whereas subsequent emer-
gence of UL54 mutations usually leads to a high level of drug resistance
(EC50 > 30 μM) [86, 122, 205]. Most clinical isolates resistant to GCV have muta-
tions in UL97 only. However, isolated UL54 mutations have been reported occa-
sionally in clinical HCMV strains [31].

8.2.4  anagement of Infections Caused by Drug-Resistant

M
HCMV Strains

HCMV resistance to antivirals should be suspected in patients failing treatment who

have been exposed to an antiviral for substantial periods of time (typically
>3–4 months of treatment in AIDS patients and after 2–3 months or more of pro-
phylaxis or treatment in transplant recipients), especially if some risk factors are
present (i.e., D+/R– SOT, lung or kidney/pancreas transplant, AIDS patients with
CD4 counts <50 cells/μl). Resistance should be suspected in pediatric patients with
shorter periods of drug exposure if they had T-cell depletion. Clinical resistance is
more likely if active viral replication (high or increasing levels of DNAemia/anti-
genemia or viremia) persists or recurs despite maximum IV doses of the antivirals
[145, 166]. On the other hand, rising CMV DNA or antigenemia levels during the
first 2–3 weeks of antiviral therapy in HSCT recipients have not been associated
with antiviral resistance but rather with host and other transplant-related factors
[101, 166]. Whenever antiviral resistance is suspected, phenotypic and/or genotypic
investigation for resistance should be undertaken. As discussed above, genotypic
methods are fast and more convenient and provide useful information for selection
of an alternative treatment. However, identification of mutations of unknown sig-
nificance remains problematic and may require phenotypic assays for validation.
Furthermore, genotypic assays do not quantitate the degree of resistance while phe-
notypic assays do. The choice of the sample to analyze may also have some impor-
tance. Some studies have reported that there is a good correlation between genotypes
detected in the eyes and the blood (93.5%) [116] or between blood and urine iso-
lates (87.5%) [122] of AIDS patients with HCMV retinitis. However, there have
been at least some reports of resistant HCMV strains restricted to specific body
compartments [81, 148]. This suggests that resistance assessment based solely on
only one bodily fluid or tissue may be misleading in some cases.
8 Resistance of Herpesviruses to Antiviral Agents 243

Fig. 8.3 Suggested algorithm for the management of suspected drug-resistant HCMV infections
in solid organ transplant recipients. Key: GCV ganciclovir, FOS foscarnet, CDV cidofovir, BID
twice a day, IV intravenous, EC50, concentration of antiviral that reduces HCMV replication in
cultured cells by 50% compared to control (without drug) determined in phenotypic assay.
(Adapted from [134])

As mentioned, resistance is more likely when stable or rising viral loads (espe-
cially DNAemia levels) or persistence of clinical symptoms are observed more than
2–3 weeks after initiating appropriate full-dose IV antiviral therapy. In this context,
clinical decisions on disease management should be based on genotypic analysis of
UL97 and UL54 genes, the patient’s immune status (e.g., high-risk D+/R– recipi-
ents, lung transplant recipients), and disease severity (i.e., sight- or life-threatening
conditions) [77, 145] (Fig. 8.3). Despite the limitation mentioned above, genotypic
resistance testing is more practical and rapid (results in 72–96 h) than phenotypic
assays. Thus, rescue therapy should be ideally based on results of the genotypic
assays. In centers where genotypic testing is unavailable or performed infrequently,
initial management should avoid the use of drugs with similar pathways of resis-
tance. For instance, patients failing GCV should be given FOS in the absence of any
sequencing data because of the possibility of UL54 mutations that could confer
resistance to both GCV and CDV. On the other hand, if UL97 and UL54 sequencing
data are available and indicate that only UL97 mutations are present, then CDV
therapy can be attempted. Other empiric options for patients failing GCV therapy
244 W. L. Drew et al.

could consist in re-inducing the patient with higher than normal doses of GCV (up
to 10 mg/kg IV BID) or combination therapy with reduced doses of GCV and FOS
[165, 208] although these strategies are associated with significant toxicity and can
be clinically risky in patients with life- or sight-threatening diseases. Leflunomide,
an anti-inflammatory compound, which appears to inhibit viral capsid assembly, has
not been systemically evaluated but is the subject of successful case reports [182].
When GCV resistance is encountered, discontinuation of the drug and the use of
foscarnet alone may hasten return to “wild,” i.e., sensitive virus [78]. Whenever pos-
sible, improvement of the patient’s immune status (i.e., reduction of immunosup-
pressive regimen in transplant patients or aggressive antiretroviral therapy in AIDS
patients) should also be considered. HCMV viral load should be carefully moni-
tored (once weekly) to quantitate a response to the change in therapy.

8.3  erpes Simplex Virus and Varicella-Zoster Virus

H
Resistance

8.3.1  henotypic and Genotypic Assays to Evaluate HSV

P
and VZV Drug Susceptibility

Four mechanisms are involved in HSV resistance to ACV and lead to different phe-
notypes: (i) a complete deficiency in viral TK activity (TK-deficient); (ii) a decreased
production of viral TK (TK low producer); (iii) a viral TK protein with altered sub-
strate specificity (TK altered), i.e., the enzyme is able to phosphorylate thymidine,
the natural substrate, but does not phosphorylate ACV; and finally (iv) a viral DNA
pol with altered substrate specificity (DNA pol altered). Alteration or absence of the
TK protein is the most frequent mechanism seen in the clinic, probably because TK
is not essential for viral replication in most tissues and cultured cells [99, 162].
Thymidine kinase mutants resistant to ACV usually exhibit a reduction in fitness
and neurovirulence. In animal models, TK low producers show some reduction in
pathogenicity compared with wild-type strains but are generally able to reactivate.
In contrast, TK-deficient mutants demonstrate impaired pathogenicity as well as a
lower efficiency to establish latency in sensory ganglia and a poor reactivation com-
pared with wild-type strains. However, it has been suggested that ultralow levels of
TK enzyme activity could be sufficient to allow reactivation [18]. Mutants with
altered DNA pol activity exhibit different degrees of neurovirulence attenuation in
animals [2, 67].
The TK phenotype can be determined by the selective incorporation of iodode-
oxycytidine and thymidine into infected cells using plaque autoradiography [156].
However, it can be difficult to evaluate residual TK activity in HSV isolates of
immunocompromised patients in whom heterogeneous populations (TK-competent/
TK-deficient) may coexist [99, 191, 223]. Nonisotopic methods have been devel-
oped using ADP-Glo kinase assay [39, 174] and DiviTum assay based on
8 Resistance of Herpesviruses to Antiviral Agents 245

b­ romodeoxyuridine as substrate [193, 194] to evaluate ACV phosphorylation and

thymidine kinase activity, respectively. However, both methods do not distinguish
between the three different TK phenotypes. More interestingly, thymidine kinase
functionality was assessed by measuring monophosphate forms of both ACV and
thymidine by high-pressure liquid chromatography with diode-array detection
[155]. Although such enzymatic assays do not allow the determination of the resis-
tance levels conferred by a specific TK mutation, they may facilitate the discrimina-
tion between resistance-associated mutations and polymorphic alterations.
Only approximately 5–10% of ACV-resistant HSV strains have polymerase
mutations. Resistance to FOS and to CDV is conferred by specific mutations within
the viral DNA pol which is the ultimate target of all current antiviral drugs.
Depending on the locus of the pol mutation, there may or may not be cross-­resistance
between ACV, FOS, and CDV [20]. At the present time, no simple enzymatic assay
has been described to rapidly assess the DNA pol activity of herpesviruses.
Levels of drug resistance (EC50 values) are best measured by cell-based (pheno-
typic) assays. Such assays are more practical in the case of HSV (and to some extent
VZV) than for HCMV considering the more rapid replication kinetics of the former
viruses. The gold standard phenotypic method to determine the susceptibility of
HSV isolates to antiviral drugs is the plaque reduction assay in Vero cells that is
approved as a standard protocol by the Clinical and Laboratory Standards Institute
(Wayne, PA) [215]. Herpes simplex virus resistance cutoffs to ACV have varied in
the literature from 4.4 μM to 13.2 μM according to the method selected, i.e., plaque
reduction or dye uptake assays and various other factors (although a cutoff of 2 μg/
ml or 8.8 μM is mostly used with the plaque reduction assay) [44, 65, 84, 89, 188].
Due to this variability, a susceptibility index is said to be a better measure of viral
resistance. The ratio of the EC50 of the patient’s isolate should be at least three times
or greater than the EC50 of a known, sensitive HSV control [190].
Susceptibility of VZV to ACV can be tested in plaque reduction assays using
fibroblastic cell lines such as MRC-5 [93]. The use of the plaque reduction assay is
limited by the low rate of VZV isolation from vesicle samples (from 20% to 43%)
and its slow growth in cell culture (typically 5–6 days) [195]. The end point for
detecting resistance is a susceptibility index greater than or equal to four, i.e., the
test strain has an EC50 greater than four times that of a control, known sensitive
strain, e.g., the Oka strain. Regarding absolute values, three resistant strains from a
single series had mean ACV EC50 values of 85 μM vs 3.3 μM for the Oka strain
[189].
An alternative to phenotypic assays is genotyping by sequence analysis. For a
comprehensive genotypic analysis, the whole TK gene and the conserved regions of
the DNA pol gene of HSV or VZV should be sequenced because of the large num-
ber of TK mutations (substitutions, deletions, and additions) as well as DNA pol
mutations associated with drug resistance [163]. The development of fast and effi-
cient methods for detecting viral mutant sequences directly in clinical specimens by
next-generation sequencing [128] should improve the evaluation of heterogeneity
and temporal changes that occur in populations of drug-resistant viruses during
antiviral therapy [4]. As some degree of inter-strain variability exists in these genes,
246 W. L. Drew et al.

mutations conferring drug resistance must be discriminated from natural polymor-

phisms. Different systems can be used to generate recombinant HSV or VZV and
evaluate their phenotype of drug resistance, such as the transfection of a set of over-
lapping cosmids and plasmids allowing rapid site-directed mutagenesis [20, 202]
and the cloning of the viral genome coupled with a reporter gene expressing a fluo-
rescent protein into a bacterial artificial chromosome [36, 37]. Compilations of con-
firmed drug resistance mutations and natural polymorphisms in the TK and DNA
pol genes of HSV [196] and VZV [177] are described in several reviews.

8.3.2  linical Significance, Incidence, and Risk Factors

C
for Drug-Resistant HSV and VZV Infections

Antiviral drugs against herpesviruses provide some of the best examples of effective
and selective antiviral therapy. However, drug-resistant viruses have been rapidly
selected in the laboratory and also identified in the clinic. Contrasting with HCMV
resistance data, no extensive survey has been performed to evaluate the rate of emer-
gence of drug-resistant HSV isolates according to the duration of antiviral therapy.
Such study would be a difficult task considering that oral and topical ACV formula-
tions are widely used.
In immunocompetent hosts, HSV resistance to ACV is not a clinically important
problem. Studies have shown that 0.1–0.6% of HSV isolates recovered from
untreated, prophylaxed, or treated immunocompetent subjects harbor a resistant
phenotype to ACV (EC50 ≥ 8.8 μM) as assessed by a plaque reduction assay, and
this seems to reflect the natural occurrence of TK-deficient mutants in a viral popu-
lation [8, 9, 33, 60, 69, 91, 159, 211, 232]. Except for a few notable cases [133, 136,
214], the occasional recovery of ACV-resistant HSV-2 from immunocompetent
hosts has not been associated with clinical failure and proved to be transient [110,
232]. A higher prevalence (6.4%) of ACV-resistant HSV-1 isolates has been reported
in immunocompetent patients with recurrent herpetic keratitis [80], and some of
these cases were clinically refractory to ACV therapy [41, 125, 170, 224]. The lower
immune surveillance in the cornea, which is an immune-privileged site, could
explain the rapid selection of drug-resistant viruses [6]. Herpes simplex virus strains
resistant to ACV are more often isolated in immunocompromised hosts, and such
isolates have been associated with persistent and/or disseminated diseases [26, 47,
60, 84, 89, 114, 162, 184, 228]. In the few clinical surveys reported, the rate of
ACV-resistant HSV isolates has varied from 4.3% to 14% among all immunocom-
promised groups [60, 69, 72, 84, 143, 169, 181, 211]. The prevalence of ACV resis-
tance has ranged from 3.5% to 7% in HIV-positive patients in several studies [60,
84, 143, 149, 181, 239]. It is estimated that 6.5% of HSV isolates obtained from
patients with cancer were resistant to ACV compared to 10% from heart or lung
transplant recipients [60] and 5–14% from other SOT recipients [84]. Of note, high
resistance rates have been reported in HSCT recipients, ranging from 4.1% to 10.9%
8 Resistance of Herpesviruses to Antiviral Agents 247

[47, 69, 85, 96, 162, 229, 233]. In another study, 8% of allogeneic cell transplant
recipients demonstrated persistent HSV excretion despite ACV therapy, whereas
5% of HSV isolates showed significant level of ACV resistance in vitro [42]. Morfin
and Thouvenot reported that patients receiving either autologous or allogenic bone
marrow have a similar incidence, i.e., 9%, of HSV infection, but resistance only
occurred in allogenic transplants, reaching a prevalence of 30% [163]. The severity
of immunosuppression and the prolonged use of ACV are considered two important
factors for the development of drug resistance. The importance of the severity of
immunosuppression is underscored by Langston et al. who studied adult patients
undergoing lymphocyte-depleted hematopoietic progenitor cell transplant from
HLA-matched family donors [140]. All seven evaluable HSV-1 or HSV-2 seroposi-
tive patients reactivated at a median of 40 days posttransplant, and the five strains
tested were all resistant to ACV. Furthermore, FOS resistance developed rapidly in
the three patients treated with this drug [140]. Importantly, the prevalence of ACV-­
resistant HSV isolates has remained stable in immunocompromised patients over
the past decades [69, 211], and there has been no unequivocal evidence of transmis-
sion of a resistant HSV strain from person to person.
The emergence of VZV isolates resistant to ACV has not been described in
immunocompetent individuals with primary VZV infection or herpes zoster, except
for one case report of a patient with an ACV-resistant VZV keratitis [109]. Acyclovir-­
resistant VZV isolates in the clinic have been mainly found in AIDS patients with
low CD4 cell counts who presented with atypical, disseminated, or relapsing zoster
lesions [27, 161, 189, 216, 226]. Cases of resistance to ACV have also been described
in SOT and HSCT recipients as well as in hemato-oncological patients with VZV
reactivations unresponsive to therapy. In these patients, VZV infections not respond-
ing to ACV therapy persist in the form of chronic skin lesions and are associated
with significant morbidity and mortality due to visceral dissemination. An unusual
verrucous form of VZV infection caused by ACV-resistant mutants has also been
described in some patients [38, 66]. Two cases of immunocompromised children
presenting herpes zoster due to the Oka vaccine strain and who developed chronic,
disseminated drug-resistant VZV infections following ACV therapy have been
reported [38, 142]. However, the prevalence of ACV-resistant cases in these differ-
ent populations is unknown because only case reports have been published so far. It
was reported that 27% of hemato-oncological patients, including HSCT recipients,
with persistent VZV infections had mutations probably associated with ACV resis-
tance [222].
Only a few FOS-resistant HSV (EC50 ≥ 330 μM or ≥ 100 μg/ml and at least a
threefold increase in EC50 value compared with the parental susceptible strain) have
been reported in the clinic [64, 119, 199]. Nine FOS-resistant HSV clinical isolates
from HIV-infected subjects in whom ACV and FOS therapy sequentially failed have
been described [19, 199]. A few reports have described the emergence of VZV
strains resistant to FOS in immunocompromised patients [16, 92, 186, 226, 227].
248 W. L. Drew et al.

Fig. 8.4 Confirmed mutations associated with acyclovir resistance in the UL23 and ORF36 gene
of clinical HSV-1, HSV-2, and VZV isolates. Conserved regions among the thymidine kinases of
Herpesviridae including the ATP-binding site (ATP) and the nucleoside-binding site (NBS) are
represented by the black boxes. Scale represents nucleotide positions in the gene. Bars (│) indicate
amino acid substitutions, whereas dots (●) represent nucleotide additions and/or deletions that
confer resistance to acyclovir

8.3.3  ole of Viral TK and DNA Polymerase Mutations

R
in Drug-Resistant Clinical Strains

Mutations in the TK of HSV (encoded by the UL23 gene) leading to ACV resistance
consist of either additions or deletions in homopolymer runs of Gs and Cs associ-
ated with a premature stop codon or single nucleotide substitutions in conserved and
non-conserved regions of the gene (Fig. 8.4) [176, 177]. Each mechanism accounts
for approximately 50% of ACV-resistant phenotypes in the clinic [99, 162].
However, recent studies reported an increased proportion of additions/deletions
which accounted for 62% [97] or even 80% [40] of TK gene mutations. Nucleotide
substitutions are scattered within the TK gene including the three catalytic sites of
the enzyme (ATP-binding site, nucleoside-binding site, and a.a. 336) [70]. Albeit
rarely seen in clinic, most mutations in the DNA polymerase of HSV (encoded by
the UL30 gene) conferring drug resistance are located in the conserved regions of
the enzyme, most specifically in regions II, VI, III, and VII (Fig. 8.5) which are
directly or indirectly involved in the recognition and binding of nucleotides or pyro-
phosphate as well as in catalysis [176, 177]. The greatest clusters of mutations in the
DNA pol enzyme have been found in conserved regions II and III. Most FOS-­
resistant clinical HSV isolates contain single-base substitutions in conserved regions
II, VI, III, or VII and in a non-conserved region (between I and VII) of the DNA pol.
Some of these isolates retain susceptibility or borderline levels of susceptibility to
ACV and CDV. However, some mutations, in particular in regions II (V715G and
S724N) and VII (Y941H) of the DNA pol, can confer resistance to both ACV and
8 Resistance of Herpesviruses to Antiviral Agents 249

Fig. 8.5 Confirmed mutations associated with resistance to acyclovir (ACVR), foscarnet (FOSR),
and/or cidofovir (CDVR) in the UL30 and ORF28 genes of clinical HSV-1, HSV-2, and VZV iso-
lates. Conserved regions among the Herpesviridae DNA polymerases are represented by the black
boxes. The roman numbers (I to VII) and δ-region C or region A corresponding to each of these
regions are indicated above the boxes. Scale represents nucleotide positions in the gene. Colored
bars (│) indicate amino acid substitutions

FOS. Mutations associated with CDV resistance mapped in DNA pol regions II
(R700M), VI (L773M), III (G841C and G850I), and VII (Y941H) and in δ-region C
located in the Exo III motif (V573M).
A recent study of Topalis et al. showed that mutations conferring drug resistance
in DNA pol of HCMV are mostly detected in the 3′-5′ exonuclease domain (60.6%)
and to a lower extent in the palm (18.2%), fingers (16.7%), and thumb (4.6%)
domains, whereas those identified in DNA pol of HSV are mainly located in the
palm (25.0%), fingers (25.0%), and thumb (21.5%) domains with a lower propor-
tion of mutations being found in the 3′-5′ exonuclease domain (27.3%) [220]. The
different distribution of mutations in DNA pol domains may reflect various mecha-
nisms of drug resistance. Mutations conferring resistance to nucleoside analogues
located within conserved regions of the pol domain might reduce the binding of the
inhibitor or the incorporation of the active drug into growing DNA [117]. It has
been suggested that mutations conferring resistance to nucleoside analogues located
in the exonuclease domain might enhance the rate of excision of the incorporated
drug [55]. However, a recent study demonstrated that mutant HCMV with reduced
exonuclease activity might efficiently synthetize DNA in the absence of drug exci-
sion [46]. It has been proposed that resistance of HSV-1 and HCMV to FOS may
250 W. L. Drew et al.

result from subtle conformational changes in the DNA pol that adopts a more open
conformation to which the drug binds with a lower affinity [178, 218, 238].
In VZV clinical isolates, resistance to ACV is mostly associated with mutations
in the viral TK (encoded by the ORF36 gene) and, less frequently, with mutations
in the viral DNA pol (encoded by the ORF28 gene) [177, 179]. The string of six
cytosines located at codons 493–498 of the TK gene emerged as a hot spot for the
insertion or deletion of nucleotides involved in ACV resistance (Fig. 8.4) [3, 27,
161, 222]. Deletions of nucleotides that result in frameshift reading leading to a stop
codon at position 231 are often detected in ACV-resistant VZV clinical isolates
[161]. In addition, nonsynonymous nucleotide substitutions conferring resistance to
ACV are widely dispersed in the TK gene. However, these amino acid changes
occur more frequently in the ATP-binding and nucleoside-binding sites of the TK
enzyme [27, 93, 161, 189, 192, 197, 216]. A few reports have described ACV-­
resistant and/or FOS-resistant VZV clinical isolates with mutations in the DNA pol
gene (Fig. 8.5) [127, 192, 226]. The amino acid substitutions are mainly found in
the catalytic site and in the conserved regions of the DNA pol and may confer cross-­
resistance to ACV and FOS [177, 179]. The TK and DNA pol genes of VZV are
highly conserved compared with those of HSVs, and only a few natural polymor-
phisms have been identified in these genes [192].

8.3.4  anagement of Infections Caused by Drug-Resistant

M
HSV and VZV Strains

With the emergence of ACV-resistant HSV infections observed in patients with

AIDS and other immunocompromised hosts, several studies have examined the util-
ity of alternative antiviral agents and treatment regimens. Standard doses of oral
ACV have no clinical benefit if the HSV isolate is resistant to ACV in vitro. Most
ACV-resistant strains isolated from immunocompromised patients are TK-deficient
and are therefore also resistant to PCV and its prodrug FCV. The persistence of
active lesions due to HSV for more than 7–10 days after initiation of high-dose oral
ACV, VACV, or FCV therapy without apparent decrease in size, an atypical appear-
ance, or the emergence of satellite lesions is suggestive of treatment failure
(Fig. 8.6). In the presence of suspected or confirmed resistance to ACV, the initial
options are either to switch to high-dose oral VACV or IV ACV (10 mg/kg of body
weight every 8 h adjusted for renal function). Acyclovir-resistant HSV strains
remain usually susceptible in vitro to vidarabine (a purine nucleoside analogue),
which is phosphorylated without TK and appears to interfere with the early steps of
viral DNA synthesis, and to FOS, which does not require phosphorylation for activ-
ity. Studies have confirmed that FOS is superior to vidarabine in the treatment of
these TK-deficient, drug-resistant HSV infections [43, 88, 187]. If lesions do not
begin to respond to high-dose oral VACV or IV ACV within 5–7 days, a switch to
IV FOS (40 mg/kg every 8 h with reduction in dose for renal dysfunction) should be
8 Resistance of Herpesviruses to Antiviral Agents 251

Fig. 8.6 Proposed algorithm for the management of suspected nucleoside analogue-resistant HSV
infections. Key: ACV acyclovir, VACV valacyclovir, FCV famciclovir, FOS foscarnet, CDV cido-
fovir, TID thrice a day, IV intravenous

considered. In parallel, isolates from the lesions should be submitted for susceptibil-
ity testing and/or genotypic analysis of the TK gene. If there is still no improvement
of HSV disease after 7–10 days, continuous infusion of high-dose ACV at a dosage
of 1.5–2.0 mg/kg/h for 6 weeks could be initiated, as it is a well-tolerated option for
severe ACV-resistant or multidrug-resistant HSV infections [83, 132]. Cidofovir
(5 mg/kg once a week for 3–4 weeks) could be considered, as it has shown some
efficacy in the treatment of progressive multidrug-resistant mucocutaneous HSV
infection in immunocompromised patients [150, 207], but it is not approved for this
indication. Although topical formulations of FOS [126] and CDV [138, 183] were
effective in the treatment of mucocutaneous infections not responding to ACV and
could avoid the adverse effects associated with their IV administration, they are not
commercially available. A topical formulation containing 5% imiquimod, an
252 W. L. Drew et al.

immunomodulatory drug, was effective in the treatment of recurrent and severe

mucocutaneous lesions due to ACV-resistant HSV-2 isolates in HIV-infected indi-
viduals [141]. Ophthalmic ointment containing 3% vidarabine is indicated for the
treatment of acute keratoconjunctivitis and recurrent epithelial keratitis due to
HSV. A final option for topical therapy is trifluorothymidine (a fluorinated deoxy-
uridine analogue that inhibits thymidylate synthase) as an ophthalmic solution
which may be applied to the affected area three to four times a day until the lesion
is completely healed [48, 130].
As with many opportunistic infections in AIDS patients, there is a high incidence
of recurrent HSV disease after successful treatment of drug-resistant HSV. Some
(but not all) relapses in this setting have been due to drug-resistant strains, suggest-
ing that these mutant viruses are capable of causing latency in the immunocompro-
mised host. Chronic prophylaxis with daily ACV, VACV, FCV, or FOS can be
considered in patients who have been treated successfully for drug-resistant HSV,
although there are no data to confirm efficacy in this setting. Foscarnet-resistant
strains of HSV have been reported, raising concerns over the possible selection for
multidrug-resistant HSV with suppressive therapy [199, 207].
Drug-resistant VZV strains have been identified in patients with AIDS, SOT, HSCT,
and hemato-oncological patients. These patients may present with atypical-­appearing
cutaneous lesions that shed VZV intermittently despite ongoing high-­dose antiviral
therapy. Visceral dissemination of the infection could also lead to significant morbidity
and mortality in these patients. The persistence of clinical signs of VZV infections for
more than 10–14 days after initiation of high-dose oral ACV is suggestive of treatment
failure, and it should lead to alternate therapy depending on the clinical severity of the
disease [1]. Strains have been isolated from patients previously treated with ACV for
recurrent VZV or HSV infection, and these strains may be resistant to ACV, VACV,
and FCV by deficiency of the TK enzyme [27, 124, 161]. Genotypic testing of the TK
gene could be performed in samples from vesicular fluids, biopsy of mucocutaneous
lesions, or other body compartments when necessary [35]. Foscarnet has been shown
to be effective in small studies conducted mainly in HIV-infected individuals [34, 186]
and some oncology patients [38, 66, 142], but, as with HSV, cross-resistance between
ACV and FOS may occur due to viral DNA polymerase mutations [186]. The IV dos-
age recommended for FOS is 60 mg/kg three times daily for at least 10 days or until
complete lesion healing is observed [1]. Clinical experience with the use of CDV in
the treatment of drug-resistant VZV diseases is very limited [198].

8.4 Conclusions and Future Directions

All currently available antiviral agents target the viral DNA pol. The development of
new anti-herpetic compounds with different mechanisms of action and with adequate
safety profile is urgently needed. Some promising compounds are currently in clinical
trials. The orally bioavailable lipid ester prodrug of CDV (hexadecyloxypropyl-cido-
fovir; brincidofovir) could avoid the dose-limiting toxicity of the parent drug and
8 Resistance of Herpesviruses to Antiviral Agents 253

provide a safe alternative for nucleoside analogue-­resistant herpesviruses in immuno-

compromised patients [115]. Treatment with oral brincidofovir significantly reduced
the incidence of HCMV infections in HSCT recipients in a phase II study [157].
Diarrhea was a dose-limiting adverse event in this population at a dose of 200 mg
twice weekly. Maribavir is a competitive inhibitor of the UL97 kinase [23]. Mutations
selected in vitro with maribavir often map to the UL97 gene; low-level resistance
mutations are also detected in the UL27 gene and seem to be the result of a loss of
UL97 kinase activity. Mutations in the UL97 gene conferring resistance to maribavir
are generally distinct from those described in GCV-resistant strains, and some have
been detected outside the conserved kinase domains [57]. Thus, maribavir retains
activity against most GCV-resistant HCMV mutants. However, some mutations, such
as mutation F342S which is located in the p-loop, confer cross-resistance to GCV and
maribavir [58]. The emergence of resistance to this drug has been reported in some
clinical cases [201, 212]. Letermovir targets the terminase complex of HCMV and
interferes with viral DNA concatemer maturation [107, 147]. Accordingly, mutations
conferring resistance to letermovir map to the UL56 gene encoding the HCMV termi-
nase [108]. Successful treatment of a multidrug-resistant HCMV infection with leter-
movir has been reported in a lung transplant recipient [129]. Preemptive treatment of
HCMV infection with letermovir was effective in kidney transplant recipients [210].
Moreover, prophylaxis with letermovir was effective in reducing the incidence of
HCMV infection in HSCT recipients [45]. Pritelivir, a potent orally bioavailable heli-
case-primase inhibitor, reduced the rate of genital HSV-2 shedding and days with
lesions in a phase II trial [230]. A humanized monoclonal antibody was shown to be
effective for immunotherapy of severe HSV infections, including those caused by
multidrug-resistant isolates, in immunocompromised mice and warrants further clini-
cal developments [135]. The bicyclic nucleoside analogue FV-100 and carboxylic
nucleoside analogue valomaciclovir were well tolerated and effective for the treatment
of herpes zoster in phase II trials [5, 221]. Novel classes of antiviral agents targeting
the ribonucleotide reductase, the helicase-primase complex, and the process of viral
DNA encapsidation are at earlier stages of development [90].

Major Points
• Resistance of herpesviruses to antiviral drugs is mostly detected in immunocom-
promised patients but it is increasingly recognized in immunocompetent indi-
viduals with herpetic keratitis.
• Genotypic testing is more frequently used for the detection of antiviral drug
resistance in herpesvirus infections.
• Interpretation of genotypic testing requires a database linking amino acid changes
to mutations associated with natural polymorphisms or drug resistance.
• Algorithms are proposed for the management of infections caused by drug-resis-
tant herpesvirus strains.
• Novel antiviral agents acting on viral targets other than the viral DNA poly-
merase are in development for the treatment of herpesvirus infections.
254 W. L. Drew et al.

Acknowledgments This study was supported by a Foundation Grant from the Canadian Institutes
of Health Research (grant no. 148361 to G.B.). G.B. is the holder of the Canada research chair on
emerging viruses and antiviral resistance.

References

1. Ahmed AM, Brantley JS, Madkan V, Mendoza N, Tyring SK. Managing herpes zoster in
immunocompromised patients. Herpes. 2007;14:32–6.
2. Andrei G, Fiten P, Froeyen M, De Clercq E, Opdenakker G, Snoeck R. DNA polymerase
mutations in drug-resistant herpes simplex virus mutants determine in vivo neurovirulence
and drug-enzyme interactions. Antivir Ther. 2007;12:719–32.
3. Andrei G, Topalis D, Fiten P, McGuigan C, Balzarini J, Opdenakker G, et al. In vitro-selected
drug-resistant varicella-zoster virus mutants in the thymidine kinase and DNA polymerase
genes yield novel phenotype-genotype associations and highlight differences between anti-
herpesvirus drugs. J Virol. 2012;86:2641–52.
4. Andrei G, Georgala A, Topalis D, Fiten P, Aoun M, Opdenakker G, et al. Heterogeneity and
evolution of thymidine kinase and DNA polymerase mutants of herpes simplex virus type 1:
implications for antiviral therapy. J Infect Dis. 2013;207:1295–305.
5. Andrei G, Snoeck R. Advances in the treatment of varicella-zoster virus infections. Adv
Pharmacol. 2013;67:107–68.
6. Andrei G, Snoeck R. Herpes simplex virus drug-resistance: new mutations and insights. Curr
Opin Infect Dis. 2013;26:551–60.
7. Arthurs SK, Eid AJ, Pedersen RA, Kremers WK, Cosio FG, Patel R, et al. Delayed-onset
primary cytomegalovirus disease and the risk of allograft failure and mortality after kidney
transplantation. Clin Infect Dis. 2008;46:840.
8. Bacon TH, Boon RJ, Schultz M, Hodges-Savola C. Surveillance for antiviral-agent-resistant
herpes simplex virus in the general population with recurrent herpes labialis. Antimicrob
Agents Chemother. 2002;46:3042–4.
9. Bacon TH, Levin MJ, Leary JJ, Sarisky RT, Sutton D. Herpes simplex virus resistance
to acyclovir and penciclovir after two decades of antiviral therapy. Clin Microbiol Rev.
2003;16:114–28.
10. Baldanti F, Silini E, Sarasini A, Talarico CL, Stanat SC, Biron KK, et al. A three-nucleotide
deletion in the UL97 open reading frame is responsible for the ganciclovir resistance of a
human cytomegalovirus clinical isolate. J Virol. 1995;69:796–800.
11. Baldanti F, Underwood MR, Stanat SC, Biron KK, Chou S, Sarasini A, et al. Single amino
acid changes in the DNA polymerase confer foscarnet resistance and slow-growth phenotype,
while mutations in the UL97-encoded phosphotransferase confer ganciclovir resistance in
three double-resistant human cytomegalovirus strains recovered from patients with AIDS. J
Virol. 1996;70:1390–5.
12. Baldanti F, Underwood MR, Talarico CL, Simoncini L, Sarasini A, Biron KK, et al. The
Cys607-->Tyr change in the UL97 phosphotransferase confers ganciclovir resistance to
two human cytomegalovirus strains recovered from two immunocompromised patients.
Antimicrob Agents Chemother. 1998;42:444–6.
13. Baldanti F, Michel D, Simoncini L, Heuschmid M, Zimmermann A, Minisini R, et al.
Mutations in the UL97 ORF of ganciclovir-resistant clinical cytomegalovirus isolates differ-
entially affect GCV phosphorylation as determined in a recombinant vaccinia virus system.
Antivir Res. 2002;54:59–67.
14. Baldanti F, Lilleri D, Campanini G, Comolli G, Ridolfo AL, Rusconi S, et al. Human cyto-
megalovirus double resistance in a donor-positive/recipient-negative lung transplant patient
8 Resistance of Herpesviruses to Antiviral Agents 255

with an impaired CD4-mediated specific immune response. J Antimicrob Chemother.

2004;53:536–9.
15. Balfour HH Jr. Antiviral drugs. N Engl J Med. 1999;340:1255–68.
16. Bendel AE, Gross TG, Woods WG, Edelman CK, Balfour HH Jr. Failure of foscarnet in dis-
seminated herpes zoster. Lancet. 1993;341:1342.
17. Benz C, Holz G, Michel D, Awerkiew S, Dries V, Stippel D, et al. Viral escape and T-cell
immunity during ganciclovir treatment of cytomegalovirus infection: case report of a
pancreatico-­renal transplant recipient. Transplantation. 2003;75:724–7.
18. Besecker MI, Furness CL, Coen DM, Griffiths A. Expression of extremely low levels of thy-
midine kinase from an acyclovir-resistant herpes simplex virus mutant supports reactivation
from latently infected mouse trigeminal ganglia. J Virol. 2007;81:8356–60.
19. Bestman-Smith J, Boivin G. Herpes simplex virus isolates with reduced adefovir susceptibil-
ity selected in vivo by foscarnet therapy. J Med Virol. 2002;67:88–91.
20. Bestman-Smith J, Boivin G. Drug resistance patterns of recombinant herpes simplex virus
DNA polymerase mutants generated with a set of overlapping cosmids and plasmids. J Virol.
2003;77:7820–9.
21. Bhorade SM, Lurain NS, Jordan A, Leischner J, Villanueva J, Durazo R, et al. Emergence of
ganciclovir-resistant cytomegalovirus in lung transplant recipients. J Heart Lung Transplant.
2002;21:1274–82.
22. Biron KK, Stanat SC, Sorrell JB, Fyfe JA, Keller PM, Lambe CU, et al. Metabolic activa-
tion of the nucleoside analog 9-[(2-hydroxy-1-(hydroxymethyl)ethoxy]methyl)guanine in
human diploid fibroblasts infected with human cytomegalovirus. Proc Natl Acad Sci U S A.
1985;82:2473–7.
23. Biron KK, Harvey RJ, Chamberlain SC, Good SS, Smith AA, Davis MG, et al. Potent and
selective inhibition of human cytomegalovirus replication by 1263W94, a benzimidazole
L-riboside with a unique mode of action. Antimicrob Agents Chemother. 2002;46:2365–72.
24. Boeckh M, Boivin G. Quantitation of cytomegalovirus: methodologic aspects and clinical
applications. Clin Microbiol Rev. 1998;11:533–54.
25. Boeckh M, Leisenring W, Riddell SR, Bowden RA, Huang ML, Myerson D, et al. Late
cytomegalovirus disease and mortality in recipients of allogeneic hematopoietic stem cell
transplants: importance of viral load and T-cell immunity. Blood. 2003;101:407–14.
26. Boivin G, Erice A, Crane DD, Dunn DL, Balfour HH Jr. Acyclovir susceptibilities of herpes
simplex virus strains isolated from solid organ transplant recipients after acyclovir or ganci-
clovir prophylaxis. Antimicrob Agents Chemother. 1993;37:357–9.
27. Boivin G, Edelman CK, Pedneault L, Talarico CL, Biron KK, Balfour HH Jr. Phenotypic
and genotypic characterization of acyclovir-resistant varicella-zoster viruses isolated from
persons with AIDS. J Infect Dis. 1994;170:68–75.
28. Boivin G, Chou S, Quirk MR, Erice A, Jordan MC. Detection of ganciclovir resistance muta-
tions quantitation of cytomegalovirus (CMV) DNA in leukocytes of patients with fatal dis-
seminated CMV disease. J Infect Dis. 1996;173:523–8.
29. Boivin G, Gilbert C, Gaudreau A, Greenfield I, Sudlow R, Roberts NA. Rate of emergence
of cytomegalovirus (CMV) mutations in leukocytes of patients with acquired immunodefi-
ciency syndrome who are receiving valganciclovir as induction and maintenance therapy for
CMV retinitis. J Infect Dis. 2001;184:1598–602.
30. Boivin G, Goyette N, Gilbert C, Roberts N, Macey K, Paya C, et al. Absence of
cytomegalovirus-­resistance mutations after valganciclovir prophylaxis, in a prospective mul-
ticenter study of solid-organ transplant recipients. J Infect Dis. 2004;189:1615–8.
31. Boivin G, Goyette N, Gilbert C, Covington E. Analysis of cytomegalovirus DNA polymerase
(UL54) mutations in solid organ transplant patients receiving valganciclovir or ganciclovir
prophylaxis. J Med Virol. 2005;77:425–9.
32. Boivin G, Goyette N, Gilbert C, Humar A, Covington E. Clinical impact of ganciclovir-­
resistant cytomegalovirus infections in solid organ transplant patients. Transpl Infect Dis.
2005;7:166–70.
256 W. L. Drew et al.

33. Boon RJ, Bacon TH, Robey HL, Coleman TJ, Connolly A, Crosson P, et al. Antiviral sus-
ceptibilities of herpes simplex virus from immunocompetent subjects with recurrent herpes
labialis: a UK-based survey. J Antimicrob Chemother. 2000;46:324–5.
34. Breton G, Fillet AM, Katlama C, Bricaire F, Caumes E. Acyclovir-resistant herpes zoster in
human immunodeficiency virus-infected patients: results of foscarnet therapy. Clin Infect
Dis. 1998;27:1525–7.
35. Brink AA, van Gelder M, Wolffs PF, Bruggeman CA, van Loo IH. Compartmentalization of
acyclovir-resistant varicella zoster virus: implications for sampling in molecular diagnostics.
Clin Infect Dis. 2011;52:982–7.
36. Brunnemann AK, Bohn-Wippert K, Zell R, Henke A, Walther M, Braum O, et al. Drug resis-
tance of clinical varicella-zoster virus strains confirmed by recombinant thymidine kinase
expression and by targeted resistance mutagenesis of a cloned wild-type isolate. Antimicrob
Agents Chemother. 2015;59:2726–34.
37. Brunnemann AK, Liermann K, Deinhardt-Emmer S, Maschkowitz G, Pohlmann A, Sodeik
B, et al. Recombinant herpes simplex virus type 1 strains with targeted mutations relevant for
aciclovir susceptibility. Sci Rep. 2016;6:29903.
38. Bryan CJ, Prichard MN, Daily S, Jefferson G, Hartline C, Cassady KA, et al. Acyclovir-­
resistant chronic verrucous vaccine strain varicella in a patient with neuroblastoma. Pediatr
Infect Dis J. 2008;27:946–8.
39. Burrel S, Bonnafous P, Hubacek P, Agut H, Boutolleau D. Impact of novel mutations of her-
pes simplex virus 1 and 2 thymidine kinases on acyclovir phosphorylation activity. Antivir
Res. 2012;96:386–90.
40. Burrel S, Aime C, Hermet L, Ait-Arkoub Z, Agut H, Boutolleau D. Surveillance of herpes
simplex virus resistance to antivirals: a 4-year survey. Antivir Res. 2013;100:365–72.
41. Burrel S, Boutolleau D, Azar G, Doan S, Deback C, Cochereau I, et al. Phenotypic and geno-
typic characterization of acyclovir-resistant corneal HSV-1 isolates from immunocompetent
patients with recurrent herpetic keratitis. J Clin Virol. 2013;58:321–4.
42. Chakrabarti S, Pillay D, Ratcliffe D, Cane PA, Collingham KE, Milligan DW. Resistance
to antiviral drugs in herpes simplex virus infections among allogeneic stem cell transplant
recipients: risk factors and prognostic significance. J Infect Dis. 2000;181:2055–8.
43. Chatis PA, Miller CH, Schrager LE, Crumpacker CS. Successful treatment with foscarnet of
an acyclovir-resistant mucocutaneous infection with herpes simplex virus in a patient with
acquired immunodeficiency syndrome. N Engl J Med. 1989;320:297–300.
44. Chatis PA, Crumpacker CS. Analysis of the thymidine kinase gene from clinically isolated
acyclovir-resistant herpes simplex viruses. Virology. 1991;180:793–7.
45. Chemaly RF, Ullmann AJ, Stoelben S, Richard MP, Bornhauser M, Groth C, et al. Letermovir
for cytomegalovirus prophylaxis in hematopoietic-cell transplantation. N Engl J Med.
2014;370:1781–9.
46. Chen H, Beardsley GP, Coen DM. Mechanism of ganciclovir-induced chain termination
revealed by resistant viral polymerase mutants with reduced exonuclease activity. Proc Natl
Acad Sci U S A. 2014;111:17462–7.
47. Chen Y, Scieux C, Garrait V, Socie G, Rocha V, Molina JM, et al. Resistant herpes simplex
virus type 1 infection: an emerging concern after allogeneic stem cell transplantation. Clin
Infect Dis. 2000;31:927–35.
48. Chilukuri S, Rosen T. Management of acyclovir-resistant herpes simplex virus. Dermatol
Clin. 2003;21:311–20.
49. Chou S, Erice A, Jordan MC, Vercellotti GM, Michels KR, Talarico CL, et al. Analysis of the
UL97 phosphotransferase coding sequence in clinical cytomegalovirus isolates and identifi-
cation of mutations conferring ganciclovir resistance. J Infect Dis. 1995;171:576–83.
50. Chou S, Marousek G, Guentzel S, Follansbee SE, Poscher ME, Lalezari JP, et al. Evolution
of mutations conferring multidrug resistance during prophylaxis and therapy for cytomegalo-
virus disease. J Infect Dis. 1997;176:786–9.
8 Resistance of Herpesviruses to Antiviral Agents 257

51. Chou S, Marousek G, Parenti DM, Gordon SM, LaVoy AG, Ross JG, et al. Mutation in region
III of the DNA polymerase gene conferring foscarnet resistance in cytomegalovirus isolates
from 3 subjects receiving prolonged antiviral therapy. J Infect Dis. 1998;178:526–30.
52. Chou S, Lurain NS, Weinberg A, Cai GY, Sharma PL, Crumpacker CS. Interstrain varia-
tion in the human cytomegalovirus DNA polymerase sequence and its effect on genotypic
diagnosis of antiviral drug resistance. Adult AIDS Clinical Trials Group CMV Laboratories.
Antimicrob Agents Chemother. 1999;43:1500–2.
53. Chou S, Meichsner CL. A nine-codon deletion mutation in the cytomegalovirus UL97
phosphotransferase gene confers resistance to ganciclovir. Antimicrob Agents Chemother.
2000;44:183–5.
54. Chou S, Waldemer RH, Senters AE, Michels KS, Kemble GW, Miner RC, et al.
Cytomegalovirus UL97 phosphotransferase mutations that affect susceptibility to ganciclo-
vir. J Infect Dis. 2002;185:162–9.
55. Chou S, Lurain NS, Thompson KD, Miner RC, Drew WL. Viral DNA polymerase mutations
associated with drug resistance in human cytomegalovirus. J Infect Dis. 2003;188:32–9.
56. Chou S, Van Wechel LC, Lichy HM, Marousek GI. Phenotyping of cytomegalovirus drug
resistance mutations by using recombinant viruses incorporating a reporter gene. Antimicrob
Agents Chemother. 2005;49:2710–5.
57. Chou S, Hakki M, Villano S. Effects on maribavir susceptibility of cytomegalovirus UL97
kinase ATP binding region mutations detected after drug exposure in vitro and in vivo. Antivir
Res. 2012;95:88–92.
58. Chou S, Ercolani RJ, Marousek G, Bowlin TL. Cytomegalovirus UL97 kinase cata-
lytic domain mutations that confer multidrug resistance. Antimicrob Agents Chemother.
2013;57:3375–9.
59. Chrisp P, Clissold SP. Foscarnet. A review of its antiviral activity, pharmacokinetic properties
and therapeutic use in immunocompromised patients with cytomegalovirus retinitis. Drugs.
1991;41:104–29.
60. Christophers J, Clayton J, Craske J, Ward R, Collins P, Trowbridge M, et al. Survey of
resistance of herpes simplex virus to acyclovir in Northwest England. Antimicrob Agents
Chemother. 1998;42:868–72.
61. Cihlar T, Chen MS. Identification of enzymes catalyzing two-step phosphorylation of cidofo-
vir and the effect of cytomegalovirus infection on their activities in host cells. Mol Pharmacol.
1996;50:1502–10.
62. Cihlar T, Fuller MD, Cherrington JM. Characterization of drug resistance-associated muta-
tions in the human cytomegalovirus DNA polymerase gene by using recombinant mutant
viruses generated from overlapping DNA fragments. J Virol. 1998;72:5927–36.
63. Cihlar T, Fuller MD, Mulato AS, Cherrington JM. A point mutation in the human cytomega-
lovirus DNA polymerase gene selected in vitro by cidofovir confers a slow replication phe-
notype in cell culture. Virology. 1998;248:382–93.
64. Collins P, Larder BA, Oliver NM, Kemp S, Smith IW, Darby G. Characterization of a DNA
polymerase mutant of herpes simplex virus from a severely immunocompromised patient
receiving acyclovir. J Gen Virol. 1989;70:375–82.
65. Collins P, Ellis MN. Sensitivity monitoring of clinical isolates of herpes simplex virus to
acyclovir. J Med Virol. 1993;(Suppl 1):58–66.
66. Crassard N, Souillet AL, Morfin F, Thouvenot D, Claudy A, Bertrand Y. Acyclovir-resistant
varicella infection with atypical lesions in a non-HIV leukemic infant. Acta Paediatr.
2000;89:1497–9.
67. Dambrosi S, Martin M, Yim K, Miles B, Canas J, Sergerie Y, et al. Neurovirulence and
latency of drug-resistant clinical herpes simplex viruses in animal models. J Med Virol.
2010;82:1000.
68. Dankner WM, Scholl D, Stanat SC, Martin M, Sonke RL, Spector SA. Rapid antiviral DNA-­
DNA hybridization assay for human cytomegalovirus. J Virol Methods. 1990;28:293–8.
258 W. L. Drew et al.

69. Danve-Szatanek C, Aymard M, Thouvenot D, Morfin F, Agius G, Bertin I, et al. Surveillance

network for herpes simplex virus resistance to antiviral drugs: 3-year follow-up. J Clin
Microbiol. 2004;42:242–9.
70. Darby G, Larder BA, Inglis MM. Evidence that the ‘active center’ of the herpes simplex virus
thymidine kinase involves an interaction between three distinct regions of the polypeptide.
J Gen Virol. 1986;67:753–8.
71. De Clercq E. Discovery and development of BVDU (brivudin) as a therapeutic for the treat-
ment of herpes zoster. Biochem Pharmacol. 2004;68:2301–15.
72. DeJesus E, Wald A, Warren T, Schacker TW, Trottier S, Shahmanesh M, et al. Valacyclovir
for the suppression of recurrent genital herpes in human immunodeficiency virus-infected
subjects. J Infect Dis. 2003;188:1009–16.
73. Drew WL, Miner RC, Busch DF, Follansbee SE, Gullett J, Mehalko SG, et al. Prevalence
of resistance in patients receiving ganciclovir for serious cytomegalovirus infection. J Infect
Dis. 1991;163:716–9.
74. Drew WL, Anderson R, Lang W, Miner RC, Davis G, Lalezari J. Failure of high-dose oral
acyclovir to suppress CMV viruria or induce ganciclovir-resistant CMV in HIV antibody
positive patients. J Acquir Immune Defic Syndr Hum Retrovirol. 1995;8:289–91.
75. Drew WL, Stempien MJ, Andrews J, Shadman A, Tan SJ, Miner R, et al. Cytomegalovirus
(CMV) resistance in patients with CMV retinitis and AIDS treated with oral or intravenous
ganciclovir. J Infect Dis. 1999;179:1352–5.
76. Drew WL, Paya CV, Emery V. Cytomegalovirus (CMV) resistance to antivirals. Am
J Transplant. 2001;1:307–12.
77. Drew WL. Cytomegalovirus disease in highly active antiretroviral therapy era. Curr Infect
Dis Rep. 2003;5:257–65.
78. Drew WL, Liu C. Repopulation of ganciclovir-resistant cytomegalovirus by wild-type virus.
Clin Transpl. 2012;26:949–52.
79. Drouot E, Piret J, Boivin G. Novel method based on "en passant" mutagenesis coupled with
a Gaussia luciferase reporter assay for studying the combined effects of human cytomegalo-
virus mutations. J Clin Microbiol. 2013;51:3216–24.
80. Duan R, de Vries RD, Osterhaus AD, Remeijer L, Verjans GM. Acyclovir-resistant corneal
HSV-1 isolates from patients with herpetic keratitis. J Infect Dis. 2008;198:659–63.
81. Eckle T, Prix L, Jahn G, Klingebiel T, Handgretinger R, Selle B, et al. Drug-resistant human
cytomegalovirus infection in children after allogeneic stem cell transplantation may have dif-
ferent clinical outcomes. Blood. 2000;96:3286–9.
82. Eid AJ, Arthurs SK, Deziel PJ, Wilhelm MP, Razonable RR. Emergence of drug-resistant
cytomegalovirus in the era of valganciclovir prophylaxis: therapeutic implications and out-
comes. Clin Transpl. 2008;22:162–70.
83. Engel JP, Englund JA, Fletcher CV, Hill EL. Treatment of resistant herpes simplex virus with
continuous-infusion acyclovir. JAMA. 1990;263:1662–4.
84. Englund JA, Zimmerman ME, Swierkosz EM, Goodman JL, Scholl DR, Balfour HH Jr.
Herpes simplex virus resistant to acyclovir. A study in a tertiary care center. Ann Intern Med.
1990;112:416–22.
85. Erard V, Wald A, Corey L, Leisenring WM, Boeckh M. Use of long-term suppressive acy-
clovir after hematopoietic stem-cell transplantation: impact on herpes simplex virus (HSV)
disease and drug-resistant HSV disease. J Infect Dis. 2007;196:266–70.
86. Erice A, Gil-Roda C, Perez JL, Balfour HH Jr, Sannerud KJ, Hanson MN, et al. Antiviral
susceptibilities and analysis of UL97 and DNA polymerase sequences of clinical cytomega-
lovirus isolates from immunocompromised patients. J Infect Dis. 1997;175:1087–92.
87. Erice A, Borrell N, Li W, Miller WJ, Balfour HH Jr. Ganciclovir susceptibilities and analysis
of UL97 region in cytomegalovirus (CMV) isolates from bone marrow recipients with CMV
disease after antiviral prophylaxis. J Infect Dis. 1998;178:531–4.
88. Erlich KS, Jacobson MA, Koehler JE, Follansbee SE, Drennan DP, Gooze L, et al. Foscarnet
therapy for severe acyclovir-resistant herpes simplex virus type-2 infections in patients with
8 Resistance of Herpesviruses to Antiviral Agents 259

the acquired immunodeficiency syndrome (AIDS). An uncontrolled trial. Ann Intern Med.
1989;110:710–3.
89. Erlich KS, Mills J, Chatis P, Mertz GJ, Busch DF, Follansbee SE, et al. Acyclovir-resistant
herpes simplex virus infections in patients with the acquired immunodeficiency syndrome. N
Engl J Med. 1989;320:293–6.
90. Field HJ, Vere Hodge RA. Recent developments in anti-herpesvirus drugs. Br Med Bull.
2013;106:213–49.
91. Fife KH, Crumpacker CS, Mertz GJ, Hill EL, Boone GS. Recurrence and resistance patterns
of herpes simplex virus following cessation of > or = 6 years of chronic suppression with
acyclovir. Acyclovir study group. J Infect Dis. 1994;169:1338–41.
92. Fillet AM, Visse B, Caumes E, Dumont B, Gentilini M, Huraux JM. Foscarnet-resistant mul-
tidermatomal zoster in a patient with AIDS. Clin Infect Dis. 1995;21:1348–9.
93. Fillet AM, Dumont B, Caumes E, Visse B, Agut H, Bricaire F, et al. Acyclovir-resistant
varicella-­zoster virus: phenotypic and genetic characterization. J Med Virol. 1998;55:250–4.
94. Fillet AM, Auray L, Alain S, Gourlain K, Imbert BM, Najioullah F, et al. Natural polymor-
phism of cytomegalovirus DNA polymerase lies in two nonconserved regions located between
domains delta-C and II and between domains III and I. Antimicrob Agents Chemother.
2004;48:1865–8.
95. Florescu DF, Qiu F, Schmidt CM, Kalil AC. A direct and indirect comparison meta-analysis
on the efficacy of cytomegalovirus preventive strategies in solid organ transplant. Clin Infect
Dis. 2014;58:785–803.
96. Frangoul H, Wills M, Crossno C, Engel M, Domm J. Acyclovir-resistant herpes simplex virus
pneumonia post-unrelated stem cell transplantation: a word of caution. Pediatr Transplant.
2007;11:942–4.
97. Frobert E, Burrel S, Ducastelle-Lepretre S, Billaud G, Ader F, Casalegno JS, et al. Resistance
of herpes simplex viruses to acyclovir: an update from a ten-year survey in France. Antivir
Res. 2014;111:36–41.
98. Fyfe JA, Keller PM, Furman PA, Miller RL, Elion GB. Thymidine kinase from herpes sim-
plex virus phosphorylates the new antiviral compound, 9-(2-hydroxyethoxymethyl)guanine.
J Biol Chem. 1978;253:8721–7.
99. Gaudreau A, Hill E, Balfour HH Jr, Erice A, Boivin G. Phenotypic and genotypic charac-
terization of acyclovir-resistant herpes simplex viruses from immunocompromised patients.
J Infect Dis. 1998;178:297–303.
100. Gerna G, Baldanti F, Zavattoni M, Sarasini A, Percivalle E, Revello MG. Monitoring of gan-
ciclovir sensitivity of multiple human cytomegalovirus strains coinfecting blood of an AIDS
patient by an immediate-early antigen plaque assay. Antivir Res. 1992;19:333–45.
101. Gerna G, Sarasini A, Lilleri D, Percivalle E, Torsellini M, Baldanti F, et al. In vitro model
for the study of the dissociation of increasing antigenemia and decreasing DNAemia and
viremia during treatment of human cytomegalovirus infection with ganciclovir in transplant
recipients. J Infect Dis. 2003;188:1639–47.
102. Gilbert C, Roy J, Belanger R, Delage R, Beliveau C, Demers C, et al. Lack of emergence
of cytomegalovirus UL97 mutations conferring ganciclovir (GCV) resistance following
preemptive GCV therapy in allogeneic stem cell transplant recipients. Antimicrob Agents
Chemother. 2001;45:3669–71.
103. Gilbert C, Bestman-Smith J, Boivin G. Resistance of herpesviruses to antiviral drugs: clinical
impacts and molecular mechanisms. Drug Resist Updat. 2002;5:88–114.
104. Gilbert C, Boivin G. Discordant phenotypes and genotypes of cytomegalovirus (CMV) in
patients with AIDS and relapsing CMV retinitis. AIDS. 2003;17:337–41.
105. Gilbert C, Boivin G. New reporter cell line to evaluate the sequential emergence of mul-
tiple human cytomegalovirus mutations during in vitro drug exposure. Antimicrob Agents
Chemother. 2005;49:4860–6.
106. Gilbert C, Boivin G. Human cytomegalovirus resistance to antiviral drugs. Antimicrob
Agents Chemother. 2005;49:873–83.
260 W. L. Drew et al.

107. Goldner T, Hewlett G, Ettischer N, Ruebsamen-Schaeff H, Zimmermann H, Lischka P. The

novel anticytomegalovirus compound AIC246 (Letermovir) inhibits human cytomegalovirus
replication through a specific antiviral mechanism that involves the viral terminase. J Virol.
2011;85:10884–93.
108. Goldner T, Hempel C, Ruebsamen-Schaeff H, Zimmermann H, Lischka P. Geno- and phe-
notypic characterization of human cytomegalovirus mutants selected in vitro after letermovir
(AIC246) exposure. Antimicrob Agents Chemother. 2014;58:610–3.
109. Gueudry J, Boutolleau D, Gueudin M, Burrel S, Miri A, Bodaghi B, et al. Acyclovir-resistant
varicella-zoster virus keratitis in an immunocompetent patient. J Clin Virol. 2013;58:318–20.
110. Gupta R, Hill EL, McClernon D, Davis G, Selke S, Corey L, et al. Acyclovir sensitivity of
sequential herpes simplex virus type 2 isolates from the genital mucosa of immunocompetent
women. J Infect Dis. 2005;192:1102–7.
111. Hamprecht K, Eckle T, Prix L, Faul C, Einsele H, Jahn G. Ganciclovir-resistant cytomega-
lovirus disease after allogeneic stem cell transplantation: pitfalls of phenotypic diagnosis by
in vitro selection of an UL97 mutant strain. J Infect Dis. 2003;187:139–43.
112. Hanson MN, Preheim LC, Chou S, Talarico CL, Biron KK, Erice A. Novel mutation in
the UL97 gene of a clinical cytomegalovirus strain conferring resistance to ganciclovir.
Antimicrob Agents Chemother. 1995;39:1204.
113. Hew K, Dahlroth SL, Veerappan S, Pan LX, Cornvik T, Nordlund P. Structure of the vari-
cella zoster virus thymidylate synthase establishes functional and structural similarities as the
human enzyme and potentiates itself as a target of brivudine. PLoS One. 2015;10:e0143947.
114. Hill EL, Hunter GA, Ellis MN. In vitro and in vivo characterization of herpes simplex virus
clinical isolates recovered from patients infected with human immunodeficiency virus.
Antimicrob Agents Chemother. 1991;35:2322–8.
115. Hostetler KY. Synthesis and early development of hexadecyloxypropylcidofovir: an oral anti-
poxvirus nucleoside phosphonate. Virus. 2010;2:2213–25.
116. Hu H, Jabs DA, Forman MS, Martin BK, Dunn JP, Weinberg DV, et al. Comparison of cyto-
megalovirus (CMV) UL97 gene sequences in the blood and vitreous of patients with acquired
immunodeficiency syndrome and CMV retinitis. J Infect Dis. 2002;185:861–7.
117. Huang L, Ishii KK, Zuccola H, Gehring AM, Hwang CB, Hogle J, et al. The enzymological
basis for resistance of herpesvirus DNA polymerase mutants to acyclovir: relationship to the
structure of alpha-like DNA polymerases. Proc Natl Acad Sci U S A. 1999;96:447–52.
118. Humar A, Kumar D, Preiksaitis J, Boivin G, Siegal D, Fenton J, et al. A trial of valganciclovir
prophylaxis for cytomegalovirus prevention in lung transplant recipients. Am J Transplant.
2005;5:1462–8.
119. Hwang CB, Ruffner KL, Coen DM. A point mutation within a distinct conserved region of the
herpes simplex virus DNA polymerase gene confers drug resistance. J Virol. 1992;66:1774–6.
120. Jabs DA, Enger C, Dunn JP, Forman M. Cytomegalovirus retinitis and viral resistance: gan-
ciclovir resistance. J Infect Dis. 1998;177:770.
121. Jabs DA, Enger C, Forman M, Dunn JP. Incidence of foscarnet resistance and cidofovir resis-
tance in patients treated for cytomegalovirus retinitis. The Cytomegalovirus Retinitis and
Viral Resistance Study Group. Antimicrob Agents Chemother. 1998;42:2240–4.
122. Jabs DA, Martin BK, Forman MS, Dunn JP, Davis JL, Weinberg DV, et al. Mutations confer-
ring ganciclovir resistance in a cohort of patients with acquired immunodeficiency syndrome
and cytomegalovirus retinitis. J Infect Dis. 2001;183:333–7.
123. Jabs DA, Martin BK, Forman MS, Dunn JP, Davis JL, Weinberg DV, et al. Longitudinal
observations on mutations conferring ganciclovir resistance in patients with acquired immu-
nodeficiency syndrome and cytomegalovirus retinitis: The Cytomegalovirus and Viral
Resistance Study Group Report Number 8. Am J Ophthalmol. 2001;132:700–10.
124. Jacobson MA, Berger TG, Fikrig S, Becherer P, Moohr JW, Stanat SC, et al. Acyclovir-­
resistant varicella zoster virus infection after chronic oral acyclovir therapy in patients with
the acquired immunodeficiency syndrome (AIDS). Ann Intern Med. 1990;112:187–91.
8 Resistance of Herpesviruses to Antiviral Agents 261

125. James SH, Prichard MN. A possible pitfall in acyclovir prophylaxis for recurrent herpetic
keratitis? J Infect Dis. 2013;208:1353–5.
126. Javaly K, Wohlfeiler M, Kalayjian R, Klein T, Bryson Y, Grafford K, et al. Treatment of
mucocutaneous herpes simplex virus infections unresponsive to acyclovir with topical foscar-
net cream in AIDS patients: a phase I/II study. J Acquir Immune Defic Syndr. 1999;21:301–6.
127. Kamiyama T, Kurokawa M, Shiraki K. Characterization of the DNA polymerase gene of
varicella-zoster viruses resistant to acyclovir. J Gen Virol. 2001;82:2761–5.
128. Karamitros T, Harrison I, Piorkowska R, Katzourakis A, Magiorkinis G, Mbisa JL. De novo
assembly of human herpes virus type 1 (HHV-1) genome, mining of non-canonical structures
and detection of novel drug-resistance mutations using short- and long-read next generation
sequencing technologies. PLoS One. 2016;11:e0157600.
129. Kaul DR, Stoelben S, Cober E, Ojo T, Sandusky E, Lischka P, et al. First report of success-
ful treatment of multidrug-resistant cytomegalovirus disease with the novel anti-CMV com-
pound AIC246. Am J Transplant. 2011;11:1079–84.
130. Kessler HA, Hurwitz S, Farthing C, Benson CA, Feinberg J, Kuritzkes DR, et al. Pilot study
of topical trifluridine for the treatment of acyclovir-resistant mucocutaneous herpes simplex
disease in patients with AIDS (ACTG 172). AIDS Clinical Trials Group. J Acquir Immune
Defic Syndr Hum Retrovirol. 1996;12:147–52.
131. Kesson AM, Zeng F, Cunningham AL, Rawlinson WD. The use of flow cytometry to detect
antiviral resistance in human cytomegalovirus. J Virol Methods. 1998;71:177–86.
132. Kim JH, Schaenman JM, Ho DY, Brown JM. Treatment of acyclovir-resistant herpes sim-
plex virus with continuous infusion of high-dose acyclovir in hematopoietic cell transplant
patients. Biol Blood Marrow Transplant. 2011;17:259–64.
133. Kost RG, Hill EL, Tigges M, Straus SE. Brief report: recurrent acyclovir-resistant genital
herpes in an immunocompetent patient. N Engl J Med. 1993;329:1777–82.
134. Kotton CN, Kumar D, Caliendo AM, Asberg A, Chou S, Danziger-Isakov L, et al. Updated
international consensus guidelines on the management of cytomegalovirus in solid-organ
transplantation. Transplantation. 2013;96:333–60.
135. Krawczyk A, Arndt MA, Grosse-Hovest L, Weichert W, Giebel B, Dittmer U, et al.
Overcoming drug-resistant herpes simplex virus (HSV) infection by a humanized antibody.
Proc Natl Acad Sci U S A. 2013;110:6760–5.
136. Kriesel JD, Spruance SL, Prichard M, Parker JN, Kern ER. Recurrent antiviral-resistant geni-
tal herpes in an immunocompetent patient. J Infect Dis. 2005;192:156–61.
137. Kruger RM, Shannon WD, Arens MQ, Lynch JP, Storch GA, Trulock EP. The impact of
ganciclovir-resistant cytomegalovirus infection after lung transplantation. Transplantation.
1999;68:1272–9.
138. Lalezari J, Schacker T, Feinberg J, Gathe J, Lee S, Cheung T, et al. A randomized, double-­
blind, placebo-controlled trial of cidofovir gel for the treatment of acyclovir-­unresponsive
mucocutaneous herpes simplex virus infection in patients with AIDS. J Infect Dis.
1997;176:892–8.
139. Landry ML, Stanat S, Biron K, Brambilla D, Britt W, Jokela J, et al. A standardized plaque
reduction assay for determination of drug susceptibilities of cytomegalovirus clinical iso-
lates. Antimicrob Agents Chemother. 2000;44:688–92.
140. Langston AA, Redei I, Caliendo AM, Somani J, Hutcherson D, Lonial S, et al. Development
of drug-resistant herpes simplex virus infection after haploidentical hematopoietic progenitor
cell transplantation. Blood. 2002;99:1085–8.
141. Lascaux AS, Caumes E, Deback C, Melica G, Challine D, Agut H, et al. Successful treatment
of aciclovir and foscarnet resistant Herpes simplex virus lesions with topical imiquimod in
patients infected with human immunodeficiency virus type 1. J Med Virol. 2012;84:194–7.
142. Levin MJ, Dahl KM, Weinberg A, Giller R, Patel A, Krause PR. Development of resistance
to acyclovir during chronic infection with the Oka vaccine strain of varicella-zoster virus, in
an immunosuppressed child. J Infect Dis. 2003;188:954–9.
262 W. L. Drew et al.

143. Levin MJ, Bacon TH, Leary JJ. Resistance of herpes simplex virus infections to nucleoside
analogues in HIV-infected patients. Clin Infect Dis. 2004;39:S248–57.
144. Limaye AP, Corey L, Koelle DM, Davis CL, Boeckh M. Emergence of ganciclovir-­
resistant cytomegalovirus disease among recipients of solid-organ transplants. Lancet.
2000;356:645–9.
145. Limaye AP. Ganciclovir-resistant cytomegalovirus in organ transplant recipients. Clin Infect
Dis. 2002;35:866–72.
146. Limaye AP, Bakthavatsalam R, Kim HW, Randolph SE, Halldorson JB, Healey PJ, et al.
Impact of cytomegalovirus in organ transplant recipients in the era of antiviral prophylaxis.
Transplantation. 2006;81:1645–52.
147. Lischka P, Hewlett G, Wunberg T, Baumeister J, Paulsen D, Goldner T, et al. In vitro and
in vivo activities of the novel anticytomegalovirus compound AIC246. Antimicrob Agents
Chemother. 2010;54:1290–7.
148. Liu W, Kuppermann BD, Martin DF, Wolitz RA, Margolis TP. Mutations in the cytomegalovi-
rus UL97 gene associated with ganciclovir-resistant retinitis. J Infect Dis. 1998;177:1176–81.
149. Lolis MS, Gonzalez L, Cohen PJ, Schwartz RA. Drug-resistant herpes simplex virus in HIV
infected patients. Acta Dermatovenerol Croat. 2008;16:204–8.
150. LoPresti AE, Levine JF, Munk GB, Tai CY, Mendel DB. Successful treatment of an acyclovir-
and foscarnet-resistant herpes simplex virus type 1 lesion with intravenous cidofovir. Clin
Infect Dis. 1998;26:512–3.
151. Lowance D, Neumayer HH, Legendre CM, Squifflet JP, Kovarik J, Brennan PJ, et al.
Valacyclovir for the prevention of cytomegalovirus disease after renal transplantation.
International Valacyclovir Cytomegalovirus Prophylaxis Transplantation Study Group. N
Engl J Med. 1999;340:1462–70.
152. Lurain NS, Weinberg A, Crumpacker CS, Chou S, Adult AIDS Clinical Trials Group-CMV
Laboratories. Sequencing of cytomegalovirus UL97 gene for genotypic antiviral resistance
testing. Antimicrob Agents Chemother. 2001;45:2775–80.
153. Lurain NS, Bhorade SM, Pursell KJ, Avery RK, Yeldandi VV, Isada CM, et al. Analysis and
characterization of antiviral drug-resistant cytomegalovirus isolates from solid organ trans-
plant recipients. J Infect Dis. 2002;186:760–8.
154. Lurain NS, Chou S. Antiviral drug resistance of human cytomegalovirus. Clin Microbiol Rev.
2010;23:689–712.
155. Malartre N, Boulieu R, Falah N, Cortay JC, Lina B, Morfin F, et al. Effects of mutations
on herpes simplex virus 1 thymidine kinase functionality: an in vitro assay based on detec-
tion of monophosphate forms of acyclovir and thymidine using HPLC/DAD. Antivir Res.
2012;95:224–8.
156. Martin JL, Ellis MN, Keller PM, Biron KK, Lehrman SN, Barry DW, et al. Plaque autora-
diography assay for the detection and quantitation of thymidine kinase-deficient and thymi-
dine kinase-altered mutants of herpes simplex virus in clinical isolates. Antimicrob Agents
Chemother. 1985;28:181–7.
157. Marty FM, Winston DJ, Rowley SD, Vance E, Papanicolaou GA, Mullane KM, et al.
CMX001 to prevent cytomegalovirus disease in hematopoietic-cell transplantation. N Engl
J Med. 2013;369:1227–36.
158. McSharry JM, Lurain NS, Drusano GL, Landay A, Manischewitz J, Nokta M, et al. Flow
cytometric determination of ganciclovir susceptibilities of human cytomegalovirus clinical
isolates. J Clin Microbiol. 1998;36:958–64.
159. Mertz GJ, Jones CC, Mills J, Fife KH, Lemon SM, Stapleton JT, et al. Long-term acyclo-
vir suppression of frequently recurring genital herpes simplex virus infection. A multicenter
double-blind trial. JAMA. 1988;260:201–6.
160. Michel D, Hohn S, Haller T, Jun D, Mertens T. Aciclovir selects for ganciclovir-cross-­
resistance of human cytomegalovirus in vitro that is only in part explained by known muta-
tions in the UL97 protein. J Med Virol. 2001;65:70–6.
8 Resistance of Herpesviruses to Antiviral Agents 263

161. Morfin F, Thouvenot D, De Turenne-Tessier M, Lina B, Aymard M, Ooka T. Phenotypic and

genetic characterization of thymidine kinase from clinical strains of varicella-zoster virus
resistant to acyclovir. Antimicrob Agents Chemother. 1999;43:2412–6.
162. Morfin F, Souillet G, Bilger K, Ooka T, Aymard M, Thouvenot D. Genetic characterization
of thymidine kinase from acyclovir-resistant and -susceptible herpes simplex virus type 1
isolated from bone marrow transplant recipients. J Infect Dis. 2000;182:290–3.
163. Morfin F, Thouvenot D. Herpes simplex virus resistance to antiviral drugs. J Clin Virol.
2003;26:29–37.
164. Mousavi-Jazi M, Schloss L, Drew WL, Linde A, Miner RC, Harmenberg J, et al. Variations in
the cytomegalovirus DNA polymerase and phosphotransferase genes in relation to foscarnet
and ganciclovir sensitivity. J Clin Virol. 2001;23:1–15.
165. Mylonakis E, Kallas WM, Fishman JA. Combination antiviral therapy for ganciclovir-­
resistant cytomegalovirus infection in solid-organ transplant recipients. Clin Infect Dis.
2002;34:1337–41.
166. Nichols WG, Boeckh M. Cytomegalovirus infections. Curr Treat Opt Infect Dis.
2001;3:78–91.
167. Nichols WG, Corey L, Gooley T, Drew WL, Miner R, Huang M, et al. Rising pp65 anti-
genemia during preemptive anticytomegalovirus therapy after allogeneic hematopoietic
stem cell transplantation: risk factors, correlation with DNA load, and outcomes. Blood.
2001;97:867–74.
168. Nichols WG, Boeckh M, Carter RA, Wald A, Corey L. Transferred herpes simplex virus
immunity after stem-cell transplantation: clinical implications. J Infect Dis. 2003;187:801–8.
169. Nugier F, Colin JN, Aymard M, Langlois M. Occurrence and characterization of acyclovir-­
resistant herpes simplex virus isolates: report on a two-year sensitivity screening survey.
J Med Virol. 1992;36:1–12.
170. Pan D, Kaye SB, Hopkins M, Kirwan R, Hart IJ, Coen DM. Common and new acyclovir
resistant herpes simplex virus-1 mutants causing bilateral recurrent herpetic keratitis in an
immunocompetent patient. J Infect Dis. 2014;209:345–9.
171. Park SY, Lee SO, Choi SH, Kim YS, Woo JH, Baek S, et al. Efficacy and safety of low-­
dose ganciclovir preemptive therapy in allogeneic haematopoietic stem cell transplant
recipients compared with conventional-dose ganciclovir: a prospective observational study.
J Antimicrob Chemother. 2012;67:1486–92.
172. Paya C, Humar A, Dominguez E, Washburn K, Blumberg E, Alexander B, et al. Efficacy
and safety of valganciclovir vs. oral ganciclovir for prevention of cytomegalovirus disease in
solid organ transplant recipients. Am J Transplant. 2004;4:611–20.
173. Paya CV, Wilson JA, Espy MJ, Sia IG, DeBernardi MJ, Smith TF, et al. Preemptive use of
oral ganciclovir to prevent cytomegalovirus infection in liver transplant patients: a random-
ized, placebo-controlled trial. J Infect Dis. 2002;185:854–60.
174. Perrier M, Desire N, Deback C, Agut H, Boutolleau D, Burrel S. Complementary assays for
monitoring susceptibility of varicella-zoster virus resistance to antivirals. J Virol Methods.
2016;233:10–4.
175. Pescovitz MD, Rabkin J, Merion RM, Paya CV, Pirsch J, Freeman RB, et al. Valganciclovir
results in improved oral absorption of ganciclovir in liver transplant recipients. Antimicrob
Agents Chemother. 2000;44:2811–5.
176. Piret J, Boivin G. Resistance of herpes simplex viruses to nucleoside analogues: mechanisms,
prevalence, and management. Antimicrob Agents Chemother. 2011;55:459–72.
177. Piret J, Boivin G. Antiviral drug resistance in herpesviruses other than cytomegalovirus. Rev
Med Virol. 2014;24:186–218.
178. Piret J, Goyette N, Eckenroth BE, Drouot E, Gotte M, Boivin G. Contrasting effects of
W781V and W780V mutations in helix N of herpes simplex virus 1 and human cytomegalo-
virus DNA polymerases on antiviral drug susceptibility. J Virol. 2015;89:4636–44.
179. Piret J, Boivin G. Antiviral resistance in herpes simplex virus and varicella-zoster virus infec-
tions: diagnosis and management. Curr Opin Infect Dis. 2016;29:654–62.
264 W. L. Drew et al.

180. Piret J, Goyette N, Boivin G. Novel method based on real-time cell analysis for drug sus-
ceptibility testing of herpes simplex virus and human cytomegalovirus. J Clin Microbiol.
2016;54:2120–7.
181. Reyes M, Shaik NS, Graber JM, Nisenbaum R, Wetherall NT, Fukuda K, et al. Acyclovir-­
resistant genital herpes among persons attending sexually transmitted disease and human
immunodeficiency virus clinics. Arch Intern Med. 2003;163:76–80.
182. Rifkin LM, Minkus CL, Pursell K, Jumroendararasame C, Goldstein DA. Utility of lefluno-
mide in the treatment of drug resistant cytomegalovirus retinitis. Ocul Immunol Inflamm.
2017;25:93–6.
183. Sacks SL, Shafran SD, Diaz-Mitoma F, Trottier S, Sibbald RG, Hughes A, et al. A multi-
center phase I/II dose escalation study of single-dose cidofovir gel for treatment of recurrent
genital herpes. Antimicrob Agents Chemother. 1998;42:2996–9.
184. Safrin S, Assaykeen T, Follansbee S, Mills J. Foscarnet therapy for acyclovir-resistant muco-
cutaneous herpes simplex virus infection in 26 AIDS patients: preliminary data. J Infect Dis.
1990;161:1078–84.
185. Safrin S, Ashley R, Houlihan C, Cusick PS, Mills J. Clinical and serologic features of herpes
simplex virus infection in patients with AIDS. AIDS. 1991;5:1107–10.
186. Safrin S, Berger TG, Gilson I, Wolfe PR, Wofsy CB, Mills J, et al. Foscarnet therapy in five
patients with AIDS and acyclovir-resistant varicella-zoster virus infection. Ann Intern Med.
1991;115:19–21.
187. Safrin S, Crumpacker C, Chatis P, Davis R, Hafner R, Rush J, et al. A controlled trial com-
paring foscarnet with vidarabine for acyclovir-resistant mucocutaneous herpes simplex in
the acquired immunodeficiency syndrome. The AIDS Clinical Trials Group. N Engl J Med.
1991;325:551–5.
188. Safrin S, Kemmerly S, Plotkin B, Smith T, Weissbach N, De Veranez D, et al. Foscarnet-­
resistant herpes simplex virus infection in patients with AIDS. J Infect Dis. 1994;169:193–6.
189. Saint-Leger E, Caumes E, Breton G, Douard D, Saiag P, Huraux JM, et al. Clinical and viro-
logic characterization of acyclovir-resistant varicella-zoster viruses isolated from 11 patients
with acquired immunodeficiency syndrome. Clin Infect Dis. 2001;33:2061–7.
190. Sarisky RT, Crosson P, Cano R, Quail MR, Nguyen TT, Wittrock RJ, et al. Comparison of
methods for identifying resistant herpes simplex virus and measuring antiviral susceptibility.
J Clin Virol. 2002;23:191–200.
191. Sasadeusz JJ, Sacks SL. Spontaneous reactivation of thymidine kinase-deficient, acyclovir-­
resistant type-2 herpes simplex virus: masked heterogeneity or reversion? J Infect Dis.
1996;174:476–82.
192. Sauerbrei A, Taut J, Zell R, Wutzler P. Resistance testing of clinical varicella-zoster virus
strains. Antivir Res. 2011;90:242–7.
193. Sauerbrei A, Liermann K, Bohn K, Henke A, Zell R, Gronowitz S, et al. Significance of
amino acid substitutions in the thymidine kinase gene of herpes simplex virus type 1 for
resistance. Antivir Res. 2012;96:105–7.
194. Sauerbrei A, Vodisch S, Bohn K, Schacke M, Gronowitz S. Screening of herpes simplex
virus type 1 isolates for acyclovir resistance using DiviTumR assay. J Virol Methods.
2013;188:70–2.
195. Sauerbrei A. Diagnosis, antiviral therapy, and prophylaxis of varicella-zoster virus infections.
Eur J Clin Microbiol Infect Dis. 2016;35:723–34.
196. Sauerbrei A, Bohn-Wippert K, Kaspar M, Krumbholz A, Karrasch M, Zell R. Database
on natural polymorphisms and resistance-related non-synonymous mutations in thymidine
kinase and DNA polymerase genes of herpes simplex virus types 1 and 2. J Antimicrob
Chemother. 2016;71:6–16.
197. Sawyer MH, Inchauspe G, Biron KK, Waters DJ, Straus SE, Ostrove JM. Molecular analy-
sis of the pyrimidine deoxyribonucleoside kinase gene of wild-type and acyclovir-resistant
strains of varicella-zoster virus. J Gen Virol. 1988;69:2585–93.
8 Resistance of Herpesviruses to Antiviral Agents 265

198. Schliefer K, Gumbel HO, Rockstroh JK, Spengler U. Management of progressive outer reti-
nal necrosis with cidofovir in a human immunodeficiency virus-infected patient. Clin Infect
Dis. 1999;29:684–5.
199. Schmit I, Boivin G. Characterization of the DNA polymerase and thymidine kinase genes of
herpes simplex virus isolates from AIDS patients in whom acyclovir and foscarnet therapy
sequentially failed. J Infect Dis. 1999;180:487–90.
200. Schnepf N, Boiteau N, Petit F, Alain S, Sanson-Le Pors MJ, Mazeron MC. Rapid determina-
tion of antiviral drug susceptibility of human cytomegalovirus by real-time PCR. Antivir Res.
2009;81:64–7.
201. Schubert A, Ehlert K, Schuler-Luettmann S, Gentner E, Mertens T, Michel D. Fast selection
of maribavir resistant cytomegalovirus in a bone marrow transplant recipient. BMC Infect
Dis. 2013;13:330.
202. Sergerie Y, Boivin G. Thymidine kinase mutations conferring acyclovir resistance in herpes
simplex type 1 recombinant viruses. Antimicrob Agents Chemother. 2006;50:3889–92.
203. Singh N. Antiviral drugs for cytomegalovirus in transplant recipients: advantages of preemp-
tive therapy. Rev Med Virol. 2006;16:281–7.
204. Smith IL, Shinkai M, Freeman WR, Spector SA. Polyradiculopathy associated with
ganciclovir-­resistant cytomegalovirus in an AIDS patient: phenotypic and genotypic charac-
terization of sequential virus isolates. J Infect Dis. 1996;173:1481–4.
205. Smith IL, Cherrington JM, Jiles RE, Fuller MD, Freeman WR, Spector SA. High-level resis-
tance of cytomegalovirus to ganciclovir is associated with alterations in both the UL97 and
DNA polymerase genes. J Infect Dis. 1997;176:69–77.
206. Smith IL, Taskintuna I, Rahhal FM, Powell HC, Ai E, Mueller AJ, et al. Clinical failure
of CMV retinitis with intravitreal cidofovir is associated with antiviral resistance. Arch
Ophthalmol. 1998;116:178–85.
207. Snoeck R, Andrei G, Gerard M, Silverman A, Hedderman A, Balzarini J, et al. Successful treat-
ment of progressive mucocutaneous infection due to acyclovir- and foscarnet-­resistant herpes
simplex virus with (S)-1-(3-hydroxy-2-phosphonylmethoxypropyl)cytosine (HPMPC). Clin
Infect Dis. 1994;18:570–8.
208. SOCA. Combination foscarnet and ganciclovir therapy vs monotherapy for the treatment of
relapsed cytomegalovirus retinitis in patients with AIDS. The Cytomegalovirus Retreatment
Trial. The studies of ocular complications of AIDS research group in collaboration with the
AIDS clinical trials group. Arch Ophthalmol. 1996;114:23–33.
209. Springer KL, Chou S, Li S, Giller RH, Quinones R, Shira JE, et al. How evolution of mutations
conferring drug resistance affects viral dynamics and clinical outcomes of cytomegalovirus-­
infected hematopoietic cell transplant recipients. J Clin Microbiol. 2005;43:208–13.
210. Stoelben S, Arns W, Renders L, Hummel J, Muhlfeld A, Stangl M, et al. Preemptive treat-
ment of cytomegalovirus infection in kidney transplant recipients with letermovir: results of
a phase 2a study. Transpl Int. 2014;27:77–86.
211. Stranska R, Schuurman R, Nienhuis E, Goedegebuure IW, Polman M, Weel JF, et al. Survey
of acyclovir-resistant herpes simplex virus in the Netherlands: prevalence and characteriza-
tion. J Clin Virol. 2005;32:7–18.
212. Strasfeld L, Lee I, Tatarowicz W, Villano S, Chou S. Virologic characterization of multidrug-­
resistant cytomegalovirus infection in 2 transplant recipients treated with maribavir. J Infect
Dis. 2010;202:104–8.
213. Sullivan V, Talarico CL, Stanat SC, Davis M, Coen DM, Biron KK. A protein kinase homo-
logue controls phosphorylation of ganciclovir in human cytomegalovirus-infected cells.
Nature. 1992;358:162–4.
214. Swetter SM, Hill EL, Kern ER, Koelle DM, Posavad CM, Lawrence W, et al. Chronic vul-
var ulceration in an immunocompetent woman due to acyclovir-resistant, thymidine kinase-­
deficient herpes simplex virus. J Infect Dis. 1998;177:543–50.
266 W. L. Drew et al.

215. Swierkosz EM, Hodinka RL, Moore BM, Sacks S, Scholl DR, Wright DK. Antiviral suscep-
tibility testing: Herpes simplex virus by plaque reduction assay; approved standard. Wayne,
PA: Clinical and Laboratory Standards Institute; 2004. p. 1–39.
216. Talarico CL, Phelps WC, Biron KK. Analysis of the thymidine kinase genes from acyclovir-­
resistant mutants of varicella-zoster virus isolated from patients with AIDS. J Virol.
1993;67:1024–33.
217. Tatarowicz WA, Lurain NS, Thompson KD. In situ ELISA for the evaluation of antiviral
compounds effective against human cytomegalovirus. J Virol Methods. 1991;35:207–15.
218. Tchesnokov EP, Gilbert C, Boivin G, Gotte M. Role of helix P of the human cytomegalovirus
DNA polymerase in resistance and hypersusceptibility to the antiviral drug foscarnet. J Virol.
2006;80:1440–50.
219. Telenti A, Smith TF. Screening with a shell vial assay for antiviral activity against cytomega-
lovirus. Diagn Microbiol Infect Dis. 1989;12:5–8.
220. Topalis D, Gillemot S, Snoeck R, Andrei G. Distribution and effects of amino acid changes
in drug-resistant alpha and beta herpesviruses DNA polymerase. Nucleic Acids Res.
2016;44:9530–54.
221. Tyring SK, Plunkett S, Scribner AR, Broker RE, Herrod JN, Handke LT, et al. Valomaciclovir
versus valacyclovir for the treatment of acute herpes zoster in immunocompetent adults: a
randomized, double-blind, active-controlled trial. J Med Virol. 2012;84:1224–32.
222. van der Beek MT, Vermont CL, Bredius RG, Marijt EW, van der Blij-de Brouwer CS, Kroes
AC, et al. Persistence and antiviral resistance of VZV in hematological patients. Clin Infect
Dis. 2013;56:335–43.
223. van Velzen M, van Loenen FB, Meesters RJ, de Graaf M, Remeijer L, Luider TM, et al.
Latent acyclovir-resistant herpes simplex virus type 1 in trigeminal ganglia of immunocom-
petent individuals. J Infect Dis. 2012;205:1539–43.
224. van Velzen M, Missotten T, van Loenen FB, Meesters RJ, Luider TM, Baarsma GS, et al.
Acyclovir-resistant herpes simplex virus type 1 in intra-ocular fluid samples of herpetic uve-
itis patients. J Clin Virol. 2013;57:215–21.
225. Vere Hodge RA, Field HJ. Antiviral agents for herpes simplex virus. Adv Pharmacol.
2013;67:1–38.
226. Visse B, Dumont B, Huraux JM, Fillet AM. Single amino acid change in DNA polymerase is
associated with foscarnet resistance in a varicella-zoster virus strain recovered from a patient
with AIDS. J Infect Dis. 1998;178:S55–7.
227. Visse B, Huraux JM, Fillet AM. Point mutations in the varicella-zoster virus DNA poly-
merase gene confers resistance to foscarnet and slow growth phenotype. J Med Virol.
1999;59:84–90.
228. Wade JC, Newton B, McLaren C, Flournoy N, Keeney RE, Meyers JD. Intravenous acyclo-
vir to treat mucocutaneous herpes simplex virus infection after marrow transplantation: a
double-blind trial. Ann Intern Med. 1982;96:265–9.
229. Wade JC, McLaren C, Meyers JD. Frequency and significance of acyclovir-resistant herpes
simplex virus isolated from marrow transplant patients receiving multiple courses of treat-
ment with acyclovir. J Infect Dis. 1983;148:1077–82.
230. Wald A, Corey L, Timmler B, Magaret A, Warren T, Tyring S, et al. Helicase-primase inhibi-
tor pritelivir for HSV-2 infection. N Engl J Med. 2014;370:201–10.
231. Weinberg A, Jabs DA, Chou S, Martin BK, Lurain NS, Forman MS, et al. Mutations confer-
ring foscarnet resistance in a cohort of patients with acquired immunodeficiency syndrome
and cytomegalovirus retinitis. J Infect Dis. 2003;187:777–84.
232. Whitley RJ, Gnann JW Jr. Acyclovir: a decade later. N Engl J Med. 1992;327:782–9.
233. Williamson EC, Millar MR, Steward CG, Cornish JM, Foot AB, Oakhill A, et al. Infections
in adults undergoing unrelated donor bone marrow transplantation. Br J Haematol.
1999;104:560–8.
8 Resistance of Herpesviruses to Antiviral Agents 267

234. Wolf DG, Lee DJ, Spector SA. Detection of human cytomegalovirus mutations associated
with ganciclovir resistance in cerebrospinal fluid of AIDS patients with central nervous sys-
tem disease. Antimicrob Agents Chemother. 1995;39:2552–4.
235. Wolf DG, Smith IL, Lee DJ, Freeman WR, Flores-Aguilar M, Spector SA. Mutations in
human cytomegalovirus UL97 gene confer clinical resistance to ganciclovir and can be
detected directly in patient plasma. J Clin Invest. 1995;95:257–63.
236. Wolf DG, Yaniv I, Honigman A, Kassis I, Schonfeld T, Ashkenazi S. Early emergence of
ganciclovir-resistant human cytomegalovirus strains in children with primary combined
immunodeficiency. J Infect Dis. 1998;178:535–8.
237. Wolf DG, Lurain NS, Zuckerman T, Hoffman R, Satinger J, Honigman A, et al. Emergence
of late cytomegalovirus central nervous system disease in hematopoietic stem cell transplant
recipients. Blood. 2003;101:463–5.
238. Zahn KE, Tchesnokov EP, Gotte M, Doublie S. Phosphonoformic acid inhibits viral replica-
tion by trapping the closed form of the DNA polymerase. J Biol Chem. 2011;286:25246–55.
239. Ziyaeyan M, Alborzi A, Japoni A, Kadivar M, Davarpanah MA, Pourabbas B, et al. Frequency
of acyclovir-resistant herpes simplex viruses isolated from the general immunocompe-
tent population and patients with acquired immunodeficiency syndrome. Int J Dermatol.
2007;46:1263–6.
Chapter 9
Heteroresistance: A Harbinger of Future
Resistance

Karl Drlica, Bo Shopsin, and Xilin Zhao

9.1 Introduction

Heteroresistance is a condition in which a microbial population contains subpopula-

tions whose minimal inhibitory concentration (MIC) is above the resistance break-
point, while the bulk population MIC is below that breakpoint. Since heteroresistant
infections usually respond favorably to antimicrobial treatment, largely due to
effective host defense systems, heteroresistance has often been seen as a minor
problem for treating individual patients. However, when heteroresistance is consid-
ered as an intermediate state in the evolution to resistance, it is a warning sign – a
window through which we can see the future.
Emergence of resistance is important for individual patients with three diseases:
tuberculosis, malaria, and HIV disease. On a global basis, these diseases are top-­
ranked in terms of mortality. However, in industrialized countries, individual
patients are more troubled by horizontal transmission of resistance, especially with

K. Drlica (*)
Public Health Research Institute, New Jersey Medical School,
Rutgers Biomedical and Health Sciences, Newark, NJ, USA
e-mail: [email protected]
B. Shopsin
Departments of Medicine and Microbiology, New York University School of Medicine,
New York, NY, USA
X. Zhao
Public Health Research Institute, New Jersey Medical School, Rutgers Biomedical
and Health Sciences, Newark, NJ, USA
Department of Microbiology, Biochemistry, & Molecular Genetics, New Jersey Medical
School, Rutgers Biomedical and Health Sciences, Newark, NJ, USA
State Key Laboratory of Molecular Vaccinology and Molecular Diagnostics, School of Public
Health, Xiamen University, Xiamen, Fujian Province, China

© Springer International Publishing AG, part of Springer Nature 2018 269

I. W. Fong et al. (eds.), Antimicrobial Resistance in the 21st Century, Emerging Infectious
Diseases of the 21st Century, https://doi.org/10.1007/978-3-319-78538-7_9
270 K. Drlica et al.

opportunistic infections caused by bacteria such as Staphylococcus aureus,

Escherichia coli, and Klebsiella pneumoniae [1, 2]. For these organisms, the de
novo emergence of resistance in individual patients is thought to occur rarely or to
have little clinical consequence. In the absence of immediate consequences, little
incentive has existed to implement dosing strategies for restricting the emergence of
resistance [3]. That perspective may soon change as heteroresistance becomes
increasingly common with many different opportunistic pathogens.
The present chapter begins with a brief overview of heteroresistance by consider-
ing detection methods, types of genetic changes involved in heteroresistance, and
general resistance features of pathogen categories. We then consider the phenomenon
in two phylogenetically distant pathogens, Mycobacterium tuberculosis and methi-
cillin-resistant S. aureus (MRSA). These two organisms, both of which pose serious
antimicrobial resistance problems, serve to illustrate how the path to resistance
depends on the pathogen, the drug, the fitness of pathogen variants, and the epidemi-
ology of infection. With M. tuberculosis, many of the genetic changes associated
with heteroresistance are the same as those causing complete resistance. Thus, DNA-
based detection methods are practical. With MRSA, we see a situation in which fit-
ness costs limit the evolution of vancomycin resistance to an intermediate state called
VISA (vancomycin-intermediate S. aureus). Multiple evolutionary pathways to
VISA exist, which makes the development of DNA tests challenging. Examination of
these two pathogens may eventually lead to an understanding of factors that deter-
mine the outcome of host-pathogen-antimicrobial encounters. Moreover, the result-
ing framework should help us predict failure of particular therapeutic interventions.
We conclude the chapter by surveying other pathogens for which heteroresistance is
beginning to threaten standard surveillance and diagnostic procedures. In sum, het-
eroresistance is an under-reported phenomenon that will become increasingly impor-
tant as we move deeper into the era of antimicrobial resistance. Readers interested in
an earlier review of heteroresistance are referred to Ref. [4].

9.2 Overview of Heteroresistance

9.2.1 Detection of Heteroresistance

Heteroresistance manifests itself in several ways. The most graphic is the growth of
bacterial colonies within a zone of inhibition created when an antimicrobial diffuses
from a central source on agar that had been covered with bacteria prior to incubation
(for example, see Fig. 1 in Ref. [7]). If the colonies in the inhibition zone test posi-
tive for antimicrobial resistance using assays that measure minimal inhibitory con-
centration (MIC), the overall population is said to be heteroresistant. When those
resistant colonies continue to test resistant following multiple rounds of growth in
or on drug-free medium, the heteroresistance is said to be stable. Many examples
exist in which the resistance phenotype is lost during subculturing in the absence of
drug. Those situations are called unstable heteroresistance. The “colonies within the
inhibition zone” is the easiest method for detecting heteroresistance and is
9 Heteroresistance: A Harbinger of Future Resistance 271

Resistant

Colonies recovered
Heteroresistant

Wild type

Drug concentration

Fig. 9.1 Population analysis profile. A bacterial culture or specimen is applied to a series of agar
plates containing different concentrations of the test antimicrobial. After incubation to allow col-
ony formation, colonies are counted and plotted for each drug concentration. A resistant culture is
unaffected by the drug until concentrations are very high, while a fully susceptible culture (wild
type) exhibits a sharp drop in colony recovery at MIC. Results for a heteroresistant culture contain-
ing a variety of subpopulations having reduced susceptibility are depicted. (Data for hVISA can be
seen in Refs [5, 6])

commonly used as an initial screen with samples that would otherwise be scored as
susceptible by diagnostic laboratories.
With some pathogens, discordance in susceptibility testing indicates heteroresis-
tance. For example, with M. tuberculosis, DNA tests may indicate the presence of
mutations associated with drug resistance, while drug susceptibility testing (deter-
mination of MIC) indicates that the isolate is in the drug-susceptible category.
Discordance can also occur between liquid-growth and agar-plate tests. In both situ-
ations the discordance arises from assay sensitivity differences.
The gold standard for establishing heteroresistance is detection of “resistant”
subpopulations in a population analysis profile (PAP, Ref. [8]). For this assay, a
series of agar plates is prepared such that each plate contains a different concentra-
tion of drug. A large number of cells, generally >106, are applied to each agar plate,
and after incubation at the appropriate growth temperature (usually 37 °C), the
number of colonies is scored. A fully susceptible pathogen isolate will exhibit a
sharp drop in colony number when the drug concentration in the agar reaches
MIC. In contrast, a heteroresistant isolate will show colonies at concentrations
above MIC. The resulting plot of colony number versus drug concentration is the
population analysis profile (Fig. 9.1). The area under the curve (AUC) generated by
the PAP provides an integrated description of the heteroresistant subpopulations;
normalization to a reference strain lacking detectible heteroresistance provides a
single number for comparing the heteroresistance status of pathogen samples.
Although PAP can be readily applied to any pathogen that forms colonies on
solid medium, including mycobacteria [9, 10] and fungi [11], the method is very
labor intensive. Thus, it is generally used only for research purposes or to confirm
the presence of heteroresistance in a clinical setting. For research it is important to
recognize that incubation time can be a factor if the antimicrobial induces resis-
272 K. Drlica et al.

tance: in the case of fluoroquinolones, the number of colonies increases dramati-

cally over the course of 2 weeks with rapidly growing bacterial species [12].
An important issue for all detection methods is how the patient specimen is han-
dled (material directly collected from a patient is called a specimen). When speci-
mens are examined without subculturing, the percent heteroresistance reflects
pathogen subpopulations at a particular location within a given patient at a particu-
lar time. Such specifications are important with diseases such as tuberculosis,
because considerable heterogeneity exists within a patient (discussed below).
Frequently a specimen is plated on agar prior to testing the predominant colonies
for drug susceptibility. Clonal expansion of those colonies generates a sample called
an isolate. When an isolate tests positive for heteroresistance, heterogeneity could
have been produced during expansion of the culture. Such isolates would have an
elevated propensity to generate heteroresistance. For example, a gene amplification
might occur more frequently in such an isolate, or a mobile resistance element
might be lost from some cells in the population. The percent of isolates showing
heteroresistance reflects the prevalence of patients having a heteroresistance-prone
infection. In contrast, direct examination of specimens reflects both the resistant
subpopulation within an individual patient and the prevalence of patients harboring
heteroresistance-prone clones.

9.2.2 Types of Heteroresistance

Antimicrobial heteroresistance represents a point along the evolutionary path that

pathogens take toward complete resistance or, in some cases, the loss of a resistance
element that exerts an excessive fitness cost in the absence of antimicrobial. The
path varies considerably among pathogen and antimicrobial species [4]. In some
cases, multiple paths exist. A major distinction among heteroresistance types con-
cerns their origin. In one type, heterogeneity arises from coinfection with multiple,
dissimilar infecting strains. Such a situation may be common with tuberculosis due
to spread of disease from one person to another that leads to superinfection (dis-
cussed below). Alternatively, diversity can evolve along clonal lines; this is the
usual scenario when superinfection is rare. Clonal heteroresistance, in turn, has two
forms. One is derived from infection by a single pathogen cell followed by clonal
expansion; the other derives from infection by multiple cells followed by clonal
expansions.
Another major distinction is whether the diversity is genetically stable. Fitness is
an important consideration, as some resistance features are maintained in the popu-
lation only when antimicrobial pressure is present, while others persist through
multiple passages in drug-free medium. In Fig. 9.2 we illustrate common types of
clonal heteroresistance.
9 Heteroresistance: A Harbinger of Future Resistance 273

Fig. 9.2 Common types of clonal heteroresistance. Four general themes are shown schematically.
Type I represents the acquisition of a resistance mutation, either spontaneously or by horizontal
transfer, that is maintained in the population. Resistant subpopulations are enriched with each
antimicrobial challenge until resistant cells dominate the population. Examples are fluoroquino-
lone and rifampicin resistance in M. tuberculosis. In Type II heteroresistance, antimicrobial pres-
sure is needed to maintain the resistant phenotype as it is enriched. When drug pressure is removed
or relaxed, susceptible members of the population regain dominance. An example of this type of
heteroresistance is represented by gene amplification in S. enterica. Type III illustrates a situation
in which multiple pathways lower susceptibility and also reduce pathogen fitness. Fitness prob-
lems can limit the loss of susceptibility to a state called intermediate resistance. An example of
Type III heteroresistance is seen with vancomycin-intermediate S. aureus. In Type IV heteroresis-
tance, a resistance determinant enters the population by horizontal transfer and is rapidly enriched
due to continuing horizontal transfer. If the element is unstable, it can be lost when antimicrobial
pressure is reduced. Heteroresistance is seen as a balance of resistance acquisition, loss, and anti-
microbial pressure. An example of Type IV heteroresistance is methicillin-resistant S. aureus

9.2.3 Heteroresistance and Antimicrobial Tolerance

Tolerance is a situation in which pathogen growth (reproduction) is blocked by an

antimicrobial, but the pathogen is not killed. In contrast, resistant pathogens repro-
duce in the presence of the antimicrobial, and susceptible ones die. As with resistance,
genes are involved in some types of tolerance [13]. Phenotypic tolerance derives from
environmental conditions that block antimicrobial lethality. For example, some
β-lactams and first-generation quinolones fail to kill E. coli in cultures that have been
grown to stationary phase. The clinical danger from tolerant pathogens is population
outgrowth following removal of the antimicrobial; in contrast, heteroresistant patho-
gens are dangerous even during treatment. Tolerance, or persistence as it is sometimes
called, is particularly problematic with tuberculosis – it is estimated that a third of the
global human population is infected with M. tuberculosis in a tolerant, asymptomatic
state called latency. For additional discussion of tolerance see Chap. 13.
274 K. Drlica et al.

9.2.4 Pathogen Types Displaying Heteroresistance

Although a single cell can acquire resistance in a single step, pathogen populations
generally require amplification of the resistant cell. During selective amplification,
the population will be heteroresistant. Thus, heteroresistance may be a general
aspect of the emergence of resistance.
Among infections exhibiting heteroresistance are those caused by commensal
bacteria that occasionally act as pathogens. Important examples include MRSA,
vancomycin-resistant Enterococcus (VRE), Acinetobacter baumannii, Escherichia
coli, and Klebsiella pneumoniae. Repeated antimicrobial exposure, sometimes
aimed at other bacterial species, results in subpopulations of resistant mutants.
These mutant subpopulations can be enriched during treatment and thereby restrict
therapeutic options when the microbes cause infection. E. coli, a common inhabit-
ant of the human digestive tract, serves as an example. Fluoroquinolone treatment
for a variety of reasons unrelated to E. coli populations can selectively enrich
fluoroquinolone-­resistant E. coli in the digestive tract. If these organisms contami-
nate the urinary tract, they can cause fluoroquinolone-resistant urinary infection,
which is now a global problem [14].
Heteroresistance is also associated with pathogens for which infection is required
for transmission. Among these obligate pathogens are M. tuberculosis and the
human immunodeficiency virus (HIV). Spontaneously resistant mutants emerge
readily, which makes every treated patient at risk for developing a resistant infec-
tion. As a result, the standard of care involves the use of multiple antimicrobials.
With tuberculosis, the antimicrobials are administered daily by healthcare workers
to assure adherence to treatment protocols. In the next section, we consider hetero-
resistance in M. tuberculosis as an example of emerging resistance with an obligate
pathogen.

9.3 Heteroresistance with Mycobacterium tuberculosis

9.3.1  mergence of Resistance in Individual Tuberculosis

E
Patients

With individual patients, monotherapy for tuberculosis often leads to the emergence
of resistance and treatment failure [15–17]. Host defense systems appear unable to
readily clear M. tuberculosis, and thus some tuberculosis patients harbor large num-
bers of pathogen cells (on the order of 109) [18, 19]. This feature, coupled with the
early finding that cell cultures contain resistant mutants at a high frequency (about
10−6 for isoniazid and 10−8 for rifampicin and streptomycin [17, 20–22]), led to the
idea that monotherapy simply enriches existing mutant subpopulations. More recent
measurement of mutation rate, which avoids the jackpot effects of frequency assays,
suggests that mutation rate for cultured M. tuberculosis is similar to that of other
9 Heteroresistance: A Harbinger of Future Resistance 275

bacteria (discussed by McGrath et al. [23]). Thus, lack of immune clearance, which
is especially obvious with patients coinfected with HIV, results in heavy bacterial
burden; the bacterial load is probably a key factor in the emergence of resistance
rather than an abnormally high mutation rate. With M. tuberculosis, mutagenesis
can be induced by DNA damage [24], which likely contributes to the mutagenic
effect of some antimicrobials. As expected, combination therapy largely overcomes
the rapid emergence of resistance [25].
Enrichment of mutant subpopulations is further favored by the long treatment
time required to achieve cure. During infection, M. tuberculosis is thought to
remodel its metabolism such that part of the bacterial population enters a drug-­
tolerant, semi-quiescent state known as dormancy (persistence). When this state is
modeled in the laboratory, dormant bacteria are difficult to eradicate [26]. In addi-
tion, infection occurs in diverse compartments [27, 28], some of which may not be
readily accessible to active compounds [28]; moreover, some cells may be non-­
culturable but viable. Consequently, antibiotic treatment must be maintained for
many months to be effective. When adherence to treatment is poor, drug exposure
becomes intermittent, which allows cycles of population expansion followed by
selective reductions. These cycles also occur with many other pathogens but usually
in different patients rather than in a single individual. The easily observed emer-
gence of resistance with M. tuberculosis has made tuberculosis a paradigm for
understanding the process.
Awareness of tuberculosis heteroresistance is high in part because PCR-based
detection of resistant subpopulations is straightforward: resistance arises from point
mutations in an otherwise highly conserved pathogen genome (reviewed in [23]).
Moreover, the need for rapid diagnosis has been a high priority, which has led to the
widespread application of DNA-based methods. These tests now show that 10–20%
of patients in localities of high tuberculosis incidence have infections containing
diverse subpopulations. Heterogeneity is seen with both HIV-positive and HIV-­
negative patients [29], and it is detected with a variety of genes, including those that
encode resistance to ethambutol [30–33], isoniazid [32–34], rifampicin [34], fluoro-
quinolones [32, 35], streptomycin [33], pyrazinamide [32], and amikacin [36].
Thus, M. tuberculosis heteroresistance within individual patients is common to
many antimicrobials.

9.3.2 Two Forms of Heterogeneity

DNA analyses of M. tuberculosis specimens reveal two general types of heterogene-

ity. In one form, bacterial isolates contain subpopulations having very different
DNA fingerprints (IS6110 RFLP or VNTR patterns) [37]. This form of heteroresis-
tance is generally attributed to mixed infections arising from superinfection, or per-
haps coinfection if the initial inoculum contained multiple bacterial cells having
mixed genotypes [38–40]. Mixed-clone infections tend to occur in localities where
tuberculosis burden is high and resistant disease is common. In an example from
276 K. Drlica et al.

Tashkent, Uzbekistan [34], sputum samples from 35 patients were examined by

culture-based, drug-susceptibility testing and by a variety of DNA-based methods.
Seven of the 35 samples contained susceptible cells mixed with cells resistant to
isoniazid, rifampicin, or both. By DNA analysis, five of the seven heteroresistant
isolates were shown to contain different strains, which indicated mixed infection
due to coinfection or superinfection. Three of the mixed infections were newly diag-
nosed in patients who had not been treated; thus, continuous antimicrobial pressure
is not required to observe mixed infection. These cases of mixed infection, which
derived from dissemination of resistant M. tuberculosis, have been taken as evi-
dence for inadequate infection control (isolation of patients, controlled air flow,
etc.). Mixed infection is likely due to multiple factors.
In two heteroresistant cases from the Tashkent collection, the resistant subpopu-
lations and the major, susceptible population had very similar DNA fingerprints
[34]. Although the study did not show identical fingerprints and although it lacked
the whole-genome sequencing or epidemiological information required to establish
a de novo origin for the major and minor populations, clonal relation is the most
likely explanation. Inadequate treatment and poor adherence to therapy regimens,
rather than lax infection control, are the likely causes of this type of
heteroresistance.
The experience of an HIV-positive Italian tuberculosis patient, who failed to
adhere to treatment protocols for more than a decade, is best explained by evolution
to resistance within an individual host. The patient first exhibited a fully susceptible
infection that appeared to respond to therapy [41]. But after 3 years, the dominant
isolate exhibited resistance to rifampicin and streptomycin. A subsequent sample
contained a mixture of streptomycin-resistant and streptomycin-susceptible cells.
The original streptomycin-resistance marker was later replaced by a different allele
that became fixed in the patient. Eventually the strain, which had the same DNA
fingerprint throughout, became resistant to rifampicin, streptomycin, isoniazid, and
pyrazinamide. Had the patient lived to continue treatment with other agents, his
pathogen population could have acquired even more resistance markers: some iso-
lates from New York City have nine different resistance markers [42].
M. tuberculosis heterogeneity is not restricted to drug-resistance markers. For
example, in a collection of largely pan-susceptible specimens from Bangladesh
[37], ten colonies were examined from each of 97 samples. When DNA analyses
(spoligotyping and IS6110 RFLP tests) were applied, most samples had identical
DNA patterns for all ten replicate colonies. However, with eight specimens, repli-
cate colonies contained similar but nonidentical DNA fingerprints. That result was
taken as evidence for clonal heterogeneity. Only two specimens had DNA finger-
prints that were distinct enough for the samples to be from mixed-clone infections.
A similar finding of mixed infections has been reported for samples collected in
Georgia [43].
9 Heteroresistance: A Harbinger of Future Resistance 277

9.3.3 Dynamics of Clonal Evolution

A study from South Africa illustrates the complex dynamics of clonal heteroresis-
tance [32]. The subjects of the study suffered from multidrug-resistant tuberculosis
(MDR-TB) that had persisted through more than 12 months of treatment. Since the
overall prevalence of MDR-TB in the community was low (0.3% in new patients,
1.7% in previously treated patients), clonal heterogeneity was more likely to occur
than mixed infections. Indeed, when sputum samples from 13 HIV-negative
MDR-TB patients were examined at 2-week intervals, no evidence was found for
superinfection: all carried bacteria having a single IS6110 RFLP type and spoligo-
type pattern. Nucleotide sequence analysis for eight resistance genes (katG,
inhA = isoniazid; pncA = pyrazinamide; embB = ethambutol; rrs = amikacin, kana-
mycin; rpsL = streptomycin; gyrA = fluoroquinolone; rpoB = rifampicin) showed
that several of the infections changed resistance patterns over the course of
sampling.
One patient in the South African study [32] was examined for mutations in katG,
embB, and gyrA during 56 weeks of therapy. At the beginning of sampling, all three
genes were wild type, but at weeks 4 and 6, the katG marker was scored as resistant.
Subsequent samples showed that it returned to wild type. The embB marker con-
verted to resistant by 6 weeks, and it remained resistant throughout the observation
period. The gyrA gene showed a mixture of alleles at week 6, and in subsequent
samples transient changes were observed among several gyrA resistance forms,
often mixed with wild-type alleles. Even after 48 weeks, gyrA was a mixture of
resistant (D94C) and wild-type alleles. By week 52 a different gyrA allele (D94G)
had emerged as the dominant form.
Specimens from two other patients [32] also contained different alleles of genes
involved in drug resistance. For example, one patient evolved a mixture of wild-type
and resistant alleles for pncA, changes in gyrA alleles over time, and a mixture of rrs
alleles at the beginning of sampling that later saw one allele emerge as dominant.
Another patient began with wild-type gyrA that after 36 weeks changed to resistant.
But after 48 weeks, gyrA returned to wild type. Wild-type pncA also persisted until
week 36, and then it shifted to resistant for the remainder of the study (week 52).
The katG gene began as wild type, but after week 6 it was resistant, except for one
sample at week 30 that was wild type. Overall, these fluctuations in drug-resistance
markers illustrate the dynamic and varied nature of clonal heteroresistance when
sputum samples are the source of information.
Examination of lung tissue provides an explanation for the allelic diversity: het-
eroresistance measured in sputum samples arises at least in part from independent
clonal evolution in different regions of the lung. When surgical samples were exam-
ined from three HIV-negative patients having undergone long-term therapy, DNA
IS6110 fingerprints were the same for bacteria from different regions of the lung;
thus, the isolates from individual patients appeared to be clonally related [27]. One
patient had a streptomycin-resistant strain in an open lesion, while wild-type cells
were detected in a closed granuloma. Wild-type cells were also recovered from
278 K. Drlica et al.

sputum. Bacteria from a second patient carried two different gyrA resistance alleles
when isolated from open lesions, while wild-type gyrA alleles were obtained from
sputum and from two closed lesions. The third patient produced three types of M.
tuberculosis: (1) cells from apparently normal lung tissue had wild-type genes for
katG, embB, and rrs, (2) cells from sputum and four pathological sites had katG and
embB resistance markers but wild-type rrs, and (3) another pathological site yielded
bacteria with resistance for all three genes. These observations, plus similar findings
in another study [44] and autopsies [45], led to the conclusion that evolution occurs
independently in different lung compartments and that wild-type cells can survive
treatment. The results of sputum-based analyses probably reflect opening of granu-
lomas and release of bacteria at different times during infection. Thus, analysis of a
single sputum sample may not accurately reflect the diversity of bacterial popula-
tions in the infection.

9.3.4 Consequences of Heteroresistance

In the early days of tuberculosis chemotherapy, diagnostic criteria were set up to

avoid mistakenly identifying a susceptible strain as resistant, because to do so would
deprive a patient of a useful treatment. For example, with the agar proportion
method, a specimen is considered to be resistant only if at least 1% of the colonies
are resistant [22] (when resistance is identified at the 1% level, enrichment to full
resistance requires only seven generations of selective growth, roughly 1 week for
M. tuberculosis). We conclude that 1% heteroresistance is a late stage in the evolu-
tion to resistance. Such an infection can still be treated with combination therapy,
but close monitoring and treatment adjustment are required to avoid conditions that
enrich mutant subpopulations.
Failure to recognize resistant subpopulations leads to inappropriate treatment,
the expansion of those subpopulations, and eventually full resistance [46]. In one
example, infection with M. tuberculosis was scored as susceptible at the time of
diagnosis, but after 3 months of first-line therapy, MDR tuberculosis was diagnosed
by drug susceptibility testing [38]. Retrospective analysis, using strain-specific
PCR-based methods, showed that an MDR subpopulation had been present through-
out treatment [38].
An added complication is that interruption of treatment can lead to reemergence
of susceptible M. tuberculosis. In an example from South Africa, the susceptible
subpopulation was not eradicated, even by 17 months of therapy; at treatment inter-
ruption, susceptible bacteria repopulated the infection [38]. In another example, an
MDR infection was treated with second-line agents, and after 3 months of treat-
ment, the infection was judged fully susceptible [38]. Reduced antibiotic pressure,
as may occur with second-line agents, allowed the susceptible strain to become
dominant. In such cases, treatment needs to be reassessed periodically, and perhaps
first-line therapy needs to be continued with MDR strains even after applying
second-­line agents. To maintain adequate therapy while resistance markers are
9 Heteroresistance: A Harbinger of Future Resistance 279

changing requires rapid and accurate diagnostic methods. Below we consider the
development of genetic (DNA-based) assays.

9.3.5 Detecting Heteroresistance Using DNA-Based Methods

Although M. tuberculosis population heterogeneity had been known for many years
from phage-typing of M. tuberculosis subpopulations [47, 48], it was recognized as
an important phenomenon only after molecular diagnostic methods emerged. When
PCR was used to amplify specific regions of M. tuberculosis DNA encoding pro-
teins associated with resistance and the amplified fragments were separated by gel
electrophoresis, the size distribution characteristic of both susceptible and resistant
alleles was observed from a single bacterial specimen [33]. Heteroresistance was
then used to explain the occasional discordance between results from drug-­
susceptibility testing and DNA-based methods: the DNA tests indicated resistance,
but only susceptibility was detected following the bacterial outgrowth required for
susceptibility testing. A fitness advantage among the susceptible bacteria was
thought to allow them to dominate during outgrowth [49].
The various DNA-based tools differ in sensitivity (Table 9.1). For example,
Sanger DNA sequencing of PCR products reported 15% of isolates as heteroresis-
tant, while with the same samples deep sequencing found almost 40% heteroresis-
tance [35]. When heteroresistance is greater than a few percent, current hybridization
methods are sufficiently sensitive. Unfortunately, PCR-based diagnostic methods
encounter a specificity problem when subpopulations are below 1%, because tem-
plates from the major bacterial population can generate false-positive, variant sig-
nals due to mis-priming, mis-incorporation, and mis-hybridization.
As pointed out above, sensitivity to 1% is unlikely to be adequate for monitoring
the emergence of resistance, because 1% is considered fully resistant for that marker
if the equivalent of monotherapy is employed. Moreover, a negative result cannot
rule out heteroresistance. In essence, current genetic diagnostics can give false-­
negative results. We conclude that other methods are needed to detect heteroresis-
tance at levels low enough to allow successful intervention.
Work in cancer biology is driving new, DNA-based tests for heteroresistance – a
priority in the cancer field is detection of a small number of transformed cells within
a large background of normal cells. One approach is called digital PCR [57]. In this
method, the sample is diluted into a series of wells in a multi-well microfluidic plate
such that only a single molecule of mutant DNA is expected to be present in a given
well (most wells will contain only wild-type DNA). Amplification of DNA in the
wells produces a digital readout: either the presence or absence of mutant DNA. The
fraction of total wells scoring positive estimates the percent of the sample contain-
ing mutant DNA. In principle, the sensitivity of this method is limited only by the
number of wells assayed. Digital PCR has been applied to M. tuberculosis isolates
by mixing wild-type cells with M. tuberculosis containing resistance mutations in
katG, rpoB, gyrA, and rrs. The method reliably detects heteroresistance at a ratio of
280 K. Drlica et al.

Table 9.1 Sensitivity of DNA-based detection methods for heteroresistance

Resistance Size of detectable
Method Gene(s)a sub-population Reference
Sanger sequencing katG, fqn, rif, rrs 28–60% [35, 50, 51,
katG 50% 52]
rif 50%
fqnb 15%
Melting curve inh 40% [53]
Sloppy molecular rif 40% [54, 55]
beacons fqn 5–10%
qPCR bacteriophage ns 10% [50]
qPCR katG, fqn, rif, rrs 10% [50]
Line probe katG 5% [51]
rif 5%,1–70%c [51, 56]
iPLEX amk 0.5% [36]
Digital PCR katG, fqn, rif, rrs 0.1% [50]
a
Abbreviations: ns not stated, amk amikacin, fqn fluoroquinolone, katG isoniazid, inh various iso-
niazid markers, rif rifampicin, rrs aminoglycosides
b
Fluoroquinolone resistance was the only marker in the population; deep sequencing identified
38% heteroresistant
c
Depends on allele

1 mutant per 1000 wild-type cells [50], which is about 10-times more sensitive than
previous PCR-based methods. For digital PCR to achieve this sensitivity with spu-
tum samples, the samples must contain more than 1000 M. tuberculosis cells per ml
(bacillary content varies among sputum samples, but it can exceed one million cfu
[58–60]).
Another approach, single-nucleotide primer extension, is used to incorporate a
nucleotide having a distinctive mass modification that can be identified by mass
spectroscopy [61]. The method, called iPLEX Gold, has the advantage of detecting
multiple resistance alleles in the same reaction mixture. In one application of the
method, a reconstruction experiment detected one amikacin-resistant cell per 200
wild-type cells [36].
A third strategy is called pyrophosphorolysis-activated polymerization [62–67].
In this method, a primer containing a dideoxyribonucleotide at its 3′ terminus (noted
as P*) is hybridized to the test DNA at the preselected mutation site. Removal of the
dideoxyribonucleotide by pyrophosphorolysis, which is highly specific for perfect
hybridization of the primer, is required for extension of the primer by DNA
­polymerase. Primer extension then amplifies the signal for real-time detection by
fluorescent probes. When a P* primer is used that contains the complement of the
mutant sequence, the polymerization assay is expected to detect mutant alleles at a
frequency as low as 10−8 of wild-type DNA, a level that approaches background
(spontaneous) mutation frequency. To our knowledge, the pyrophosphorolysis-­
activated polymerization method has not been applied to detection of heteroresis-
tant M. tuberculosis.
9 Heteroresistance: A Harbinger of Future Resistance 281

A fourth strategy, which is also derived from cancer diagnosis, uses what are
called SuperSelective primers for real-time PCR assays [68]. In this system, a DNA
primer is designed in which one region hybridizes strongly to a portion of the target
DNA being queried. This anchor region is separated from a detector region, the
“foot”, by a long region expected to mispair with the target and thus form a loop.
The foot is designed to hybridize only with the mutant sequence in the target. By
adjusting the length of the loop and the foot, conditions can be obtained in which
hybridization only occurs with mutant DNA. The resulting hybrid then primes real-­
time PCR. The system can detect multiple mutations in the same reaction tube by
using fluorophores having different colors to discriminate the amplification prod-
ucts. This method has not yet been applied to diagnosis of heteroresistance.
A fifth strategy is based on CRISPR, a bacterial system that recognizes and
destroys foreign nucleic acids. The underlying idea is as follows. A DNA sample
from the pathogen is transcribed in vitro and incubated with the Cas13a protein
system plus a quenched, fluorescently labeled reporter RNA. Recognition of the
target RNA by Cas13a, which is designed to occur only if the resistance mutation is
present, will cause collateral damage in the reporter RNA, eliminate the quenching,
and generate a fluorescent signal. This method, which has been dubbed SHERLOCK
[69], has single-molecule sensitivity, similar to droplet digital PCR and quantitative
PCR (qPCR). Moreover, it has point-of-care diagnostic features. To our knowledge
SHERLOCK has not been applied to detection of heteroresistant M. tuberculosis.
However, the CRISPR system has been modified to function in this pathogen [70].
A general problem associated with PCR-based diagnosis of resistant bacterial
subpopulations is laboratory contamination by amplicons present in the laboratory
from previous tests. Published estimates of laboratory cross-contamination using
open-tube methods are presently almost 4% [38, 40]. Although closed-tube methods
exist [53, 71, 72], current closed-tube methods require refinement to be sensitive
enough for heteroresistance detection.

9.4 Heteroresistance with Staphylococcus aureus

9.4.1 Methicillin Heteroresistance

Heteroresistance with S. aureus, which has been known for many years [73], is not
routinely detected by standard susceptibility testing (MIC determination). Such
determinations typically examine only 104 to 105 cells, and the frequency of resis-
tant subpopulations is generally below 10−5. However, when susceptibility testing
uses a large number of cells, on the order of 107 to 1010, subpopulations having
reduced susceptibility can be seen. For example, heterogeneity is a distinctive fea-
ture of methicillin resistance due to the presence of a mobile chromosomal element
called SCCmec [6, 74]. SCCmec elements, which vary in size, contain a gene, mecA,
that encodes a low-affinity penicillin-binding protein (PBP2’ or PBP2a). PBP2’ is a
transpeptidase [75] that allows S. aureus to form cell walls in the presence of
282 K. Drlica et al.

methicillin. Clinical isolates carrying mecA usually show moderate-level heteroge-

nous resistance to all β-lactams [6]. However, subpopulations emerge in which
resistance levels are high. Indeed, repeated β-lactam challenge leads to homo-resis-
tant S. aureus, sometimes, but not always, due to enhanced mecA expression [6, 74].
Most homogenous, high-level resistance strains revert to heterogeneity, although
some laboratory isolates, such as strain COL, demonstrate stable high-level resis-
tance. Overall, the evolution of methicillin heteroresistance is a classic example of
antimicrobial resistance emerging in an opportunistic pathogen.
Early work identified a chromosomal mutation, chr*, as being important for
high-level methicillin resistance [74]. Whole-genome sequencing and genetic
reconstruction experiments subsequently showed that at least one type of chr* is a
substitution in the β subunit of RNA polymerase that, along with mecA, confers
high-level resistance to methicillin [76]. The molecular basis for rpoB action on
mecA is unknown, but it is likely to be important, because RNA polymerase substi-
tutions are also involved in intermediate resistance to vancomycin [77]. One specu-
lation is that the RNA polymerase variants alter the expression of genes that protect
from antimicrobial activity.
One of the more relevant examples of S. aureus heteroresistance concerns cef-
taroline, a cephalosporin (β-lactam) that shows activity against MRSA and
vancomycin-­intermediate S. aureus (VISA, discussed below). In one study, a col-
lection of 57 isolates contained 12 heteroresistant members, some of which also
exhibited reduced susceptibility to vancomycin, daptomycin, or linezolid [78]. We
conclude that controlling MRSA with new β-lactams is likely to be difficult.

9.4.2 Vancomycin-Intermediate Heteroresistance

MRSA infection is commonly treated with the glycopeptide vancomycin. The result
has been the emergence of an intermediate level of resistance (VISA, which is dis-
tinct from the rare, vanA-mediated, fully vancomycin-resistant S. aureus). VISA is
associated with a poorly defined thickening of the bacterial cell wall that reduces the
uptake of vancomycin [79]. Other features associated with VISA are excess pepti-
doglycan production, low fitness manifested by reversion toward susceptibility dur-
ing growth in vitro [80], and attenuated virulence in animal models of infection
[81–84]. VISA probably represents the end stage of evolution from heteroresistant
strains (hVISA) in which subpopulations slightly elevate the overall MIC of an
MRSA isolate. Since clinical isolates of MRSA are heteroresistant due to instability
of the SCCmec elements [85], hVISA emerging during vancomycin treatment can
be co-heteroresistant (heteroresistant for two or more antimicrobials).
With both hVISA and VISA, S. aureus populations exhibit considerable hetero-
geneity in their susceptibility to vancomycin (see examples in Ref. [5]). To better
detect hVISA, the breakpoint for full susceptibility was lowered to an MIC of 2 μg/
ml [86]. In some samples, hVISA cells are abundant enough to raise vancomycin
9 Heteroresistance: A Harbinger of Future Resistance 283

MIC to the high end of the susceptible range (MIC = 0.5 to <2 μg/ml; with most
antimicrobial-pathogen combinations, heteroresistance has no effect on MIC
because the subpopulations are small). Thus, a slightly elevated MIC can indicate
hVISA, but a more sensitive test is needed.
PAP (population analysis profile) determination can definitively identify hVISA,
but the method is too labor intensive for routine clinical use. Consequently, efforts
have shifted toward other agar-plate screens [87]. Developing DNA tests is still
challenging, because multiple genetic pathways lead to VISA. For example, when
seven successive samples of MRSA were obtained from a single patient and ana-
lyzed by vancomycin-PAP, a clear evolution from susceptible to hVISA to VISA
was observed [88]. Whole-­genome sequencing then identified six mutations that
generated five distinct genetic profiles that correlated with evolution along three
pathways involving cell wall biology.
In another whole-genome sequencing study, a VISA strain was compared to a
closely related, susceptible isolate. Several gene differences were found, and wild-­
type alleles were introduced into the VISA strain to regain susceptibility and thereby
identify genes involved in VISA [79, 89]. Among these were the GraSR and VraSR
two-component systems that contribute to the evolution of VISA by upregulating
cell wall synthesis (for additional detail, see Chap. 15). Other associated alterations,
listed in Ref. [77], include a substitution in RNA polymerase (H481Y/L/N) that also
confers rifampicin resistance (whether rifampicin resistance is acquired before
VISA is not known). We conclude that unlike the situation with M. tuberculosis
heteroresistance, a simple DNA-based diagnostic for hVISA is not likely to be
available soon.
The clinical importance of VISA and hVISA can be assessed by surveillance
studies. VISA represents a few percent of MRSA recovered from serious infections,
and hVISA prevalence is 4 to 5 times higher [90, 91]. Apparently, the genetic
changes that create VISA have fitness costs that trap most S. aureus in the
intermediate-­resistance state. Since antimicrobial resistance is inherently a local
phenomenon, variation in prevalence is expected (assay methods also differ, which
adds to variation). For example, surveys from Asia (excluding China) indicate a
prevalence for hVISA of a few percent [92], but a report from Taiwan places it at
about 10% (2012–2013; Ref. [91]). A hospital study from Turkey also reported a
high prevalence of hVISA in blood isolates (almost 14%, Ref. [93]). In a troubling
study from Michigan (cited in Ref. [94]), the prevalence of hVISA in blood isolates
increased from 2% (1987–1993) to 8% (2003–2007), and in a multicenter US study,
the prevalence increased from 0.4% in 2009 to 1.2% in 2011. Thus, hVISA dissemi-
nation within and between hospitals is taken seriously; it may require special con-
sideration by infection control departments to limit transmission from patients
colonized by or infected with hVISA.
hVISA surveillance data also suggest that treatment changes are needed.
Although VISA is associated with treatment failure, reports on hVISA are mixed.
For example, a Michigan study of MRSA bacteremia found that hVISA was not
clearly associated with treatment failure [90]. However, in another study, hVISA
284 K. Drlica et al.

correlated with a doubling in mortality from pneumonia, relative to vancomycin-­

susceptible S. aureus (treatment failure was also higher, but not significantly so
[95]). Although a clear statement about outcome from hVISA cannot be made,
hVISA remains important as a likely precursor to VISA.
Studies of hVISA are also contributing to a test for restricting the emergence of
resistance. One approach is to increase antimicrobial dose. We pointed out that
resistant mutants are selectively enriched when drug concentrations fall inside a
specific concentration range called the mutant selection window; keeping drug con-
centrations above that window should restrict amplification of mutant subpopula-
tions [3, 96]. The increasing problem of hVISA and VISA with serious infections
led to a medical commission concluding from pharmacokinetic studies that dosing
to generate vancomycin AUC/MIC >400 h should control most serious MRSA
infections [97]. It was argued that measuring this pharmacodynamic parameter was
not practical for routine clinical use, but as a surrogate goal, the recommendation
was to maintain a minimum serum concentration between 15 and 20 μg/ml [97].
The commission did not know the upper boundary of the selection window. Several
years later, that boundary was measured and found to be 19 μg/ml (over 400 MRSA
isolates were examined, Ref. [98]). Thus, the proposed vancomycin target level for
favorable clinical outcome (15–20 μg/ml) fit with the value needed to restrict the
emergence of new resistant mutants. Clinical studies, again in Michigan [99],
reported increases in the minimum serum concentration of vancomycin from 10 μg/
ml in 2002–2003 to 19.7 in 2010–2012. During that time, the prevalence of hVISA
dropped from 9.7% to 2.1%. This vancomycin work with heteroresistance is the first
example for convergence between efforts to achieve favorable patient outcome and
efforts to restrict the emergence of resistance.

9.5 Other Pathogens Displaying Heteroresistance

Consideration of heteroresistance with M. tuberculosis and S. aureus provides an

introduction to two important features. First, some resistance mutations, such as
gyrase-mediated resistance to fluoroquinolones, have little fitness cost and are read-
ily enriched; in contrast, high fitness cost, as seen with VISA, limits the evolution to
a state of intermediate vancomycin resistance, at least in nature. Second, it is
straightforward to develop a DNA-based diagnostic to query a limited number of
mutations associated with antibiotic resistance, as with M. tuberculosis; however,
design is difficult when numerous mutations are associated with drug resistance, as
with VISA. Applying these ideas to heteroresistance with other pathogens has not
been done, since much less is known. Nevertheless, heteroresistance is clearly a
widespread phenomenon (Table 9.2). Below we list recent work that establishes the
potential importance of heteroresistance.
9 Heteroresistance: A Harbinger of Future Resistance 285

Table 9.2 Selected examples of heteroresistance

Pathogen species Antimicrobial Prevalence Locality Reference
Acinetobacter baumannii Carbapenem Nsa Greece [7, 100]
Acinetobacter baumannii Cephalosporin, Case study Taiwan [101]
Penicillins
Acinetobacter baumannii Colistin Case S. Korea, [102,
study; Argentina 103]
Candida glabrata Fluconazole 58% Israel [11]
Clostridium difficile Metronidazole 29% Spain [104]
Corynebacterium Daptomycin Case study USA [105]
striatum
Escherichia coli Cefepime 22% China [106]
Escherichia coli Carbapenem 34% China [107]
Haemophilus influenzae Imipenem 37% Switzerland [108]
Helicobacter pylori Severalb 48% Tunisia [109]
Klebsiella pneumoniae Carbapenem Ns Spain, Greece [110,
111]
Klebsiella pneumoniae Colistin 75% Greece [112]
Mycobacterium Fluoroquinolone 23% China [113]
tuberculosis
Pseudomonas aeruginosa Carbapenem 24; 19% Greece, China [114,
115]
Salmonella enterica Colistin Laboratory Nad [116]
Staphylococcus aureus Ceftaroline 21% USA [78]
Staphylococcus aureus Vancomycin-­ 10% Taiwan [91]
intermediate
Streptococcus Penicillin 44% Multinational [117]
pneumoniae
a
Ns no surveillance
b
Multiple infection
c
Not applicable

9.5.1 Gram-Negative Bacteria

Acinetobacter baumannii has become an important source of opportunistic nosoco-

mial infection, largely due to widespread multidrug resistance. Indeed, isolates have
been reported that are resistant to all commonly used antimicrobials. Heteroresistance
in A. baumannii is well known, having been observed in carbapenem Etest analyses
more than a decade ago [7]. Individual carbapenems may differ in the genes involved
in resistance, since for one carbapenem (meropenem), heteroresistance persists dur-
ing subculturing on drug-free agar, while that stability is not seen with another
(imipenem) [100]. A. baumannii also displays heteroresistance to cephalosporins
and penicillins [101]. Population analysis profiles for these β-lactams can be com-
plex, as illustrated by a report in which PAP showed colony numbers dropping at
low concentrations of cefepime and climbing at high concentrations [101]. This
phenomenon is not yet understood.
286 K. Drlica et al.

Heteroresistance to colistin, an agent of last resort, is also seen among isolates of

A. baumannii [102, 103]. In a survey performed at an Argentine hospital, heterore-
sistance doubled (46–95%) from 2004 to 2012, a period in which colistin consump-
tion increased by more than fourfold [103]. Colistin resistance in Argentina tends to
be unstable, and the increase in heteroresistance did not presage an increase in resis-
tance [103]. Nevertheless, the widespread occurrence of heteroresistance with A.
baumannii does not bode well for antimicrobial success with this pathogen.
Escherichia coli is a common inhabitant of the human digestive tract that is
becoming a serious urinary pathogen as multidrug-resistant forms become more
prevalent. In a study that examined more than 300 isolates for cephalosporin
(cefepime) heteroresistance, almost a quarter displayed colony growth inside the
zone of inhibition on agar plates [105]. In two-thirds of the cases, the patients had
received prior treatment with a cephalosporin. These observations are consistent with
a model in which antimicrobial pressure enriches mutant subpopulations. E. coli also
causes septicemia, and invasive E. coli has exhibited clonally diverse, carbapenem
heteroresistance [107]. In one case, examination of consecutive samples from the
same patient showed a gradual shift of the E. coli subpopulation profile (PAP) to
higher carbapenem concentrations and eventually to complete resistance. Such data
establish heteroresistance as an intermediate step along the evolutionary climb toward
complete carbapenem resistance, at least for E. coli. To our knowledge, the contribu-
tion of plasmid-mediated resistance, which is common, has not been addressed.
Haemophilus influenzae is an opportunistic pathogen that colonizes the human
airway. Resistance to β-lactams is commonly due to plasmid-mediated β-lactamases
and altered penicillin-binding protein-3 [108]. While imipenem resistance is rare,
heteroresistant H. influenzae isolates have been described [108]. In one report, PAP
revealed heteroresistance in 46/124 isolates that had an intermediate Etest MIC. With
H. influenzae, β-lactam heteroresistance arises from multiple genetic and biochemi-
cal factors, which will make DNA testing a challenge.
Helicobacter pylori causes a chronic infection of the human gastric mucosa that
is thought to be central to peptic ulcer disease, chronic gastritis, and gastric cancer.
Extensive use of antimicrobials has led to loss of antimicrobial susceptibility among
isolates of H. pylori. Clinical testing of gastric biopsies is complicated by the het-
erogeneous distribution of H. pylori in the stomach. In a survey of 66 patients in
which isolates were obtained from two distinct gastric regions, 15% exhibited infec-
tion of clonal origin in which the isolate from one compartment was susceptible to
the antibiotics tested, while the sample from the other compartment was resistant to
at least one of four agents (clarithromycin, metronidazole, levofloxacin, and
­rifabutin) [118]. In this situation, simply labeling an infection as heteroresistant
would have obscured the compartmentalization associated with H. pylori.
Since transmission of H. pylori is common and since infection persists for long
times, heteroresistant infections may arise from multiple superinfection. The frequency
of multiple infection may be less common in industrialized countries, as indicated by
a comparison of isolates from university hospitals in France and Tunisia [109]. For 21
isolates examined from each country, multiple infection was observed 10-times more
often with Tunisian patients than with French ones (clonal heteroresistance was similar
9 Heteroresistance: A Harbinger of Future Resistance 287

for the two countries). While the reasons for differences in heteroresistance are com-
plex, these data show that clinicians in developing countries should be watchful for
multiple infections that might impact susceptibility testing.
Klebsiella pneumoniae causes serious diseases, such as pneumonia, meningitis,
and urinary infections. Since K. pneumoniae inhabits the human digestive tract, it
readily disseminates in hospitals by fecal contamination. Thus, when multidrug-­
resistant K. pneumoniae strains develop resistance to carbapenems, they become a
major nosocomial problem. Low reproducibility of MIC tests for carbapenems, fol-
lowed by population analysis profiling, led to the conclusion that K. pneumoniae
heteroresistance is overlooked by automated susceptibility testing [110].
Heteroresistance appears to arise from drug-induced expression of carbapenemases,
since heteroresistance to meropenem is lost when drug pressure is withdrawn [111].
As the prevalence of resistance to the major antimicrobials mounts, colistin is
being used to treat K. pneumoniae infections. The result has been a sharp increase
in colistin resistance. For example, in one Greek hospital, resistance to colistin rose
from 0% in 2007, to 8% in 2008, and 24% in 2009 [112]. When PAP was performed
on a small set of patient isolates, heteroresistance to colistin was observed in 12/16
isolates that had been deemed susceptible by standard MIC assays [112]. With K.
pneumoniae, colistin heteroresistance is associated with the PhoPQ regulatory sys-
tem [119], as pointed out below for E. cloacae. The PhoPQ system alters the lipo-
polysaccharide of cell surfaces (the negative charge on lipid A is reduced, thereby
lowering the affinity for colistin, a cationic peptide). Colistin monotherapy is con-
traindicated for serious disease caused by K. pneumoniae.
Pseudomonas aeruginosa is an opportunistic pathogen that is particularly prob-
lematic for patients suffering from cystic fibrosis. Antimicrobial resistance with P.
aeruginosa is mediated by multiple efflux systems and production of β-lactamases.
In a study from Greece, 27% of presumably susceptible isolates exhibited stable
carbapenem heteroresistance [114]. This result may be common for P. aeruginosa,
as a similar finding was reported from China [115]. With P. aeruginosa, it may be
necessary to perform heteroresistance testing on many isolates, since automated
methods do not reliably detect heteroresistance.
Salmonella enterica serovar Typhimurium is noted for causing outbreaks of food
poisoning. Since isolates that exhibit multidrug resistance are associated with
increased mortality and morbidity, colistin is being considered for treatment of S.
enterica-associated diseases. A study of laboratory-generated colistin heteroresis-
tance with S. enterica revealed a correlation between heteroresistance and a
­moderate gene dosage of pmrD, a gene that upregulates proteins that modify lipid A
and thereby lower susceptibility to colistin [116]. Successive passages in the pres-
ence of colistin increased amplification of pmrD, while the number of amplified
copies declined when cells were passaged on drug-free medium. A similar phenom-
enon may have contributed to tetracycline heteroresistance in a clinical isolate
[116]. Antimicrobial resistance arising from gene amplification has also been
observed with M. tuberculosis [120], suggesting that it may underlie heteroresis-
tance in a variety of pathogens.
288 K. Drlica et al.

9.5.2 Gram-Positive Bacteria

Clostridium difficile causes serious diarrhea, especially in nosocomial settings

where antibiotic resistance plays an important role in driving outbreaks. C. difficile
is an anaerobic pathogen that is frequently treated with metronidazole. In a Spanish
study [104], almost 30% of C. difficile samples showed metronidazole heteroresis-
tance when examined for colony formation within inhibition zones on agar plates.
Thus, a major treatment option for this opportunistic pathogen is threatened by
resistance.
Corynebacterium striatum is a commensal skin inhabitant that occasionally
causes infection. A case was reported [105] in which a patient with a C. striatum
infection, treated with daptomycin, developed endocarditis. C. striatum was recov-
ered from the patient, and after plating for an Etest, colonies formed within the zone
of inhibition. Bacteria from those colonies, when purified and retested, had very
high MICs for daptomycin, while the bulk of the culture was daptomycin suscepti-
ble. These data show that daptomycin is subject to heteroresistance issues.
Enterobacter cloacae is a nosocomial pathogen that causes a wide range of infec-
tions, largely in the very young and the elderly. The pathogen is readily distributed
within hospitals on medical devices and via hospital workers. Due to multidrug resis-
tance, colistin is being used in the hospital setting. Colistin heteroresistance is readily
detected by colonies in the zone of inhibition during susceptibility testing on agar, but
examination of individual colonies shows that resistance is lost upon subculturing on
drug-free agar [121]. During infection of mice with the heteroresistant isolate, the
fraction of resistant cells increased even in the absence of colistin. This enrichment
was due to a portion of the innate immune response exerted by macrophages: hetero-
resistance rendered E. cloacae refractory to colistin if administered after infection
was established, but experimental depletion of macrophages maintained colistin sus-
ceptibility [121]. Thus, host functions can expand the effect of heteroresistance. Such
data emphasize that automated susceptibility testing can be misleading.
Transcriptional analysis revealed increased expression of PhoQ in the transiently
resistant strain of E. cloacae (for additional detail, see Chap. 15). PhoQ expression
leads to a modification of membrane lipid A, which then restricts the action of colis-
tin. How the innate immune system stimulates expression PhoQ is not yet known.
Streptococcus pneumoniae is responsible for roughly half of all pneumonia
cases. Since S. pneumoniae is commonly carried in the nasopharynx of young chil-
dren and since children are treated with many antibiotics, resistance is expected to
be a problem. Penicillin has been used extensively to treat infections caused by S.
pneumoniae, and penicillin heteroresistance has been reported [117]. In an effort to
expand the number of useful antibiotic agents for S. pneumoniae-related infections,
a Swiss study examined S. pneumoniae isolates for heteroresistance to fosfomycin
[122]. Even though fosphomycin is not currently used for treatment, 10 of 11 iso-
lates exhibited fosfomycin-heteroresistance. These data, which indicate that fosfo-
mycin resistance may emerge quickly, show that heteroresistance can be used as a
way discriminate against certain new antimicrobials.
9 Heteroresistance: A Harbinger of Future Resistance 289

9.5.3 Invasive Fungus

Candida glabrata is an important fungal pathogen that can be lethal to immuno-

compromised patients. Fluconazole, a common antifungal agent, readily enriches
stable heteroresistant strains of C. glabrata [11]. It is likely that many genes are
involved in heteroresistance, since population analysis profiles showed a wide dis-
tribution. As with bacterial pathogens, heteroresistance in C. glabrata is not readily
detected by standard drug susceptibility testing; consequently, some isolates may be
misclassified as susceptible. To assess the relevance of fluconazole heteroresistance,
mice were infected with C. glabrata and treated with fluconazole. Persistent infec-
tion was observed four times as often with a highly heteroresistant isolate. Thus,
heteroresistance in disease caused by C. glabrata is likely to be clinically
important.

9.6 Concluding Remarks

Efforts to control the expansion of resistance by reducing antimicrobial consump-

tion have met with mixed results (e.g., [123, 124]), and heteroresistance is becom-
ing widespread (Table 9.2). A preemptive attack on heteroresistance may slow the
emergence of resistance. In the case of tuberculosis, that entails identifying hetero-
resistant infections and then adjusting treatment protocols. In the case of MRSA, it
requires treating infections with higher vancomycin concentrations. With many
other pathogens, detection of heteroresistance needs to be improved (automated
susceptibility testing currently fails to detect heteroresistance); then treatment pro-
tocols need to be modified to block further mutant enrichment. A central problem is
that raising doses to suppress evolution to resistance is likely to increase toxic side
effects. Thus, strategies that may be good for the community as a whole may be
harmful to some individual patients. A long-term solution requires more research
focus on chemical adjuvants that will increase antimicrobial lethality to allow non-
toxic, anti-mutant dosing.

Major Points
• Antimicrobial heteroresistance derives from a variety of phenomena ranging
from subpopulations of stable, fully resistant mutants to reversible, antimicrobial-­
mediated induction or amplification of protective genes.
• Heteroresistance is common: it has been observed in many different pathogenic
bacterial species and found in almost 25% of patient isolates
• Heteroresistance can evolve to full drug resistance.
• The importance of heteroresistance has been underappreciated, because infec-
tions containing heteroresistant pathogen populations can often be treated
successfully.
290 K. Drlica et al.

• Detection provides an opportunity to adjust antimicrobial treatment to slow the

evolution of heteroresistant populations into populations exhibiting complete
drug resistance.
• DNA-based methods can be used to detect heteroresistance when specific genetic
alterations are known to be responsible for reduced susceptibility; methods
developed for cancer diagnostics may apply to detection of M. tuberculosis
heteroresistance.

Acknowledgments We thank the following for helpful discussions and critical comments:
Veronique Dartois, Dorothy Fallows, Marila Gennaro, Ben Gold, Barry Kreiswirth, Richard Pine,
and George Zhanel. The authors’ work was supported by NIH grants 1DP20D007423,
1R01AI073491, 1R21A03781, and 1R01AI87671.

References

1. Bush K, Fisher J. Epidemiological expansion, structural studies, and clinical challenges of

new β-lactamases from gram-negative bacteria. Annu Rev Microbiol. 2011;65:455–78.
2. Chambers H, Deleo F. Waves of resistance: Staphylococcus aureus in the antibiotic era. Nat
Rev Microbiol. 2009;7:629–41.
3. Drlica K, Zhao X. Mutant selection window hypothesis updated. Clin Inf Dis. 2007;44:681–8.
4. El-Halfawy O, Valvano M. Antimicrobial heteroresistance: an emerging field in need of clar-
ity. Clin Microbiol Rev. 2015;28:191–207.
5. Liu C, Chambers H. Staphylococcus aureus with heterogeneous resistance to vancomy-
cin: epidemiology, clinical significance, and critical assessment of diagnostic methods.
Antimicrob Agents Chemother. 2003;47:3040–5.
6. Hiramatsu K, Cui L, Kuroda M, Ito T. The emergence and evolution of methicillin-resistant
Staphylococcus aureus. Trends Microbiol. 2001;9:486–93.
7. Pournaras S, Ikonomidis A, Markogiannakis A, Maniatis A, Tsakris A. Heteroresistance to
carbapenems in Acinetobacter baumannii. J Antimicrob Chemother. 2005;55(6):1055.
8. deLencastre H, SáFigueiredo A, Urban C, Rahal J, Tomasz A. Multiple mechanisms of methi-
cillin resistance and improved methods for detection in clinical isolates of Staphylococcus
aureus. Antimicrob Agents Chemother. 1991;35:632–9.
9. Dong Y, Zhao X, Domagala J, Drlica K. Effect of fluoroquinolone concentration on selection
of resistant mutants of Mycobacterium bovis BCG and Staphylococcus aureus. Antimicrob
Agents Chemother. 1999;43:1756–8.
10. Dong Y, Zhao X, Kreiswirth B, Drlica K. Mutant prevention concentration as a measure of
antibiotic potency: studies with clinical isolates of Mycobacterium tuberculosis. Antimicrob
Agents Chemother. 2000;44:2581–4.
11. Ben-Ami R, Zimmerman O, Finn T, Amit S, Novikov A, Wertheimer N, et al. Heteroresistance
to fluconazole is a continuously distributed phenotype among Candida glabrata clinical
strains associated with in vivo persistence. MBio. 2016;7:pii: e00655–16.
12. Malik M, Hoatam G, Chavda K, Kerns R, Drlica K. Novel approach for comparing qui-
nolones for emergence of resistant mutants during quinolone exposure. Antimicrob Agents
Chemother. 2010;54:149–56.
13. Levin-Reisman I, Ronin I, Gefen O, Braniss I, Shoresh N, Balaban N. Antibiotic tolerance
facilitates the evolution of resistance. Science. 2017;355:826–30.
14. Nicolas-Chanoine M, Bertrand X, Madec J. Escherichia coli ST131, an intriguing clonal
group. Clin Microbiol Rev. 2014;27:543–74.
9 Heteroresistance: A Harbinger of Future Resistance 291

15. East African Hospitals and British Medical Research Council. Comparative trial of isoniazid
alone in low and high dosage and isoniazid plus PAS in the treatment of acute pulmonary
tuberculosis in East Africa. Tubercle. 1960;40:83–102.
16. Crofton J, Mitchison D. Streptomycin resistance in pulmonary tuberculosis. Br Medical
J. 1948;2:1009–15.
17. Canetti G, leLerzin M, Porven G, Rist N, Grumbach F. Some comparative apects of rifampi-
cin and isoniazid. Tubercle. 1968;49:367–76.
18. Mitchison DA. Drug resistance in mycobacteria. Brit. Med Bull. 1984;40:84–90.
19. Canetti G. The J. Burns Amberson Lecture: present aspects of bacterial resistance in tubercu-
losis. Am Rev Resp Dis. 1965;92:687–703.
20. Lorian V, Finland M. In vitro effect of rifampicin on mycobacteria. App Microbiol.
1969;17:202–7.
21. Hobby G, Lenert T, Mater-Engallena J. In vitro activity of rifampin against the H37Rv strain
of Mycobacterium tuberculosis. Am Rev Respir Dis. 1968;99:453–6.
22. Canetti G, Froman S, Grosset J, Hauduroy P, Langerová M, Mahler H, et al. Mycobacteria:
laboratory methods for testing drug sensitivity and resistance. Bull World Health Organ.
1963;29:564–78.
23. McGrath M, GeyvanPittius N, vanHelden P, Warren R, Warner D. Mutation rate and the
emergence of drug resistance in Mycobacterium tuberculosis. J Antimicrob Chemother.
2014;69:292–302.
24. Warner D, Ndwandwe D, Abrahams G, Kana B, Machowski E, Venclovas C, et al. Essential
roles for imuA’- and imuB-encoded accessory factors in DnaE2-dependent mutagenesis in
Mycobacterium tuberculosis. Proc Natl Acad Sci U S A. 2010;107:13093–8.
25. Fox W, Elklard G, Mitchison D. Studies on the treatment of tuberculosis undertaken by the
British Medical Research Council Tuberculosis Units, 1946–1986, with relevant subsequent
publications. Int J Tuberc Lung Dis. 1999;3:S231–S79.
26. Wayne LG, Hayes LG. An in vitro model for sequential study of shiftdown of Mycobacterium
tuberculosis through two stages of nonreplicating persistence. Infect Immun. 1996;64:2062–9.
27. Kaplan G, Post F, Moreira A, Wainwright H, Kreiswirth B, Tanverdi M, et al. Mycobacterium
tuberculosis growth at the cavity surface: a microenvironment with failed immunity. Infect
Immun. 2003;71:7099–108.
28. Prideaux B, Via L, Zimmerman M, Eum S, Sarathy J, O'Brien P, et al. The association between
sterilizing activity and drug distribution into tuberculosis lesions. Nat Med. 2015;21:1223–7.
29. Zetola N, Modongo C, Moonan P, Ncube R, Matlhagela K, Sepako E, et al. Clinical out-
comes among persons with pulmonary tuberculosis caused by Mycobacterium tuberculosis
isolates with phenotypic heterogeneity in results of drug-susceptibility tests. J Infect Dis.
2014;209:1754–63.
30. Cullen M, Sam N, Kanduma E, McHugh T, Gillespie S. Direct detection of heteroresistance in
Mycobacterium tuberculosis using molecular techniques. J Med Microbiol. 2006;55(8):1157.
31. Plinke C, Cox H, Kalon S, Doshetov D, Rüsch-Gerdes S, Niemann S. Tuberculosis etham-
butol resistance: concordance between phenotypic and genotypic test results. Tuberculosis.
2009;89:448–52.
32. Post F, Willcox P, Mathema B, Steyn L, Shean K, Ramaswamy S, et al. Genetic polymor-
phism in Mycobacterium tuberculosis isolates from patients with chronic multidrug-resistant
tuberculosis. J Inf Dis. 2004;190:99–106.
33. Rinder H, Mieskes K, Loscher T. Heteroresistance in Mycobacterium tuberculosis. Int
J Tuberc Lung Dis. 2001;5:339–45.
34. Hofmann-Thiel S, van Ingen J, Feldmann K, Turaev L, Uzakova G, Murmusaeva G, et al.
Mechanisms of heteroresistance to isoniazid and rifampin of Mycobacterium tuberculosis in
Tashkent, Uzbekistan. Eur Respir J. 2009;33:368–74.
35. Eilertson B, Maruri F, Blackman A, Herrera M, Samuels D, Sterling T. High proportion of
heteroresistance in gyrA and gyrB in fluoroquinolone-resistant Mycobacterium tuberculosis
clinical isolates. Antimicrob Agents Chemother. 2014;58:3270–5.
292 K. Drlica et al.

36. Zhang X, Zhao B, Huang H, Zhu Y, Peng J, Dai G, et al. Co-occurrence of amikacin-­resistant
and -susceptible Mycobacterium tuberculosis isolates in clinical samples from Beijing,
China. J Antimicrob Chemother. 2013;68:1537–42.
37. Shamputa I, Rigouts L, Eyongeta L, El Aila N, van Deun A, Salim A, et al. Genotypic and
phenotypic heterogeneity among Mycobacterium tuberculosis isolates from pulmonary
tuberculosis patients. J Clin Microbiol. 2004;42:5528–36.
38. van Rie A, Victor T, Richardson M, Johnson R, vanderSpuy G, Murray E, et al. Reinfection
and mixed infection cause changing Mycobacterium tuberculosis drug-resistance patterns.
Am J Respir Crit Care Med. 2005;172:636–42.
39. Blaas S, Mütterlein R, Weig J, Neher A, Salzberger B, Lehn N, et al. Extensively drug resis-
tant tuberculosis in a high income country: a report of four unrelated cases. BMC Infect Dis.
2008;8:60.
40. Warren R, Victor T, Streicher E, Richardson M, Beyers N, GeyvanPittius N, et al. Patients
with active tuberculosis often have different strains in the same sputum specimen. Am J Crit
Care Med. 2004;169:610–4.
41. Meacci F, Orrù G, Iona E, Giannoni F, Piersimoni C, Pozzi G, et al. Drug resistance evolu-
tion of a Mycobacterium tuberculosis strain from a noncompliant patient. J Clin Microbiol.
2005;43:3114–20.
42. Xu C, Kreiswirth B, Sreevatsan S, Musser J, Drlica K. Fluoroquinolone resistance associ-
ated with specific gyrase mutations in clinical isolates of multidrug-resistant Mycobacterium
tuberculosis. J. Inf. Dis. 1996;174:1127–30.
43. Shamputa I, Jugheli L, Sadradze N, Willery E, Portaels F, Supply P, et al. Mixed infection and
clonal representativeness of a single sputum sample in tuberculosis patients from a peniten-
tiary hospital in Georgia. Respir Res. 2006;7:99.
44. Vadwai V, Daver G, Udwadia Z, Sadani M, Shetty A, Rodrigues C. Clonal population
of Mycobacterium tuberculosis strains reside within multiple lung cavities. PLoS One.
2011;6:e24770.
45. Lieberman T, Wilson D, Misra R, Xiong L, Moodley P, Cohen T, et al. Genomic diversity in
autopsy samples reveals within-host dissemination of HIV-associated Mycobacterium tuber-
culosis. Nat Med. 2016;22:1470–4.
46. Behr M. Tuberculosis due to multiple strains: a concern for the patient? A concern for tuber-
culosis control? Am J Respir Crit Care Med. 2004;169(5):554.
47. Mankiewicz E, Liivak M. Phage types of Mycobacterium tuberculosis in cultures isolated
from Eskimo patients. Am Rev Respir Dis. 1975;111:307–12.
48. Bates J, Stead W, Rado T. Phage type of tubercle bacilli isolated from patients with two or
more sites of organ involvement. Am Rev Respir Dis. 1976;114:353–8.
49. Parsons L, Salfinger M, Clobridge A, Dormandy J, Mirabello L, Polletta V, et al. Phenotypic
and molecular characterization of Mycobacterium tuberculosis isolates resistant to both iso-
niazid and ethambutol. Antimicrob Agents Chemother. 2005;49:2218–25.
50. Pholwat S, Stroup S, Foongladda S, Houpt E. Digital PCR to detect and quantify heteroresis-
tance in drug resistant Mycobacterium tuberculosis. PLoS One. 2013;8:e57238.
51. Folkvardsen D, Svensson E, Thomsen V, Rasmussen E, Bang D, Werngren J, et al. Can
molecular methods detect 1% isoniazid resistance in Mycobacterium tuberculosis? J Clin
Microbiol. 2013a;51:1596–9.
52. Folkvardsen D, Thomsen V, Rigouts L, Rasmussen E, Bang D, Bernaerts G, et al. Rifampin
heteroresistance in Mycobacterium tuberculosis cultures as detected by phenotypic and geno-
typic drug susceptibility test methods. J Clin Microbiol. 2013b;51:4220–2.
53. Hu S, Li G, Li H, Liu X, Niu J, Quan S, et al. Rapid detection of isoniazid resistance in
Mycobacterium tuberculosis isolates by use of real-time-PCR-based melting curve analysis.
J Clin Microbiol. 2014;52:1644–52.
54. Chakravorty S, Kothari H, Aladegbami B, Cho E, Lee J, Roh S, et al. Rapid, high-throughput
detection of rifampin resistance and heteroresistance in Mycobacterium tuberculosis by use of
sloppy molecular beacon melting temperature coding. J Clin Microbiol. 2012;50:2194–202.
9 Heteroresistance: A Harbinger of Future Resistance 293

55. Chakravorty S, Aladegbami B, Thoms K, Lee J, Lee E, Rajan V, et al. Rapid detection of
fluoroquinolone-resistant and heteroresistant Mycobacterium tuberculosis by use of sloppy
molecular beacons and dual melting-temperature codes in a real-time PCR assay. J Clin
Microbiol. 2011;49:932–40.
56. de Oliveira M, da Silva-Rocha A, Cardoso-Oelemann M, Gomes H, Fonseca L, Werneck-­
Barreto A, et al. Rapid detection of resistance against rifampicin in isolates of Mycobacterium
tuberculosis from Brazilian patients using a reverse-phase hybridization assay. J Microbiol
Methods. 2003;53:335–42.
57. Vogelstein B, Kinsler K. Digital PCR. Proc Natl Acad Sci U S A. 1999;96:9236–41.
58. Yajko D, Wagner C, Tevere V, Kocagöz T, Hadley W, Chambers H. Quantitative culture of
Mycobacterium tuberculosis from clinical sputum specimens and dilution endpoint of its
detection by the Amplicor PCR assay. J Clin Microbiol. 1995;33:1944–7.
59. Diacon A, Patientia R, Venter A, van Helden P, Smith P, McIlleron H, et al. Early bactericidal
activity of high-dose rifampin in patients with pulmonary tuberculosis evidenced by positive
sputum smears. Antimicrob Agents Chemother. 2007;51:2994–6.
60. Brindle R, Odhiambo J, Mitchison D. Serial counts of Mycobacterium tuberculosis in sputum
as surrogate markers of the sterilising activity of rifampicin and pyrazinamide in treating
pulmonary tuberculosis. BMC Pulm Med. 2001;1:2.
61. Bouakaze C, Keyser C, Gonzalez A, Sougakoff W, Veziris N, Dabernat H, et al. Matrix-assisted
laser desorption ionization-time of flight mass spectrometry-based single nucleotide polymor-
phism genotyping assay using iPLEX gold technology for identification of Mycobacterium
tuberculosis complex species and lineages. J Clin Microbiol. 2011;49(9):3292.
62. Liu Q, Sommer S. Pyrophosphorolysis-activated polymerization (PAP): application to allele-­
specific amplification. BioTechniques. 2000;29:1072–83.
63. Liu Q, Nguyen VQ, Li X, Sommer SS. Multiplex dosage pyrophosphorolysis-activated
polymerization: application to the detection of heterozygous deletions. BioTechniques.
2006;40(5):661–8.
64. Liu Q, Sommer SS. Pyrophosphorolysis-activatable oligonucleotides may facilitate detection
of rare alleles, mutation scanning and analysis of chromatin structures. Nucleic Acids Res.
2002;30(2):598–604.
65. Liu Q, Sommer SS. PAP: detection of ultra rare mutations depends on P* oligonucle-
otides: “sleeping beauties” awakened by the kiss of pyrophosphorolysis. Hum Mutat.
2004a;23(5):426–36.
66. Liu Q, Sommer SS. Detection of extremely rare alleles by bidirectional pyrophosphorolysis-­
activated polymerization allele-specific amplification (Bi-PAP-A): measurement of mutation
load in mammalian tissues. BioTechniques. 2004b;36(1):156–66.
67. Liu Q, Sommer SS. Pyrophosphorolysis by Type II DNA polymerases: implications for
pyrophosphorolysis-­activated polymerization. Anal Biochem. 2004c;324(1):22–8.
68. Vargas D, Kramer F, Tyagi S, Marras S. Multiplex real-time PCR assays that measure the
abundance of extremely rare mutations associated with cancer. PLoS One. 2016;11:e0156546.
69. Gootenberg J, Abudayyeh O, Lee J, Essletzbichler P, Dy A, Joung J, et al. Nucleic acid detec-
tion with CRISPR-Cas13a/C2c2. Science. 2017;pii: eaam9321.
70. Rock J, Hopkins F, Chavez A, Diallo M, Chase M, Gerrick E, et al. Programmable transcrip-
tional repression in mycobacteria using an orthogonal CRISPR interference platform. Nat
Microbiol. 2017;2:16274.
71. Rice J, Reis A, Rice L, Carver-Brown R, Wangh L. Fluorescent signatures for variable DNA
sequences. Nucleic Acids Res. 2012;40:e164.
72. Huang Q, Liu Z, Liao Y, Chen X, Zhang Y, Li Q. Multiplex fluorescence melting curve analysis
for mutation detection with dual-labeled, self-quenched probes. PLoS One. 2011;6:e19206.
73. Seligman S. Methicillin-resistant staphylococci: genetics of the minority population. J Gen
Microbiol. 1966;42:315–22.
74. Ryffel C, Strässle A, Kayser F, Berger-Bächi B. Mechanisms of heteroresistance in
methicillin-­resistant Staphylococcus aureus. Antimicrob Agents Chemother. 1994;38:724–8.
294 K. Drlica et al.

75. Lim D, Strynadka N. Structural basis for the beta lactam resistance of PBP2a from methicillin-­
resistant Staphylococcus aureus. Nat Struct Biol. 2002;9:870–6.
76. Aiba Y, Katayama Y, Hishinuma T, Murakami-Kuroda H, Cui L, Hiramatsu K. Mutation
of RNA polymerase β-subunit gene promotes heterogeneous-to-homogeneous conversion
of β-lactam resistance in methicillin-resistant Staphylococcus aureus. Antimicrob Agents
Chemother. 2013;57:4861–71.
77. Alam M, Petit R, Crispell E, Thornton T, Conneely K, Jiang Y, et al. Dissecting vancomycin-­
intermediate resistance in Staphylococcus aureus using genome-wide association. Genome
Biol Evol. 2014;6:1174–85.
78. Saravolatz SN, Martin H, Pawlak J, Johnson LB, Saravolatz LD. Ceftaroline-heteroresistant
Staphylococcus aureus. Antimicrob Agents Chemother. 2014;58(6):3133.
79. Cui L, Neoh H, Shoji M, Hiramatsu K. Contribution of vraSR and graSR point mutations
to vancomycin resistance in vancomycin-intermediate Staphylococcus aureus. Antimicrob
Agents Chemother. 2009;53:1231–4.
80. Boyle-Vavra S, Berke S, Lee J, Daum R. Reversion of the glycopeptide resistance phenotype
in Staphylococcus aureus clinical isolates. Antimicrob Agents Chemother. 2000;44:272–7.
81. Cameron D, Ward D, Kostoulias X, Howden B, Moellering R, Eliopoulos G, et al. Serine/
threonine phosphatase Stp1 contributes to reduced susceptibility to vancomycin and viru-
lence in Staphylococcus aureus. J Inf Dis. 2012;205:1677–87.
82. Gao W, Cameron D, Davies J, Kostoulias X, Stepnell J, Tuck K, et al. The RpoB H481Y
rifampicin resistance mutation and an active stringent response reduce virulence and increase
resistance to innate immune responses in Staphylococcus aureus. J Inf Dis. 2013;207:929–39.
83. Majcherczyk P, Barblan J, Moreillon P, Entenza J. Development of glycopeptide-­intermediate
resistance by Staphylococcus aureus leads to attenuated infectivity in a rat model of endocar-
ditis. Microb Pathog. 2008;45:408–14.
84. Peleg A, Monga D, Pillai S, Mylonakis E, Moellering R, Eliopoulos G. Reduced susceptibil-
ity to vancomycin influences pathogenicity in Staphylococcus aureus infection. J Inf Dis.
2009;199:532–6.
85. Donnio P, Oliveira D, Faria N, Wilhelm N, LeCoustumier A, deLencastre H. Partial exci-
sion of the chromosomal cassette containing the methicillin resistance determinant results in
methicillin-susceptible Staphylococcus aureus. J Clin Microbiol. 2005;43:4191–3.
86. Tenover F, Moellering R. The rationale for revising the Clinical and Laboratory Standards
Institute vancomycin minimal inhibitory concentration interpretive criteria for Staphylococcus
aureus. Clin Inf Dis. 2007;44:1208–15.
87. Satola S, Farley M, Anderson K, Patel J. Comparison of detection methods for heteroresis-
tant vancomycin-intermediate Staphylococcus aureus, with the population analysis profile
method as the reference method. J Clin Microbiol. 2011;49:177–83.
88. Chen C, Huang Y, Chiu C. Multiple pathways of cross-resistance to glycopeptides and dapto-
mycin in persistent MRSA bacteraemia. J Antimicrob Chemother. 2015;70:2965–72.
89. Hafer C, Lin Y, Kornblum J, Lowy F, Uhlemann A. Contribution of selected gene muta-
tions to resistance in clinical isolates of vancomycin-intermediate Staphylococcus aureus.
Antimicrob Agents Chemother. 2012;56:5845–51.
90. Khatib R, Jose J, Musta A, Sharma M, Fakih M, Johnson L, et al. Relevance of vancomycin-­
intermediate susceptibility and heteroresistance in methicillin-resistant Staphylococcus
aureus bacteraemia. J Antimicrob Chemother. 2011;66:1594–9.
91. Huang S, Chen Y, Chuang Y, Chiu S, Fung C, Lu P, et al. Prevalence of vancomycin-­
intermediate Staphylococcus aureus (VISA) and heterogeneous VISA among methicillin-­
resistant S. aureus with high vancomycin minimal inhibitory concentrations in Taiwan: a
multicenter surveillance study, 2012–2013. J Microbiol Immunol Infect. 2016;49:701–7.
92. Chung D, Lee C, Kang Y, Baek J, Kim S, Ha Y, et al. Genotype-specific prevalence of
heterogeneous vancomycin-intermediate Staphylococcus aureus in Asian countries. Int
J Antimicrob Agents. 2015;46:338–41.
9 Heteroresistance: A Harbinger of Future Resistance 295

93. Sancak B, Yagci S, Gür D, Gülay Z, Ogunc D, Söyletir G, et al. Vancomycin and dapto-
mycin minimum inhibitory concentration distribution and occurrence of heteroresistance
among methicillin-resistant Staphylococcus aureus blood isolates in Turkey. BMC Infect Dis.
2013;13:583.
94. Gomes D, Ward K, LaPlante K. Clinical implications of vancomycin heteroresistant and
intermediately susceptible Staphylococcus aureus. Pharmacotherapy. 2015;35:424–32.
95. Claeys K, Lagnf A, Hallesy J, Compton MT, Gravelin A, Davis S, et al. Pneumonia caused
by methicillin-resistant Staphylococcus aureus: does vancomycin heteroresistance matter?
Antimicrob Agents Chemother. 2016;60:1708–16.
96. Zhao X, Drlica K. Restricting the selection of antibiotic-resistant mutants: a general strategy
derived from fluoroquinolone studies. Clin Inf Dis. 2001;33(Suppl 3):S147–S56.
97. Rybak M, Lomaestro B, Rotschafer J, Moellering R, Craig W, Billeter M, et al. Therapeutic
monitoring of vancomycin in adults summary of consensus recommendations from the
American Society of Health-System Pharmacists, the Infectious Diseases Society of America,
and the Society of Infectious Diseases Pharmacists. Pharmacotherapy. 2009;29:1275–9.
98. Mei Q, Ye Y, Zhu Y, Cheng J, Yang HF, Liu Y, et al. Use of Monte Carlo simulation to evaluate
the development of vancomycin resistance in meticillin-resistant Staphylococcus aureus. Int
J Antimicrob Agents. 2015a;45(6):652.
99. Khatib R, Sharma M, Johnson L, Riederer K, Shemes S, Szpunar S. Decreasing prevalence of
isolates with vancomycin heteroresistance and vancomycin minimum inhibitory concentra-
tions ≥2 mg/L in methicillin-resistant Staphylococcus aureus over 11 years: potential impact
of vancomycin treatment guidelines. Diagn Microbiol Infect Dis. 2015;82:245–8.
100. Ikonomidis A, Neou E, Gogou V, Vrioni G, Tsakris A, Pournaras S. Heteroresistance
to meropenem in carbapenem-susceptible Acinetobacter baumannii. J Clin Microbiol.
2009;47:4055–9.
101. Hung K, Wang M, Huang A, Yan J, Wu J. Heteroresistance to cephalosporins and penicillins
in Acinetobacter baumannii. J Clin Microbiol. 2012;50:721–6.
102. Choi H, Kil M, Choi J, Kim S, Park K, Kim Y, et al. Characterisation of successive
Acinetobacter baumannii isolates from a deceased haemophagocytic lymphohistiocytosis
patient. Int J Antimicrob Agents. 2017;49:102–6.
103. Rodríguez C, Nastro M, Fiorilli G, Dabos L, Lopez-Calvo J, Fariña M, et al. Trends in the
resistance profiles of Acinetobacter baumannii endemic clones in a university hospital of
Argentina. J Chemother. 2016;28:25–7.
104. Álvarez-Pérez S, Blanco J, Harmanus C, Kuijper E, García M. Subtyping and antimicrobial
susceptibility of Clostridium difficile PCR ribotype 078/126 isolates of human and animal
origin. Vet Microbiol. 2017;199:15–22.
105. Tran T, Jaijakul S, Lewis C, Diaz L, Panesso D, Kaplan H, et al. Native valve endocar-
ditis caused by Corynebacterium striatum with heterogeneous high-level daptomycin
resistance: collateral damage from daptomycin therapy? Antimicrob Agents Chemother.
2012;56:3461–4.
106. Ma W, Sun J, Yang S, Zhang L. Epidemiological and clinical features for cefepime hetero-
resistant Escherichia coli infections in Southwest China. Eur J Clin Microbiol Infect Dis.
2016;35:571–8.
107. Sun J, Huang S, Yang S, Pu S, Zhang C, Zhang LZ. Impact of carbapenem heteroresistance
among clinical isolates of invasive Escherichia coli in Chongqing, southwestern China. Clin
Microbiol Infect. 2014;21:469.e1–10.
108. Cherkaoui A, Diene S, Renzoni A, Emonet S, Renzi G, François P, et al. Imipenem heterore-
sistance in nontypeable Haemophilus influenzae is linked to a combination of altered PBP3,
slow drug influx and direct efflux regulation. Clin Microbiol Infect. 2017;23:118.e9–e19.
109. Ben-Mansour K, Fendri C, Battikh H, Garnier M, Zribi M, Jlizi A, et al. Multiple and mixed
Helicobacter pylori infections: comparison of two epidemiological situations in Tunisia and
France. Infect Genet Evol. 2016;37:43–8.
296 K. Drlica et al.

110. Tato M, Morosini M, García L, Albertí S, Coque M, Cantón R. Carbapenem heteroresistance

in VIM-1-producing Klebsiella pneumoniae isolates belonging to the same clone: conse-
quences for routine susceptibility testing. J Clin Microbiol. 2010;48:4089–93.
111. Pournaras S, Kristo I, Vrioni G, Ikonomidis A, Poulou A, Petropoulou D, et al. Characteristics
of meropenem heteroresistance in Klebsiella pneumoniae carbapenemase (KPC)-producing
clinical isolates of K. pneumoniae. J Clin Microbiol. 2010;48:2601–4.
112. Meletis G, Tzampaz E, Sianou E, Tzavaras I, Sofianou D. Colistin heteroresistance in
carbapenemase-­producing Klebsiella pneumoniae. J Antimicrob Chemother. 2011;66:946–7.
113. Zhang X, Zhao B, Liu L, Zhu Y, Zhao Y, Jin Q. Subpopulation analysis of heteroresistance
to fluoroquinolone in Mycobacterium tuberculosis isolates from Beijing, China. J Clin
Microbiol. 2012;50:1471–3.
114. Pournaras S, Ikonomidis A, Markogiannakis A, Spanakis N, Maniatis A, Tsakris
A. Characterization of clinical isolates of Pseudomonas aeruginosa heterogeneously resis-
tant to carbapenems. J Med Microbiol. 2007;56:66–70.
115. Mei S, Gao Y, Zhu C, Dong C, Chen Y. Research of the heteroresistance of Pseudomonas
aeruginosa to imipenem. Int J Clin Exp Med. 2015b;8:6129–32.
116. Hjort K, Nicoloff H, Andersson D. Unstable tandem gene amplification generates heterore-
sistance (variation in resistance within a population) to colistin in Salmonella enterica. Mol
Microbiol. 2016;102:274–89.
117. Morand B, Mühlemann K. Heteroresistance to penicillin in Streptococcus pneumoniae. Proc
Natl Acad Sci U S A. 2007;104:14098–103.
118. Selgrad M, Tammer I, Langner C, Bornschein J, Meißle J, Kandulski A, et al. Different anti-
biotic susceptibility between antrum and corpus of the stomach, a possible reason for treat-
ment failure of Helicobacter pylori infection. World J Gastroenterol. 2014;20:16245–51.
119. Jayol A, Nordmann P, Brink A, Poirel L. Heteroresistance to colistin in Klebsiella pneu-
moniae associated with alterations in the PhoPQ regulatory system. Antimicrob Agents
Chemother. 2015;59:2780–4.
120. Kling A, Lukat P, Almeida DV, Bauer A, Fontaine E, Sordello S, et al. Antibiotics. Targeting
DnaN for tuberculosis therapy using novel griselimycins. Science. 2015;348:1106–12.
121. Band V, Crispell B, Napier B, Herrera C, Tharp G, Vavikolanu K, et al. Antibiotic failure
mediated by a resistant subpopulation in Enterobacter cloacae. Nat Microbiol. 2016;1:16053.
122. Engel H, Gutiérrez-Fernández J, Flückiger C, Martínez-Ripoll M, Mühlemann K, Hermoso
J, et al. Heteroresistance to fosfomycin is predominant in Streptococcus pneumoniae and
depends on the murA1 gene. Antimicrob Agents Chemother. 2013;57:2801–108.
123. Sedláková M, Urbánek K, Vojtová V, Suchánková H, Imwensi P, Kolář M. Antibiotic
consumption and its influence on the resistance in Enterobacteriaceae. BMC Res Notes.
2014;7:454.
124. Antoniadou A, Kanellakopoulou K, Kanellopoulou M, Polemis M, Koratzanis G,
Papademetriou E, et al. Impact of a hospital-wide antibiotic restriction policy program on the
resistance rates of nosocomial Gram-negative bacteria. Scand J Infect Dis. 2013;45:438–45.
 Part II
Biology of Resistance
Chapter 10
Epidemiology of Bacterial Resistance

Patricia A. Bradford

10.1 Introduction

Bacterial pathogens have developed resistance to antibacterial agents via multiple

routes. When any given pathogen mutates and becomes resistant, it can rapidly
result in immeasurable resistant daughter cells. Mutants that develop following
exposure to antibiotics favor mechanisms that confer resistance with the least cost
to fitness, that is, the strains that are least burdened by their resistance will survive.
This enhanced survival may also include increased virulence. Antimicrobial resis-
tance complicates the treatment for bacterial infections, resulting dosing with mul-
tiple antibiotics, prolonged courses of therapy, and excess hospitalizations. The
Centers for Disease Control and Prevention (CDC) published their first report on
antibiotic resistance in the USA in 2013, regarding the continued threat in the treat-
ment of bacterial infections [1]. In this report, the CDC estimated that at least two
million people acquired serious infections from antibiotic-resistant pathogens and
that at least 23,000 deaths in the USA could be attributed to infectious caused by
these organisms. It is important to understand not only the mechanisms by which
bacteria become resistant but also how resistance spreads from organism to organ-
ism and then from person to person. By understanding the epidemiology of resis-
tance, we can then learn how to address it with infection control and/or new
therapies. This chapter will examine the epidemiology of resistance by looking at
the mechanisms by which resistance spreads, examining the molecular methods
used for tracking resistance in bacterial pathogens, and reviewing some instances of
successful resistance dissemination within the hospital and in some populations of
people within the community.

P. A. Bradford (*)
Antimicrobial Development Specialists, LLC, Nyack, NY, USA
e-mail: [email protected]

© Springer International Publishing AG, part of Springer Nature 2018 299

I. W. Fong et al. (eds.), Antimicrobial Resistance in the 21st Century, Emerging Infectious
Diseases of the 21st Century, https://doi.org/10.1007/978-3-319-78538-7_10
300 P. A. Bradford

10.2 Development of Resistance

Bacterial resistance to antibiotics can result via three main pathways: modification
of the bacterial target for the antibacterial, decreased intracellular concentrations
due to reduced permeability and efflux, or enzymatic inactivation of the drug. In
some cases, all members of a given species might be resistant to a particular antibi-
otic. For example, all isolates of the Gram-negative non-fermenter Stenotrophomonas
maltophilia express a chromosomally encoded metallo-β-lactamase. Therefore,
resistance to imipenem and other carbapenems is a diagnostic tool for identifying
this organism. Alternatively, resistance can develop in previously susceptible organ-
isms through genetic mutation or by acquisition of foreign DNA encoding resis-
tance genes. The specific mechanisms that affect various classes of antibiotics are
discussed in other chapters. The discussion here will focus on the selection and
spread of resistance when it occurs.

10.2.1 Selection of Bacterial Pathogens with Innate Resistance

The use of antibacterial drugs disrupts the microbiome of the patient being treated.
In turn, the hospital unit or other groups of people in close proximity such as in
daycare centers or in long-term care facilities can be affected. As a consequence, an
entire species of bacterial pathogen might be selected with antibiotic pressure due
to natural resistance occurring in that species. For example, the increased role of
enterococci as opportunist pathogens in the 1980s and 1990s correlated with the
introduction and increased usage of fluoroquinolones and cephalosporins, as these
organisms are inherently resistant to those agents [2]. Similarly, the increasing inci-
dence of coagulase-negative staphylococci and α-hemolytic streptococci in hema-
tology patients, especially those who have indwelling central lines, correlated with
the increased use of fluoroquinolones in these patients [3]. Among Gram-negative
pathogens, Acinetobacter baumannii and S. maltophilia have become increasingly
prevalent in many intensive care units (ICUs) following the increased usage of car-
bapenems, especially among patients with mechanical ventilation [4, 5]. S. malto-
philia has a naturally occurring metallo-β-lactamase that renders it resistant to
carbapenems, and A. baumannii is often resistant to all antibacterials except
trimethoprim-­sulfamethoxazole. The introduction of each of these new therapies
has led to the unexpected consequence of shifting the etiology of some of the com-
mon hospital-based infections to species that are naturally more resistant than the
pathogens they replaced.
10 Epidemiology of Bacterial Resistance 301

10.2.2 Resistance by Mutation

As bacteria grow, the DNA (both the chromosome and plasmids) is replicated
through a process that is highly prone to errors in base incorporation. These errors,
leading to base substitutions, occur randomly, at a frequency of approximately 10−9
per gene [6]. Even one amino acid substitution can greatly alter the functionality of
a gene. For example, the substitution of serine for glycine at residue 238 in the
SHV-1 β-lactamase led to the first extended-spectrum β-lactamase (ESBL), SHV-2,
that conferred resistance to expanded spectrum cephalosporins [7]. In addition to
these random point mutations, replication errors may lead to deletions or insertions
of small pieces of genes. Each of these mutations may result in the altered interac-
tion of antibacterial agents with the bacterium through changes of the drug target,
enzymatic inactivation of the drug inactivation, or changes in the efflux or uptake
inactivated by the importation of insertion sequences, such as the case with the ccrA
gene expressing a metallo-β-lactamase in Bacteroides fragilis that is only expressed
only if an insertion sequence has inserted upstream of this structural gene [8].
Exposure to antibiotics does not cause the mutations but rather selects for strains
that have pre-existing mutations that allow the bacterial cell to survive in the pres-
ence of the antibiotic.
Most mutations occurring in the drug target or in an antibiotic-modifying enzyme
affect only a single antibacterial class. However, mutations also occur in genes
encoding outer membrane porin proteins that allow penetration through the outer
membrane by passive diffusion, or efflux systems that expel out of the cell multiple
antibiotic classes as well as other cell toxins such as dyes can greatly impact the
susceptibility of a bacterial cell to the antibiotic [9]. The maintenance of a mutation
in a bacterial pathogen causing antibiotic resistance is completely dependent upon
whether or not that mutation affects the fitness or virulence of that organism. If
resistant mutants emerge at high frequency and are still able to replicate and cause
disease, they can gain a foothold in the bacterial population that is further selected
through continued use of the drug [10]. There have been several antimicrobials
introduced in the 1980s and 1990s that had reduced utility following mutational
resistance in certain species. Resistance to fluoroquinolones among staphylococci
rapidly emerged by the upregulation of NorA-mediated efflux [11]. Another exam-
ple was the use of imipenem that led to the selection of P. aeruginosa that have lost
the OprD porin, which provides carbapenem-specific pores through the outer mem-
brane [9]. Interestingly, the recent development of resistance to linezolid due modi-
fication of the domain V of 23S rRNA (the binding site for linezolid) in
Staphylococcus aureus and Enterococcus spp. has not led to widespread resistance
among clinical isolates [12, 13].
302 P. A. Bradford

10.2.3 Acquired Resistance by DNA Transfer

DNA transfer among bacteria primarily occurs via plasmids, some of which are
self-transmissible, in that they carry genes to initiate the direct transfer to another
bacterium. Many plasmids are large and are able to accommodate multiple resis-
tance genes. These large transferrable plasmids are the ideal vector for the dissemi-
nation of resistance genes. Within plasmids, resistance genes are often carried by
transposons, which can transfer determinants between plasmids, or transport them
into and out of the chromosome [14]. In addition, resistant bacteria often contain
integrons that have the capability to acquire and express resistance determinants
behind a single promoter. They are widely distributed among Gram-negative bacte-
ria and are found within plasmids and transposons [14]. Very diverse resistance
determinants have been found in integrons, including genes conferring target-based
resistance to trimethoprim and fosfomycin, efflux-mediated quinolone resistance,
and metallo-β-lactamase-mediated carbapenem resistance [15–17]. Mechanisms of
transferrable resistance are presented in detail in Chap. 11.
The dissemination of plasmids, transposons, and integrons among bacterial
pathogens has resulted in “gene epidemics” [10]. The TEM-1 plasmid-mediated
β-lactamase was first described in 1965 in an Escherichia coli isolate from a patient
in Greece but has since spread globally to multiple species. It has been found in up
to 60% of clinical isolates of Enterobacteriaceae, to a few Pseudomonas aerugi-
nosa, and up to 50% of Haemophilus influenzae and Neisseria gonorrhoeae isolates
[18]. There are probably multiple factors that determine whether or not a mobilized
gene will spread widely, but these are not well understood. For example, TEM-2
β-lactamase differs from TEM-1 by only a single amino acid substitution and pro-
vides an identical spectrum of resistance. It is also found on similar kinds of plas-
mids and transposons. However, the β-lactamase TEM-2 is at least tenfold less
prevalent than TEM-1 in every region [18].
Many of the resistance determinants now found on plasmids, integrons, and
transposons are believed to have originated in the chromosomes of other bacterial
species, a phenomenon that has been well-documented in plasmid-mediated
β-lactamases. The SHV-type β-lactamases are derived from the chromosomal
β-lactamases of Klebsiella pneumoniae; plasmid-encoded AmpC enzymes expressed
in K. pneumoniae and E. coli are nearly identical to chromosomal AmpC genes
found in E. cloacae (ACT-1, MIR-1), Citrobacter freundii (CMY-type), Hafnia
alvei (ACC-1), Morganella morganii (DHA-1), and the very successful cefotaxime-­
hydrolyzing CTX-M-type ESBLs from Kluyvera spp. [7, 19–21]. In addition, many
aminoglycoside-modifying enzymes found in pathogenic bacteria were determined
to have originated in environmental species of Acinetobacter [22, 23]. Many genes
that are responsible for resistance to antibiotics that are natural products have
migrated from the antibiotic-producing organisms (mostly streptomycetes), which
have developed and retained these genes in order protect themselves against their
own by-products. For example, the erm determinants that methylate 23S rRNA
block binding of macrolides, lincosamides, and group B streptogramins to the target
10 Epidemiology of Bacterial Resistance 303

ribosome are thought to have originated with the producing organism,

Saccharopolyspora erythraea. Most plasmids, integrons, and transposons now carry
multiple resistance genes conferring resistance to antibacterials of many different
drug classes. Selection for any one of these resistance determinants will concur-
rently select for all of the resistance genes contained on this plasmid.
A few bacterial genera, such as α-hemolytic Streptococcus spp., Neisseria spp.,
and Haemophilus spp., are naturally transformable and can absorb and incorporate
fragments of DNA that have been released by lysed organisms in close proximity,
resulting in the creation of “mosaic” genes [24]. Mosaic gene formation is primarily
responsible for penicillin resistance in Streptococcus pneumoniae [25].

10.3 Methods for Tracking Resistance

Typing systems for the epidemiological study of bacterial pathogens are based on
the observation that, although different isolates of the same genus and species share
microbiological, biochemical, serological, and physiological characteristics that
distinguish them from other species, they also have detectable genetic differences
that make discrimination at the intraspecies level possible [26]. In many circum-
stances, intraspecies variability is very high among unrelated isolates, and therefore,
it is easily detectable. However, when dealing with human disease, several species
of bacterial pathogens share overlapping niches and are subjected to identical envi-
ronmental selective pressures. Molecular genetic studies of bacterial populations
have demonstrated that there is some degree of homogeneity between pathogenic
and environmental strains and making genetic differentiation relatively more com-
plicated [27]. Consequently, one must understand that different typing methods give
different, sometimes somewhat contradictory information that should be viewed as
a totality of information for an examination of the phylogenetic and epidemiologi-
cal relationships between pathogens. Molecular typing methods that utilize the
genetic structure of bacterial pathogens have been used to address many different
problems such as the study of genomic organization and evolution. In the context of
bacterial resistance, they are now being used for the identification of patterns of
infection and sources of transmission, the epidemiological surveillance of infec-
tious diseases, and outbreak investigations [28].

10.3.1 Phenotypic Typing Methods

10.3.1.1 Antibiogram

Susceptibility testing can be performed with a number of antimicrobial agents,

including drugs and antiseptic agents, to determine patterns of resistance on a macro
level for most microbial species. Resistance breakpoints that are used clinically for
304 P. A. Bradford

the detection of acquired resistance determinants may not coincide with therapeutic
breakpoints used in the clinical microbiology laboratory. In addition, minimal inhib-
itory concentration (MIC) values are more informative than qualitative resistance
patterns. However, discrimination is dependent on the diversity and relative preva-
lence of detectable resistance in the isolates in question. One drawback to using the
antibiogram for epidemiology is that the stability of resistance pattern can be insuf-
ficient for use as a clonal marker, because resistance determinants may be encoded
on plasmids or resistance genes may be expressed under control of complex regula-
tory systems [29–31]. The antibiogram is often the most valuable first-­line typing
methods in clinical laboratories that can quickly be used to assess the prevalence of
resistance or the appearance of an outbreak strain. However, the integrity of data
used to generate the antibiogram is crucial and is dependent upon the methods used
for determining susceptibility. Many automated systems use short dilution ranges
that surround the breakpoint for a given drug and may not provide enough informa-
tion to discriminate between strains [32]. Nevertheless, the generation of an antibio-
gram has the advantage of being technically easy to use and interpret, even in small
and resource-limited laboratories. It is relatively low-cost test suitable for testing
large numbers of isolates and relies on routine clinical practice. Good reproducibil-
ity allows its use for definitive typing if a standard method such as MIC or disk
diffusion as well as a standard set of marker antibiotics are utilized [28].

10.3.1.2 Serotyping

Traditional serotyping is applicable to single bacterial genus or species by using a

defined set of polyclonal or monoclonal antibodies that detect specific surface anti-
gens on the bacterial cell surface. The discrimination and frequency of cross-­
reactions of serotyping schemes are variable according to the specificity of reagents
[28]. It is considered to be accurate and definitive, but only moderately discrimina-
tory and requires the availability of high quality antisera [33]. In recent years,
molecular serotyping assays have been developed that utilize DNA microarrays to
detect sequences that encode various serovars of a bacterial pathogen. This has been
applied to typing of the O antigen of Salmonella spp. [34, 35]. In addition, gene
sequencing has been used to detect flagellin genes in Campylobacter spp. (flaA),
capsular proteins in S. pneumoniae (cps), and M protein in Group A streptococci
(emm) [36–38]. These arrays and sequencing schemes have been shown to have
comparable results to traditional serotyping [33].

10.3.2 Molecular Typing Methods

Different high resolution molecular-based procedures have been used to detect the
unique features of each individual organism. As a result, guidelines and some inter-
pretive criteria have been proposed in an attempt to standardize what constitutes the
“same strain” [27, 39].
10 Epidemiology of Bacterial Resistance 305

10.3.2.1 Plasmid Analysis

The profile of the number and size of bacterial plasmids, some of which carry anti-
microbial resistance determinants, can be used to determine the relatedness of
strains during an epidemiologic investigation especially when combined with the
utilization of restriction endonucleases to generate a restriction fragment length
polymorphism (RFLP) analysis [40, 41]. However, plasmids that can be transferred
even to other strains, including those of different bacterial species, are often unsta-
ble and may be lost or new ones acquired spontaneously. This makes plasmid fin-
gerprinting somewhat difficult to reproduce [26]. Because of this variability in
plasmid content, the use of plasmid profiling has been found to be insufficient for
use as a clonal marker in some studies [29, 31, 41]. It is best combined with other
genomic typing methods (at the chromosome level) to distinguish between spread
of a clone and that of a plasmid [28].

10.3.2.2 Ribotyping

For ribotyping, chromosomal DNA is cleaved with a frequently cutting restriction

endonuclease such as EcoRI or HindIII followed by conventional gel electrophore-
sis that resolves fragments from 50 to 0.5 kb. This is then followed by Southern blot
hybridization with a probe that detects rRNA genes (rrn) [42, 43]. Because of the
multiple copies of rRNA that are carried by most bacterial pathogens, this results in
a pattern of 5–15 fragments [44]. The level of discrimination achieved with ribotyp-
ing varies depending on the bacterial species and the restriction enzyme used, but is
typically low [45]. However, this can be improved with the use of a second restric-
tion enzyme. Ribotyping has been used to determine whether pretreatment and
posttreatment isolates of ESBL-producing Enterobacteriaceae were the same strain
[46]. Ribotyping was also used to track an outbreak of Clostridium difficile with
reduced susceptibility to vancomycin in a long-term care facility in Israel [47]. At
least one automated system for performing ribotyping has been developed, which
provides consistent data that can be compared across studies (Riboprinter® System,
Qualicon). Using this system, ribotyping identified strains of methicillin-resistant S.
aureus (MRSA) that were genotypically related to community-associated strains
(CA-MRSA) isolated from Phase 3 clinical trials for complicated skin and skin
structure and complicated intraabdominal infections [48].
A somewhat different approach to ribotyping from the RFLP-based method
described above uses PCR to amplify the intergenic region between the genes
encoding 16S and 23S rRNA. Most organisms contain more than one copy of
the rRNA operon; therefore, the size of the intergenic region varies both within
the same strain and between strains [49]. The amplified fragments are often
separated with capillary gel electrophoresis [50]. This method is very reproduc-
ible, but the discriminatory power is moderate [33]. PCR-ribotyping can be
applied to any organism, but in practice, it is mainly used for tracking and sub-
typing C. difficile [50].
306 P. A. Bradford

10.3.2.3 Polymerase Chain Reaction (PCR) Fingerprinting

Several amplification techniques using PCR have been proposed as bacterial typing
systems. The various PCR-based fingerprinting methods may involve the entire
genome by the use of either arbitrary primers or primer pairs directed at the short
sequences lying between repeat motifs in the bacterial genome [28, 51]. They are
universal typing methods that can be applied to most bacterial species and exhibit a
high level of discrimination between strains [51]. Major advantages of these tech-
niques include flexibility, technical simplicity, wide availability of equipment and
reagents, and same-day results [28].

RAPD/AP-PCR

One such PCR fingerprinting technique is the random amplification of polymorphic

DNA with arbitrarily primed PCR (RAPD/AP-PCR). With this method, small
genomic fragments are amplified using short primers (usually <14 bp) with random
sequences that are hybridized under low stringency [52, 53]. Under these conditions,
the primers will bind to both matching sequences and those regions with a few mis-
matches. The result is the amplification of numerous fragments of various sizes,
which is unique to each strain. This technique is slightly less discriminatory than
PFGE and is sometimes difficult to standardize because of the low-stringency PCR
conditions [33]. RAPD fingerprinting has been used to detect the emergence of highly
virulent strains of Helicobacter pylori following incomplete therapy [54]. It has also
been used to track nosocomial outbreaks of CA-MRSA in China and Iran [55, 56].

Rep-PCR

Subtyping of strains based on the amplification of the region between interspersed

repetitive loci in the DNA is called repetitive-element PCR (Rep-PCR). These
duplicative sequences are present in many copies in most bacterial pathogens.
Depending on the proximity of these sequences, they may or may not be amplified
by PCR, resulting in a different sized amplification fragments in each strain [33].
Several different repetitive elements have been identified, but the ones most often
used are extragenic palindromic sequences that are 33–40 bp in length, enterobacte-
rial repetitive intergenic consensus sequences (124–127 bp), and BOX elements
(154 bp, containing three subunits present in varying combinations) [57, 58]. Rep-­
PCR performed by manual methods is not always reproducible; therefore, it is dif-
ficult to compare results across different laboratories [33]. An automated platform
for performing Rep-PCR has been developed (DiversiLab™ System, bioMérieux)
that has standardized the method and improved the reproducibility [59]. Rep-PCR
has recently been used to characterize the isolates of an outbreak of A. baumannii in
Iran [60]. It was also able to track the dominance of CTX-M-type ESBLs among
Enterobacteriaceae from environmental samples in Australia [61].
10 Epidemiology of Bacterial Resistance 307

VNTR/MLVA Analysis

The intergenic regions of many bacterial genomes contain repetitive DNA sequences
that are highly variable with regard to the number or structure of these repeat units.
These variable-number tandem repeats (VNTR) allow for the differentiation
between strains using a PCR-based method [62]. When multiple regions of these
repeat sequences are analyzed, VNTR analysis is sometimes also referred to as mul-
tiple locus variable analysis (MLVA). The speed at which these intergenic regions
change and evolve is also highly variable, making the various loci important for
determining the evolutionary clock for the strains. This adds to the discriminatory
power of this tool [33]. MLVA was famously applied during the 2001 bioterrorism-­
associated anthrax outbreak in the USA [63]. Bacillus anthracis is notoriously dif-
ficult to type using most molecular strain typing methods because there is very little
variability among strains. We used MLVA to subtype 135 isolates from infected
patients, powders from mail sources, and environmental samples. Subtyping of B.
anthracis allowed anthrax cases to be linked to environmental specimens and pow-
ders and provided information about potential sources. MLVA was able to not only
determine that all of the samples involved in the terror attack were identical to each
other but that they were distinguishable from other Ames strains that had been iso-
lated from other sources [63]. VNTR/MLVA has been used for multiple other appli-
cations. VNTR analysis is considered to be the gold standard for molecular typing
of Mycobacterium tuberculosis [64]. It was also shown to be superior to Rep-PCR
for strain discrimination when typing Mycobacterium bovis [65]. In addition, VNTR
analysis was used to characterize an outbreak of linezolid-resistant isolates of
enterococci in Polish hospitals [66].

10.3.2.4 Pulsed-Field Gel Electrophoresis (PFGE)

PFGE analysis (also known as macrorestriction) can simplify the fingerprint of bac-
terial pathogens by utilizing restriction endonucleases that infrequently cut the
chromosomal DNA into a relatively small number of large fragments (20 kb to
>1 Mb) [33]. PFGE uses a current that pulses in alternating directions to separate
and resolve significantly larger fragments of DNA than is possible using constant
field gel electrophoresis. In addition, shearing of these large fragments is avoided by
stabilizing the DNA by embedding samples into an agarose plug [67]. PFGE can be
applied to isolates of most species, although the restriction enzymes used may be
specific to a particular organism [29]. The profiles generated by PFGE are highly
reproducible [26]. One limitation for using PFGE to track resistance is that it is very
labor intensive and can take up to 4 days to complete. Therefore, it cannot be used
in a rapid response to an outbreak.
Interpretive criteria for chromosomal DNA macrorestriction patterns produced
by PFGE have been proposed and guidelines applied successfully for different bac-
terial organisms [39]. Strains are considered to be unrelated if there are three or
more bands (number and/or size of fragments) differing between two isolates. This
308 P. A. Bradford

standardization has allowed for not only the use of PFGE for comparing isolates at
a local level but also across many sites. The Centers for Disease Control (CDC)
developed PulseNet in 1996, which is a national laboratory network that uses PFGE
to detect thousands of local and multistate outbreaks of foodborne illnesses https://
www.cdc.gov/pulsenet/index.html [68]. PFGE has been applied in many different
scenarios for tracking resistance, including carbapenem-resistant K. pneumoniae in
China [69], following NDM-1 among A. baumannii in Israel [70], and determining
nosocomial transmission of MRSA in Malaysia [71].

10.3.2.5 Multilocus Sequence Typing

Multilocus sequence typing (MLST) utilizes nucleotide sequencing to detect varia-

tion due to mutations or recombination in fragments of five to ten housekeeping
genes. Even a single point mutation difference between genes is considered to be a
new allele [33]. The allele types for each of the housekeeping genes are used
together to determine the sequence type (ST). MLST results can be analyzed using
clustering software that can compare the genetic relatedness of strains belonging to
different ST. Isolates with a high degree of similarity (e.g., differing by only one
allele) may be placed into clonal complexes [72]. One advantage to MLST over
PFGE is that the nucleotide sequence data is unambiguous and is not subject to
variations in experimental technique. MLST data can be shared and tracked across
laboratories via several websites such as http://pubmlst.org and http://www.mlst.
net. In recent years, this has been increasingly replaced by whole genome sequenc-
ing with examination of the various loci [73, 74]. In addition, other new technolo-
gies such as the determination of base composition using electrospray ionization-mass
spectrometry have been used [75].
MLST has been used extensively to track and monitor the spread of resistance,
and several ST types have been noted that are highly associated with resistance
mechanisms and have disseminated widely. E. coli sequence type 131 (ST131) has
been identified as the predominant lineage among extraintestinal pathogenic E. coli
and has been attributed to the rapid increase in antimicrobial resistance in that
pathogen [76]. The global dissemination of this sequence type has contributed
immensely to the worldwide emergence of fluoroquinolone-resistant and CTX-M-­
type β-lactamase-producing E. coli [76–78]. Surveillance studies have shown that
the prevalence of ST131 comprises up to 30% of all E. coli clinical isolates, up to
80% of fluoroquinolone-resistant isolates, and up to 60% of ESBL-producing iso-
lates [79].
K. pneumoniae ST258 is a prototype of a high-risk clone and has been largely
responsible for the global spread of carbapenem resistance among the
Enterobacteriaceae [80]. In particular, this ST type is highly associated with the
spread of the KPC-2 carbapenemase [81]. In a global survey of KPC-producing K.
pneumoniae, Peirano et al. found that 290 of 522 (55.6%) isolates from 19 different
countries belonged to ST258 [81]. These isolates were evenly divided between two
subclone groups, and blaKPC was encoded on IncFIIK2-like plasmids in the majority
10 Epidemiology of Bacterial Resistance 309

of the strains. A large outbreak of KPC-producing K. pneumoniae in Warsaw,

Poland, was also determined to be caused by strains belonging to ST258 [82].

10.3.2.6 Whole Genome Sequencing

In recent years, next generation sequencing technology has become an easy and
cost-effective method for performing molecular epidemiology by sequencing the
entire genome of pathogens of interest. The strain relatedness of VRE isolated dur-
ing three outbreaks in a hospital in Sweden was investigated to determine how WGS
would compare to PFGE and MLST. The whole genome sequencing (WGS) data
was analyzed using the average nucleotide identity analysis. PFGE analysis of the
isolates confirmed what was already known by the clinical epidemiological investi-
gation: that three outbreaks had occurred. However, there was no indication of fur-
ther strain relatedness, or if there was a larger cluster. In contrast, the WGS analysis
could clearly distinguish six clusters among the isolates [74]. WGS was also used to
investigate a prolonged outbreak of KPC-producing K. pneumoniae and E. cloacae
in a burn unit in the USA. WGS revealed that this outbreak, which seemed epide-
miologically unrelated, was in fact genetically linked. The outbreak was primarily
maintained by a clonal expansion of E. cloacae sequence ST114 that contained
multiple resistance determinants including blaKPC-3 that was transmitted via plas-
mids containing an identical Tn4401b [83].
Looking at the genome for any differences between strains can be overwhelming
with the amount of data that this generates. Therefore, it is often more useful to
focus on a subset of conserved genes in the bacterial species. Using this approach,
carbapenemase-producing K. pneumoniae isolates from two distinct outbreaks that
occurred in Switzerland in 2013 and 2015 were analyzed. The analysis correctly
identified the two clusters of strains from the two outbreaks and differentiated these
from K. pneumoniae that were unrelated to the outbreak [84]. Many of the previ-
ously described typing methods that utilize PCR and sequencing to detect differ-
ences in strains can now be done by WGS. The PulseNet International network
conducts global laboratory-based surveillance for foodborne illnesses. PulseNet
relies on MLST typing to track outbreaks of many pathogens. Previously, the MLST
was done by PCR and sequencing; however, they have now transitioned into the
standardized use of WGS to perform this subtyping [73, 85]. WGS was recently
used to track an outbreak of carbapenem-resistant K. pneumoniae expressing OXA-­
232 to a contaminated duodenoscope in a California hospital [86].

10.4 Patterns of Resistance

Resistance patterns among bacterial pathogens should be measured and monitored

at many levels. At one level, the epidemiology of resistance is extremely local, and
unique patterns of resistant pathogens can be noted between different wards of the
310 P. A. Bradford

same hospital. In the USA, a study found that the incidence of MRSA, VRE,
ceftazidime-­resistant E. cloacae and P. aeruginosa, and imipenem-resistant P. aeru-
ginosa was two times higher in ICU patients than in patients in general wards or
outpatients at the same hospitals [87]. Likewise, in Europe, the prevalence of MRSA
was noted to be higher in ICUs than in the general patient population [88]. Most
outbreaks and cluster cases of resistant pathogens involve a few patients in a single
unit. The prevalence of resistance is often highest in units where the most debili-
tated patients are located. These units are also often where antibacterial usage is the
greatest, resulting in a constant selective pressure for resistant strains [10].
The epidemiology of resistance can also vary greatly depending on region. In
North America, resistance rates are generally higher in the USA than in Canada.
The prevalence of MRSA among hospitalized adults in Canada was 22–28% but
was 42–45% in the USA [89, 90]. In Europe, an extreme variation in the prevalence
of resistance between countries is noted in that there is a very low incidence in the
Scandinavian countries and very high percentages in the Mediterranean countries.
The incidence of MRSA among S. aureus ranges from <1% in Sweden to 5–9% in
Austria;10–28% in the UK and France; 25–49% in Portugal, Italy, and Greece; and
a high of 50–>75% in Ukraine [91]. Very high prevalence of resistance has been
noted in Asia and Latin America. Prevalence of penicillin-resistant S. pneumoniae
(PRSP) in pediatric patients was 91.3% in Taiwan, 85.8% in Korea, and 70.4% in
Vietnam, compared to <1%–5% in the UK and Scandinavian countries [91, 92].
Differences in the prevalence of specific pathogens are likely due to a combination
of factors including variations in medical care and antibiotic prescribing habits.
Strains with similar resistance phenotypes to organisms causing outbreaks and
epidemics have been found in the same hospitals or patient groups; however, they
fail to spread [10]. The reason for the success of a particular resistant clone is not
well understood. However, some potential factors may include (i) increased
­adherence or virulence mechanisms; (ii) retention of the fitness of the bacterium in
the presence of the resistance genes; (iii) greater tolerance of desiccation, thus
remaining longer in the environment outside of the human body; and (iv) resistance
to disinfectants. A few of these outbreak strains have been studied in detail with
regard to virulence factors contributing to their proliferation. In one study, one of
the major subunits of a new fimbrial protein KPF-28 that aids adherence to the gut
mucosa was found to be co-localized on a plasmid that also encoded the ESBL
SHV-4 in serotype K25 K. pneumoniae strains that were circulating in France and
Belgium in the 1990s [93].

10.5 The Increase and Spread of Resistance

Resistant strains have now disseminated globally. In part, this is due to increased
international travel and open immigration. Several examples of the spread of resis-
tant strains are discussed in the sections below.
10 Epidemiology of Bacterial Resistance 311

10.5.1 MRSA

Shortly after methicillin (then called Celbenin) was introduced into clinical use in
the 1950s in the UK, there were reports of clinical isolates of S. aureus that were
resistant to methicillin [94]. S. aureus strains became methicillin-resistant by the
acquisition and expression of PBP 2a, a low-affinity PBP not native to S. aureus and
encoded by mecA [95]. The mecA gene confers resistance to all β-lactam antibiotics
except for the anti-MRSA cephalosporins ceftaroline and ceftobiprole that can bind
to PBP 2a [96]. The mecA gene is located within a mobile element in the S. aureus
chromosome known as the staphylococcal chromosomal cassette (SCCmec) region.
Numerous different mec regions (SCCmec I to XI) have been described [97–99].
However, transfer of the mec region between staphylococcal strains has never been
documented.
Humans are a natural reservoir for S. aureus, including MRSA, with asymptom-
atic colonization of the nasopharynx, perineum, or skin being commonplace. This
often occurs shortly after birth and may be transient throughout one’s lifetime [100].
Transmission usually occurs by direct contact to a colonized carrier. In adults, the
incidence of MRSA carriage is 25–50% for the general public. A higher incidence
is observed in injection drug users, persons with insulin-dependent diabetes, patients
with dermatologic conditions, and patients with long-term indwelling intravascular
catheters. It is thought that healthcare workers may be an important reservoir for
MRSA [101]. Young children tend to have higher colonization rates, most likely due
to frequent contact with bodily secretions from other children [102]. Epidemic
strains have spread through entire hospitals and even cities. The stability of the
mecA genetic environments differ between various clones of S. aureus clones,
­suggesting a potential explanation for the limited lineages within which the resis-
tance determinant has been found [103]. In 2011, MRSA of both human and animal
origin isolates from Europe which were found to have a divergent mecA homologue
termed mecC were first reported [104]. Isolates harboring mecC can be detected by
susceptibility testing but not by commercial assays targeting mecA or PBP 2a.
Multilocus enzyme electrophoresis and other molecular population genetic tech-
niques were used to determine the extent of mec distribution among phylogenetic
lineages of the species and genetic relationships among MRSA strains isolated from
1961 to 1992 from various geographic regions [105]. This study found that the mec
gene was harbored by many divergent phylogenetic lineages representing wide
chromosomal diversity in the S. aureus. The conclusions were that multiple epi-
sodes of horizontal transfer and recombination contributed to the spread of this
resistance determinant in natural populations of bacteria [105]. However, there can
be regional difference in the spread. This study also identified a single multilocus
enzyme genotype among MRSA isolates recovered in the UK, Denmark,
Switzerland, Egypt, and Uganda, from the 1960s, which indicated that MRSA iso-
lates recovered from those countries at that time were progeny of a single ancestral
cell that had probably recently acquired the mec determinant.
312 P. A. Bradford

Within a few years after that first occurrence, hospital outbreaks caused by
MRSA occurred in Europe. By the mid-1970s, MRSA were recognized as an impor-
tant hospital infection control problem in the USA, and subsequently, these organ-
isms have now achieved global distribution [27]. In a study of bacteremic patients
infected with S. aureus in England and Wales, the incidence of MRSA remained low
(1–3%) for the first years of the study (1989–1993), but then the percentage of
MRSA rapidly increased, reaching 42% by the year 2000 [106]. This increase coin-
cided with the emergence of two new epidemic strains, (designated EMRSA) 15
and 16, which now comprise 95% of all S. aureus bacteremias in England and Wales
[107]. Presently, many industrialized countries report that MRSA comprises at least
25–50% of S. aureus isolated in hospitals from infected patients [89]. In Japan, it
was noted that high antibiotic consumption rates lead to increased MRSA burden
over time [108]. In contrast, some countries such as the Netherlands and the
Scandinavian countries have a low incidence of MRSA infections (around 1%)
[109]. This is likely due to rigid infection control and surveillance policies, as well
as restraint in antibiotic prescription in those countries.
At first, when cases of MRSA infection were identified in the community setting,
an investigation usually exposed a history of recent hospitalization; close contact
with a person who has been hospitalized; or previous antimicrobial-drug therapy.
However, there were some notable exceptions. During 1980–1981, there was an
outbreak of MRSA infections in Detroit among people with no history of hospital-
ization [110]. The majority of these patients were found to be injection drug users.
The source of the Detroit outbreak was never identified, but it is assumed that fre-
quent needle sharing was the mode of transmission in the community setting. The
unexpected deaths of four Native American children with no known risk factors for
acquiring MRSA in the late 1990s launched a new worry of community-associated
MRSA (CA-MRSA) [111]. This more recent development in the spread of MRSA
represents the migration of MRSA being completely healthcare associated
(HA-MRSA) to now often finding CA-MRSA. These strains of CA-MRSA tend to
be microbiologically distinct from HA-MRSA, typically more susceptible to com-
monly used antibiotics than are HA-MRSA. CA-MRSA is thought to have devel-
oped as an independent acquisition of mecA by a strain of methicillin-susceptible S.
aureus (MSSA) [100]. A common clone of CA-MRSA belonging to pulsed-field gel
electrophoresis (PFGE) type USA300 was originally widespread only in the USA;
however, it has now become widespread throughout North America, Latin America,
and the Caribbean [112–114]. CA-MRSA is discussed in detail in Chap. 3.

10.5.2 Penicillin-Resistant S. pneumoniae

S. pneumoniae is a major cause of illness and death worldwide, causing otitis media,
acute sinusitis, community-acquired pneumonia, and meningitis [115]. Since the
introduction of penicillin in the mid-1940s, the treatment of pneumococcal infec-
tions has primarily been with penicillin and other β-lactam antibiotics. Only 20 years
10 Epidemiology of Bacterial Resistance 313

later, strains with decreased susceptibility to β-lactams were detected for the first
time in 1967 [116]. PRSP were found periodically during the late 1970s but then
rapidly emerged and disseminated soon after. In the USA, approximately 5% of
pneumococcal isolates recovered during 1979 to 1987 were reported as testing with
penicillin MICs of ≥1 μg/ml [117]. During the years 1993 to 1994, the percentage
of nonsusceptible isolates was 14.1% but had increased to 25% by 1999 [118].
Antibiotic resistance in S. pneumoniae is now a global public health problem,
although the incidence of PRSP differs by geographic region. 60–89% of S. pneu-
moniae are resistant to penicillin in parts of Latin America and Asia [27].
Penicillin resistance in S. pneumoniae is due to the expression of altered high-­
molecular-­weight penicillin-binding proteins (PBPs) that have reduced affinity for
binding and subsequent inhibition by β-lactam antibiotics [119]. It has been shown
that alterations in the five high molecular-weight PBPs 1A, 2X, 2A, and 2B corre-
lated with resistant patient isolates [25]. Acquisition of penicillin-binding protein
gene segments from foreign donors, such as oral streptococci, thereby creating
mosaic genes, is the primary mechanism for acquired resistance [120]. Following a
recombination event in a single bacterium, drug-resistant progeny can proliferate
with antibiotic pressure and subsequently can be transmitted both locally and glob-
ally by person-to-person spread [121]. The mechanisms of resistance in S. pneu-
moniae are discussed in detail in Chap. 2.
A good illustration of the rapid local clonal spread of antibiotic-resistant S. pneu-
moniae can be found by looking at the history of the development of resistance in
Iceland. During a program to monitor of antibiotic resistance in S. pneumoniae in
Iceland, no PRSP were isolated in years 1983 through most of 1988. However, in
December of 1988, the first penicillin-resistant strain was isolated. The percentage
of PRSP among S. pneumoniae then rose sharply 2.3% in 1989 to 17% by the first
part of 1992 [122]. Approximately 70% of the isolates of PRSP were also resistant
to multiple other antibiotics. Some of these isolates were characterized for related-
ness, including serotype, PBP pattern, PFGE, and multilocus enzyme electrophore-
sis typing [123]. All of the isolates were found to be serotype 6B and had similar or
identical patterns for each of the molecular markers examined. Interestingly, the
PRSP isolated in Iceland were indistinguishable from a subgroup of serotype 6B
PRSP that was frequently isolated in Spain. It was thought that the Spanish clone
was imported to Iceland, as Spain was a popular vacation spot for Icelandic families
with young children [124]. Continuing on with the study, it was noted that in 1995
PRSP comprised 24.2% of all pneumococci. There was then an interesting decline
of the incidence of PRSP to 13.6% in 2001, followed by another rapid increase to
38.6% in 2010. The study found that at some point after 2001, 19F replaced type 6B
in frequency, and by the end of the study, it accounted for 85.8% of all serogrouped
PRSP [124]. The factors responsible for the spread of these clones of PRSP in
Iceland remain unknown. However, it was noted that there was a high amount of
trimethoprim-sulfamethoxazole and tetracycline usage in Iceland compared to other
Nordic countries, which may have contributed to selecting the resistant clones. In
addition, a high proportion of Icelandic children attend day care centers, which may
have contributed to the rapid spread of PRSP in that country [27].
314 P. A. Bradford

The incidence and spread of PRSP took a dramatic downturn following the intro-
duction of the pneumococcal seven-valent conjugate vaccine (PCV7) in the year
2000 that included serotypes 4, 6B, 9V, 14, 18C, 19F, and 23F, which are the most
common serotypes found in invasive pneumococcal disease. As shown in Fig. 10.1,
the incidence of PRSP among pediatric cases of invasive pneumococcal disease was
21.9 per 100,000 cases in 1999 [125]. Approximately 80% of PRSP isolated in the
pre-PCV7 era in the USA were of serotypes commonly carried by healthy children
(serotypes 6B, 9V, 14, 19F, and 23F), and serotype 19A was rare [118]. Following
the introduction of the vaccine, the incidence dropped to 2.3 in just 10 years, fol-
lowed by a further reduction to <1 PRSP in 2013. However, there was also a noted
difference in the serotypes now responsible for β-lactam resistance in S. pneu-
moniae. Following the widespread implementation of the PCV7 vaccine, the highly
resistant serotype 19A became the predominant serotype among PRSP. Fortunately,
a 13 valent vaccine was introduced in 2010 that included serotype 19A; therefore,
the further dissemination of PRSP was limited [126].

10.5.3 Vancomycin-Resistant Enterococcus spp.

Vancomycin-resistant Enterococcus (VRE) faecium was first encountered in clini-

cal isolates in Europe in 1986, followed the next year by isolation of VRE faecalis
in the USA [127, 128]. In Europe, the increased prevalence of VRE was primarily
in the community setting, thought to arise from the use of a glycopeptide antibiotic
avoparcin as a growth promoter in livestock, thus causing transmission from animal
food products to humans [129]. In contrast, in the USA, the predominance of VRE
was in the hospital setting a consequence of the increasing use of the glycopeptide
antibiotic vancomycin due to the prevalence of MRSA in North America [130].
Subsequently, the USA experienced a rapid spread of VRE in hospitals in the 1990s,
followed eventually by a worldwide spread of this resistance [131, 132].
The majority of VRE colonization occurs in the GI tract, but can also be found
on the skin, and in the genitourinary tract [133]. GI colonization with VRE can
persist for months to years and is often refractory to decolonization efforts [134].
Transmission of VRE within the hospital has been traced to the hands of healthcare
workers [135]. In most patients, the VRE colonization does not result in infection.
However, when a patient becomes immunosuppressed, the VRE can flourish, caus-
ing a clinical illness [133]. Additional risk factors for colonization and subsequent
disease has also been linked to previous exposure to antibiotics [136].
E. faecalis is the most common cause of enterococcal infections. However, E.
faecium is intrinsically more resistant to antibiotics with more than half of nosoco-
mial isolates in the USA expressing resistance to ampicillin, vancomycin, and ami-
noglycosides [137]. Among Western countries, the prevalence of VRE is the highest
in North America, with VRE comprising 35.5% of enterococcal hospital-associated
infections. This ranks as the second most common cause of nosocomial infections
in the USA [138]. In contrast, Canada has a much lower incidence of VRE, with
 10 Epidemiology of Bacterial Resistance

Fig. 10.1 Approximate numbers of cases per 100,000 individuals caused by penicillin resistance for each year (left). These data encompass all clonal com-
315

plexes and serotypes associated with penicillin resistance (MIC of 2 μg/ml) in cases of pediatric invasive pneumococcal disease during 1999, 2009, and 2013.
The circle diameters reflect relative incidence of disease [125]
316 P. A. Bradford

only 6% of enterococci in Canada expressing resistance to vancomycin [139]. In

Europe, VRE is less prevalent than in the USA but appears to be increasing in fre-
quency. The European Antimicrobial Resistance Surveillance System (EARSS)
reported a VRE faecium prevalence of 8.3% in 2015 [140]. The prevalence in the
EU is variable depending on the country, with VRE ranging from less than 1% in
France, Spain, and Sweden to greater than 20% in Greece, Portugal, and the UK, all
the way up to 45% in Ireland.

10.5.4 KPC

The first incidence of the Klebsiella pneumoniae carbapenemase (KPC)-type

β-lactamase was first identified in a 1996 isolate of K. pneumoniae from a single
patient in North Carolina [141]. In the next decade, K. pneumoniae expressing
KPC-2 gained a foothold in New York City [142]. Unfortunately, KPC-producing
strains have spread very far beyond the Northeast USA and have spread worldwide
[80]. During the mid-2000s, several outbreaks were detected in hospitals in Israel
due to K. pneumoniae type ST258 that were nearly identical to the KPC-producing
strains in New York, most likely due to the frequent travel back and forth of citi-
zens of both countries [143]. In addition, KPC-type enzymes have been found
among many different genera of Enterobacteriaceae and non-fermentative Gram-
negative pathogens [144, 145]. In a recent survey of Gram-negative pathogens,
KPC-­producing isolates were found in 22 countries on four continents (Fig. 10.2).
Although K. pneumoniae remained the most prevalent pathogen, KPC-type
β-lactamases were also found in Enterobacter aerogenes, Klebsiella oxytoca,
Citrobacter freundii, and several other genera [144]. Recent outbreaks of KPC-
producing K. pneumoniae have been reported across the globe, including China,
Korea, Poland, Portugal, and Switzerland [82, 84, 146–148]. A widely publicized
outbreak of KPC-producing K. pneumoniae was documented in a clinical trial unit
at the National Institutes of Health [149]. In this outbreak, 11 of 18 infected patients
died as a result of their infections. The source of the outbreak was tracked to a
single index case that left the unit 3 weeks prior to the next apparent case. The
outbreak was subsequently halted using strict infection control practices; however,
sporadic cases still occur [149]. In the last few years, the prevalence of KPC among
in K. pneumoniae isolates in NYC has declined. This is most likely due to con-
certed effort to reduce the use of indwelling catheters and increased infection con-
trol [150].
KPC-2 and KPC-3 are the most prevalent isoenzymes among KPC-type
β-lactamases. Bacterial pathogens expressing these enzymes are resistant to a vari-
ety of most β-lactam drugs including expanded-spectrum cephalosporins and car-
bapenems [151]. Some of the newer β-lactam/β-lactamase inhibitor combinations
are active against strains expressing KPC [144, 152]. Bacterial pathogens express-
ing KPC significantly impact clinical management of patients with serious infec-
tions, in that they are often multi- or pan-resistant to many of the currently available
10 Epidemiology of Bacterial Resistance 317

Fig. 10.2 Distribution of KPC-positive Enterobacteriaceae and P. aeruginosa collected in 2012 to

2014 from surveillance. (Figure from [144])

first-line therapeutic options [153, 154]. The successful spread of KPC has been
primarily due to the spread of K. pneumoniae isolates belonging to the successful
clonal complex ST258 [153, 154]. For isolates belonging to ST258, blaKPC is most
often found inside of the Tn4401 transposon. Furthermore, the promiscuous plas-
mids harboring blaKPC also commonly carry genes encoding resistance to aminogly-
cosides and additional β-lactamases, including ESBLs and metallo-β-lactamases
[144, 145]. KPC has also been found in the same strain as metallo-β-lactamases.
blaKPC has also been identified the bacterial chromosome [155].

10.5.5 NDM

The New Delhi metallo-β-lactamase (NDM) was first described in an isolate of K.

pneumoniae from a patient that had previously been admitted to two different hos-
pitals in India during 2007 and then transferred to Sweden where the organism was
identified and characterized [156]. The DNA sequence of NDM-1 shares very little
identity with those of the other common MBLs found in Enterobacteriaceae, with
only 32.4% amino acid identity share with VIM-1, its closest relative [157]. A 2010
study by Kumarasamy et al. showed that NDM-1 was endemic to all parts of India
and Pakistan [158]. They also found that it had spread to the UK, mostly due to
patients with a travel history to the Indian subcontinent (Fig. 10.3). In only 7 years
since those initial few reports, NDM-1 has spread across the globe and has now
been found in a multitude of countries. A recent surveillance study showed that
NDM-type enzymes comprised 44.2% of all MBL-producing Enterobacteriaceae
collected worldwide (Fig. 10.4) [159]. This study also showed that NDM has spread
from K. pneumoniae and E. coli into multiple species of Enterobacteriaceae, P.
318 P. A. Bradford

Fig. 10.3 Distribution of NDM-1 producing Enterobacteriaceae strains in Bangladesh, India,

Pakistan, and the UK in 2010. (Figure from [158])

aeruginosa, and A. baumannii. In addition to the Indian subcontinent, NDM appears

to now be endemic in the Balkan countries, Northern Africa, and on the Arabian
peninsula, countries which may serve as a secondary reservoir [157]. Documented
outbreaks caused by NDM-producing Enterobacteriaceae have recently been
reported in a single hospital in Mexico, multiple centers within the Netherlands, and
in a neonatal unit in China [160–162]. Unlike the case with KPC, high-risk clones
and epidemic plasmids do not seem to play important roles in the global dissemina-
tion of NDMs [80].

10.5.6 Fluoroquinolone Resistance

Resistance to ciprofloxacin appeared quickly as single step mutations among

many Gram-positive pathogens, following its widespread usage beginning in the
1990s. In Taiwan, 11% of Streptococcus pyogenes were found to be resistant to
fluoroquinolones, and this was highly associated with the presence of the erythro-
mycin resistance determinant emm12 [163]. However, susceptibility of S. aureus
to levofloxacin remains high [164]. In contrast, fluoroquinolone resistance among
Enterobacteriaceae is a growing concern. Current surveillance data indicates that
worldwide levofloxacin resistance rates for E. coli have steadily increased. Reports
from the National Healthcare Safety Network that monitors the prevalence fluoro-
quinolone resistance reports that among E. coli isolated from bloodstream isolates
in the USA fluoroquinolone resistance increased from 30.8% in 2006–2007 to
41.1% in 2011 to 49.3% in 2014 (greater than 2% increase per year) [165]. In
Europe, the prevalence appears to be stable with percent fluoroquinolone resis-
tance in invasive E. coli isolates from blood or cerebrospinal fluid at
10 Epidemiology of Bacterial Resistance 319

Fig. 10.4 Global distribution of metallo-β-lactamase-positive Enterobacteriaceae and P. aerugi-

nosa, including NDM-type enzymes collected from 2012 to 2014 from surveillance. (Figure from
[159])

approximately 22.8% during 2012–2015 [140]. However, there was a wide varia-
tion in the percent resistance among different European countries. The fluoroqui-
nolone resistance for Europe ranged from 2.9% in Iceland up to 70% in Slovakia.
A recent survey of 10,000 E. coli isolates from UTI in US hospitals showed an
overall incidence of resistance to fluoroquinolones of 34.5% [166]. The epidemic
CTX-M-15 expressing clone E. coli ST131 is often also resistant to fluoroquino-
lones, and this is thought to have played a significant role in the spread and main-
tenance of this strain [79]. The potential for the spread of fluoroquinolone
resistance has increased with the development of plasmid-meditated fluoroquino-
lone resistance genes such as qnr [167]. In several longitudinal surveys, the inci-
dence of qnr among K. pneumoniae in Taiwan increased from zero in 2000 to
7.6% in 2005 [168]. Similarly, among clinical isolates of Enterobacter spp. iso-
lates from Israel, there were none positive for qnr during 1990–1993, but 6.8%
were positive in 2005 [169].

10.5.7 Aminoglycoside Resistance

Aminoglycosides are used clinically to treat serious infections caused by Gram-­

negative pathogens, especially P. aeruginosa. However, as with most antibiotics that
are widely used, resistance has become widespread. In the EU, the population-­
weighted mean percentage for aminoglycoside resistance in E. coli was 10.4% in
2015; however, there was a wide variation in the percentages of resistant isolates by
country, with a low of 2.9% in Iceland to 24.2% in Slovakia in 2015 [140]. For this
same population, the overall percent resistance among K. pneumoniae isolated in
the EU was 22.5%. A survey of Enterobacteriaceae, Acinetobacter spp., and P.
320 P. A. Bradford

aeruginosa showed that 91.9%, 79.6%, and 63.5%, respectively, of aminoglycoside-­

resistant isolates carried at least one aminoglycoside-modifying enzyme (AME)
[170]. The most prevalent AMEs were aac(6′)-Ib (37.5%), ant(3″)-Ia (25.5%), and
aac(3)-IIa (22.5%), with 13% of isolates carrying two or more AMEs. In another
study from Spain, among 330 aminoglycoside-resistant isolates of
Enterobacteriaceae, the predominant resistant determinant was aph(3″)-Ib (65.4%
of isolates), which correlated to the streptomycin resistance in these strains [171].
AMEs are very often found in clinical isolates that produce other important resis-
tance determinants, such as ESBLs and carbapenemases. A study showed that 98%
of KPC-producing K. pneumoniae that were isolated from two hospitals in the US
(N = 50) also expressed one or more AME [172]. Of these, 98% were positive for
aac(6′)-Ib, which correlated with resistance to tobramycin. Aminoglycoside resis-
tance due to the 16S rRNA methyltransferases (RMT) is much less widespread, and
prevalence varies by region. A surveillance study conducted on isolates from 2007
to 2008 in Asia detected the presence of RMTs in 10.5% from India, 6.9% of iso-
lates from China, 6.1% from Korea, 5.0% from Taiwan, and only 1.5% in Hong
Kong [173]. A lower prevalence of RMTs (≤1.3%) was noted among isolates of
Enterobacteriaceae from Europe [174]. RMTs are frequently found in the same
isolates harboring the NDM-1 metallo-β-lactamase [175, 176].

10.5.8 Colistin Resistance

Until the recent years, the incidence of resistance to colistin was very low. However,
because of the rise of carbapenem resistance among Enterobacteriaceae, there has
been an increased usage of the drugs of polymyxin class, chiefly colistin [177].
Subsequently, reports of colistin resistance are being seen with increasing fre-
quency. A recent study of colistin resistance among a global collection of clinical
isolates of Enterobacteriaceae from 2012 to 2013 showed only 1.6% in the overall
population; 12% in carbapenemase positive isolates were resistant [178]. Colistin
susceptibility was higher among MBL-positive isolates (92.6%) than those positive
for a KPC (87.9%) or OXA-48 (84.2%). In this study, approximately 2.4% of all K.
pneumoniae isolates were colistin resistant compared to only 0.3% of E. coli.
Interestingly, the prevalence of colistin resistance was relatively high among
Enterobacter spp. but varied from 39% in E. asburiae to 0.4% in E. aerogenes. This
study also noted regional differences in the prevalence of colistin resistance. For K.
pneumoniae, the prevalence ranged from 5% in Greece, 4.7% in Italy, and 3.2% in
Romania to 1.3% in North America [178]. Other studies have also shown K. pneu-
moniae resistance rates ranging from 10.5–20% in resistance hotspots such as
Greece to 2.9% in Canada [179].
Previously, the spread of colistin resistance was limited to the expansion of resis-
tant clones because the development of resistance depended upon mutations in the
bacterial chromosome. However, the recent discovery of colistin resistance due to
the plasmid borne mcr-1 may have a very great impact in the coming years. The
10 Epidemiology of Bacterial Resistance 321

mcr-1 gene was first described in a porcine isolate of E. coli in China [180]. Since
that time, a multitude of reports from all over the globe have emerged. It has been
found in patients with Salmonella spp. in China, E. coli in Belgium and Oman, and
K. pneumoniae in South Africa [181–184]. The first detection of the mcr-1 gene in
E. coli from a patient in the USA was reported in 2016 [185]. In a survey of 390
colistin-resistant isolates collected in 2014–2015, none of the K. pneumoniae iso-
lates harbored mcr-1 (0 of 331 isolates), whereas 19 of 59 E. coli isolates had mcr-1
[186]. A study of fecal carriage of mcr-1 in the Netherlands showed detection of the
gene, although among very few patients (0.35%), but it was detected nonetheless
[187]. There continues to be a high correlation of animal to human spread of mcr-1.
This gene was found among patient and animal samples of Salmonella spp. in China
[188]. Similarly, a recent study detected mcr-1 in E. coli and Salmonella spp. iso-
lates of animal origin in Europe, in every year from 2004 to 2014 [189]. It was also
found among E. coli from a pig farm in Germany [190]. There have also been
reports of bacteria containing mcr-1 in sewage and wastewater in China, Germany,
and Spain [191–193]. It is difficult to know if mcr-1 is really spreading or if it is just
an increased awareness of this resistance determinant which is causing its increased
detection.

10.6 Epidemiology of Resistance in Special Populations

Most resistance among bacterial pathogens has been documented in hospitalized

patients, many of which are in the ICU setting. Some infections are spread by con-
tact with healthcare workers. However, studies have shown that drains, sinks, and
faucets in ICU patient rooms were frequently colonized [194]. Although the hospi-
tal setting remains the most common location where resistant pathogens are encoun-
tered, resistant clones have become endemic in certain populations of people within
the community. In addition, some outbreaks have occurred in some interesting
places and groups.

10.6.1 Daycare Centers

Daycare centers that look after infants and preschool-aged children have been
shown to harbor and spread resistant isolates. This is likely due to the fact that these
children frequently share bodily secretions and receive multiple courses of antibiot-
ics in their first few years of life. The high incidence of PRSP in Iceland (discussed
in Sect. 10.5.2 above) was partially attributed to the large number of children that
attend daycare in that country [27]. A study to document the increased carriage of
PRSP by children in daycare centers was undertaken in Japan. Researchers sampled
the nasopharyngeal passages of children from newborns to 6 years attending two
daycare centers [195]. From 363 children cultured, they found that the overall
322 P. A. Bradford

carriage incidence of S. pneumoniae was 3.3%. Of these, the percentages of penicil-

lin nonsusceptible (PNSP) and erythromycin nonsusceptible S. pneumoniae strains
were 36.73% and 71.3%, respectively. Interestingly, using PFGE, they found many
different strain types among the isolates and attributed the genetic diversity of the
resistant strains to the high turnover among the children in the daycare centers
[195]. A longitudinal study of children in Guatemala revealed that from the years
2001 to 2006, the percentage of children colonized with PNSP rose from 1.5% to
33.3% [196].
Similarly, the prevalence of MRSA was studied in two daycare centers in the
USA [102]. In one center, 2 of 61 (3%) children were colonized with MRSA, and 9
of 40 (24%) children were colonized at the second. Ten of 11 of the children were
in classes for 2–3-year-old toddlers. In the second center, all of the MRSA belonged
to one of two PFGE types. None of the daycare workers nor household members of
the toddlers had any previous contact with a hospital, or healthcare worker, indicat-
ing that the daycare center itself was the reservoir for MRSA in that community
[102]. A similar study was conducted in 500 children attending multiple daycare
centers in Brazil [197]. They found that 48% of the children were colonized with S.
aureus and 6.2% were colonized with MRSA. This study also looked at socioeco-
nomic factors associated with resistance. They found that children attending day-
care in low-income public housing sectors were 3.3 times more likely to be colonized
with MRSA than children in other areas [197].

10.6.2 Long-Term Care Facilities

In the early 1990s, long-term care facilities (nursing homes), mostly comprised of
elderly patients, became recognized as an important reservoir for antibiotic-­resistant
pathogens during one of the earliest outbreaks of ESBL producing (TEM-10) that
occurred in Chicago [198]. The source of the outbreak was tracked to an index
patient that had been admitted to the hospital from a long-term care facility.
Subsequently, the authors conducted a study to determine the extent of prevalence
of ESBL-producing strains among these patients [40]. With that goal, they identi-
fied 55 hospitalized patients who were colonized or infected with ESBL-producing
E. coli or K. pneumoniae, during an 18-month period in 1990–1992. Of these 31
patients that were admitted from 8 different nursing homes were positive for TEM-­
10 expressing pathogens, all of which were resistant to ceftazidime, gentamycin,
and tobramycin. As a case-control study, 24 nursing home patients colonized with
resistant strains on hospital admission were compared with 16 patients admitted
from nursing homes that were not colonized on hospital admission. A strong cor-
relation ESBL carriage aligned with poor cognitive function, presence of a gastros-
tomy tube or decubitus ulcers, and prior treatment with antibiotics [40]. At the same
time across the country, another outbreak of ESBL-producing E. coli (SHV-7) was
described among elderly patients from nursing homes admitted to a New York hos-
pital [199]. More recently, the prevalence of ESBL-producing Enterobacteriaceae
10 Epidemiology of Bacterial Resistance 323

was studied among nursing home residents in Germany [200]. Using rectal swabs to
survey, they found that 14.7% of the residents were colonized with ESBL-producing
E. coli. All of the isolates expressed a variant of blaCTX-M, with the most common
isolated being CTX-M-15 in 65.2% and CTX-M-27 in 21.7%. Moreover, 69.6%
could be assigned by polymerase chain reaction (PCR)-typing to the epidemic
clonal lineage E. coli ST131 [200].
Other studies have examined the prevalence of multiple types of resistant patho-
gens among residents of long-term care facilities. Trick et al. conducted a study to
determine the frequency of and risk factors for colonization of nursing home resi-
dents by MRSA, vancomycin-resistant Enterococcus (VRE), and ESBL-producing
K. pneumoniae or E. coli [201]. Of 117 residents that were sampled, 43% were
culture positive in at least one antimicrobial-resistant pathogen as follows: MRSA
24%, VRE 3.5%, ESBL-producing K. pneumoniae 18%, and ESBL-producing E.
coli 15%. At the time of the study, only three of the residents were under contact
isolation precautions. Risk factors for colonization included a total dependence on
healthcare workers for activities of daily living and prior therapy with antibiotics.
Another study examined the molecular epidemiology and antimicrobial suscepti-
bilities of C. difficile strains in long-term care facilities in Israel. Toxigenic C. dif-
ficile isolates were recovered 23.6% of the nursing home patients that were sampled.
Many of these were of the predominant ribotype 027, which had increased MICs of
vancomycin and metronidazole [47]. The authors recommend increased contact iso-
lation and other infection control measures be implemented by the facilities [201].

10.6.3 Sports Teams

Outbreaks of MRSA usually have been associated with healthcare institutions.

However, CA-MRSA is emerging as a cause of skin infections in the community,
many of which have been documented in outbreaks among players of competitive
sports, especially athletes who play contact sports [202]. These outbreaks have
occurred in young, otherwise healthy high school and college students participating
in wrestling, rugby, and American football [203, 204]. In the outbreaks among foot-
ball players, risk factors have included skin trauma from turf burn and direct contact
with lesions of other players [204]. The transmission of MRSA has also been docu-
mented through sharing equipment, towels, or personal items such as skin ointment,
soap, or shaving razors [205, 206]. Among football players, MRSA infections were
more likely to occur in linemen, a position that is more prone to receiving abrasions
than some of the other players, and second in prevalence among cornerbacks and
wide receivers that have frequent direct person-to-person contact with each other
during scrimmage play [204, 207]. MRSA infections among wrestlers has been
linked to contaminated wrestling mats, likely contaminated by infected lesions of
the participants [203, 206]. Transmission has also been linked to whirlpools that are
found in athletic department training facilities at most universities. A study found a
strong correlation between the number of athletes using the whirlpool and the
324 P. A. Bradford

presence of S. aureus in and around the whirlpools [206]. In a systematic review of

the MRSA incidence, a statewide survey was conducted among high school athletes
in Nebraska during the school years 2006–2007 and 2007–2008. Physician docu-
mented MRSA infections were reported among one or more athletes by 4.4% of the
schools during the first year and 14.4% of the schools during the second. During
2007–2008, MRSA incidence per 10,000 wrestlers was 60.1 and 25.1 per 10,000
among football players [208].
Although less common, MRSA outbreaks have also been seen among athletes in
noncontact sports. An outbreak of CA-MRSA was also reported among athletes
participating in a fencing club in Colorado [205]. Five cases with a skin infection
among the fencers plus one household contact were identified by culture. Because
of the patients’ low-risk history, patients with confirmed cases reported recurrent
infections for which they made multiple healthcare visits before their wounds were
cultured. Although the fencers did not share face masks or other equipment, they did
share a sensor wire that is worn under clothing to record when they have been
touched by an opponent’s weapon. Following the implementation of routine clean-
ing of the sensor wires, no additional cases occurred [205].
A number of high profile college and professional football players have been
infected with MRSA. Some of these have resulted in career-ending episodes, includ-
ing amputation [209, 210]. In a very well-documented outbreak among the Saint
Louis Rams, 5 of the 58 Rams players (9%) developed infections with MRSA dur-
ing the 2003 football season [211]. All of the infections developed at turf-abrasion
sites and was significantly associated with the lineman or linebacker position and a
higher body mass index. Four of the five players had direct contact on the field
­during intra-team scrimmages (Fig. 10.5). None of the players were colonized with
MRSA, but it was recovered from whirlpools and taping gel. Using PFGE, the
MRSA from the Rams were determined to be of common CA-MRSA clone USA-­
300. An identical clone was found among some MRSA isolated from an opposing
team that the Rams played shortly after the outbreak started, indicating the clone
had been transferred from the Rams to the second team. In a review of the practices
of the team, several things were noted that probably contributed to the spread of
MRSA among the athletes: (i) hand hygiene products were not available to trainers
that provided wound care to the players near the areas where wound care or physical
therapy was provided; (ii) towels were frequently shared on the field during practice
and games, with as many as three players using the same towel; (iii) players often
did not shower before using communal whirlpools; and (iv) weight training and
therapy equipment at the training facility was not routinely cleaned. The outbreak
was contained after the implementation of infection control hygiene practices in the
training facility (Fig. 10.5) [211].
Physicians should be aware of the potential for MRSA infections in sports par-
ticipants when evaluating patients and making treatment decisions. As demonstrated
by the cluster of MRSA infections among fencers, patients with recurrent MRSA
infections might make multiple healthcare visits before a wound culture is obtained.
Recurrence of infections might be avoided if physicians obtain cultures more rou-
tinely when athletes have infected wounds [205].
10 Epidemiology of Bacterial Resistance 325

Fig. 10.5 Epidemic curve graph (top) and field position diagram (bottom) of cases of MRSA
infection among St. Louis Rams professional football players in 2003. Each box on the epidemic
curve graph and field diagram represents an MRSA infection; different colors designate different
players; boxes of the same color thus represent recurrent infections. On the field diagram, X repre-
sents a defensive player position and O an offensive player position [211]

10.6.4 Resistance Among Travelers

With the ease of international travel in today’s world, there has been an increase in
documented transplantation of previously regional resistant strains into a naïve pop-
ulation of people. This mainly occurs when people return home from abroad, either
infected or colonized with a resistant pathogen, and then transmit it to another indi-
vidual. Recently, there have also been cases of resistant bacterial pathogens being
imported to a destination country by refugees or those seeking asylum from their
country of origin. A very well-documented example of the importation of a resis-
tance mechanism was NDM-1 arriving in the UK in an individual who had recently
returned from India (discussed above in Sect. 10.5.9) [158]. A study in Finland
revealed that 46% of travelers returning from South Asia were colonized with
ESBL-producing Enterobacteriaceae [212].
It has been shown that popular travel destinations can be a localized reservoir for
resistant pathogens. In a systematic review of the literature from 2002–2017,
Leangapichart et al. looked at reports of antibiotic resistance among pilgrims and
326 P. A. Bradford

workers attending the Hajj in Mecca, Saudi Arabia [213]. MRSA carriage was
reported in 20% of pilgrims and food handlers. Carbapenem resistance was detected
in less than 10% of E. coli isolates, but up to 100% of K. pneumoniae and A. bau-
mannii isolates. Colistin-resistant Salmonella spp., E. coli, and K. pneumoniae,
including mcr-1-mediated colistin resistance, were detected among the pilgrims
[213]. A cluster of genetically related, azithromycin-resistant N. gonorrhoeae was
detected in Oahu, Hawaii, in a 1-month period in 2016. The majority of the isolates
were also resistant to ceftriaxone, which eliminated both of the currently recom-
mended therapies for treatment from consideration [214].
Antibiotic resistance has also been noted in cases of traveler’s diarrhea, some of
which caused larger outbreaks in the home country after return. Kim et al. studied
an outbreak of intestinal illness caused by Shigella sonnei in a daycare center in
South Korea. The outbreak strain was resistant to extended-spectrum cephalospo-
rins (via CTX-M-15) and fluoroquinolones. The index case was a child who had
recently returned from traveling to Vietnam [215]. Another study in Belgium exam-
ined antibiotic resistance patterns among isolates of Campylobacter spp. obtained
from returned international travelers that were symptomatic for traveler’s diarrhea
and analyzed the data based on travel destination [216]. For the group as a whole,
60.9% of the isolates of Campylobacter spp. were resistant to ciprofloxacin, with
the prevalence ranging from 50.8% in Africa to 75.0% in Asia. Resistance to
­erythromycin was 4.6%, with the highest incidence (15.2%) seen in isolates from
individuals who had traveled to Southern Asia (six of seven patients returning from
India) and 48.3% were resistant to tetracycline [216]. A recent study examined 453
cases of enteric fever in London caused by Salmonella enterica subspecies Typhi
and Paratyphi. For patients with a history of travel, 88% of S. Paratyphi A isolates
were resistant to ciprofloxacin. For isolate of S. Typhi, 80% were resistant to cipro-
floxacin, 26% to ampicillin, and 27% to chloramphenicol [217].
Migration due to refugees and asylum seekers has been noted to be one of the
risk factors for the spread of multidrug-resistant organisms, which can challenge
local healthcare systems. Ravensbergen et al. analyzed cultures performed in Dutch
hospital on asylum seekers during 2014–2015 and compared the results to those
obtained from Dutch nationals. Of 118 asylum seekers with S. aureus in clinical
cultures, almost 19% were MRSA positive compared to only 1.3% in the general
public. In addition 20.3% were infected with Enterobacteriaceae that produced
ESBLs [109]. Similarly, 20% E. coli isolated from refugees admitted to a German
hospital were found to be ESBL producers [218].

10.7 Conclusions

The spread of resistance to antibacterial agents is dynamic and, at times, unpredict-

able. Although the major resistance determinants have been fully characterized and
addressed with new therapies, new resistance genes continue to develop in response
to new and changing therapies. The changeable nature of resistance due to the
10 Epidemiology of Bacterial Resistance 327

movement of genes by plasmid, transposons, and integrons, plus the upregulation or

downregulation factors such as of multidrug efflux and outer membrane permeabil-
ity, contributes to the complexity of understanding resistance. Many global epidem-
ics of resistance are the result of the spread of a few multidrug-resistant clones, but
the exact reasons contributing to the success of particular lineages remain a mystery.
Regardless of the cause of spread, we now have at our disposal a number of different
molecular methods assisting in the tracking of resistant strains. Studies using these
tools have aided in the identification of the source and elimination of several local-
ized outbreaks of resistant pathogens. Resistance is a significant cause of excess
morbidity, mortality, and cost to hospitalized patients, and with increasing fre-
quency, previously healthy people in the community. In addition to the need for new
and better antibacterial agents, continued monitoring of the epidemiology of
infection-­causing resistant pathogens paired with improved infection control will
hopefully enable us to keep one step ahead of this threat to modern medicine.

Acknowledgment I thank Charles R. Dean for assistance with information gathering.

References

1. CDC. Antibiotic resistance threats in the United States, 2013. 2014. Available from: https://
www.cdc.gov/drugresistance/threat-report-2013/index.html.
2. Woodford N. Glycopeptide-resistant enterococci: a decade of experience. J Med Microbiol.
1998;47(10):849–62.
3. Oppenheim BA. The changing pattern of infection in neutropenic patients. J Antimicrob
Chemother. 1998;41 Suppl D:7–11.
4. Gomez J, Simarro E, Banos V, Requena L, Ruiz J, Garcia F, et al. Six-year prospective study
of risk and prognostic factors in patients with nosocomial sepsis caused by Acinetobacter
baumannii. Eur J Clin Microbiol Infect Dis. 1999;18(5):358–61.
5. Elting LS, Khardori N, Bodey GP, Fainstein V. Nosocomial infection caused by Xanthomonas
maltophilia: a case-control study of predisposing factors. Infect Control Hosp Epidemiol.
1990;11(3):134–8.
6. Miller JH. Mutational specificity in Bacteria. Annu Rev Genet. 1983;17(1):215–38.
7. Bradford PA. Extended-spectrum β-lactamases in the 21st century: characterization,
epidemiology, and detection of this important resistance threat. Clin Microbiol Rev.
2001;14(4):933–51.
8. Podglajen I, Breuil J, Collatz E. Insertion of a novel DNA sequence, 1S1186, upstream of the
silent carbapenemase gene cfiA, promotes expression of carbapenem resistance in clinical
isolates of Bacteroides fragilis. Mol Microbiol. 1994;12(1):105–14.
9. Poole K. Outer membranes and efflux: the path to multidrug resistance in Gram-negative
bacteria. Curr Pharm Biotechnol. 2002;3(2):77–98.
10. Livermore DM. Bacterial resistance: origins, epidemiology, and impact. Clin Infect Dis.
2003;36(Suppl 1):S11–23.
11. Piddock LJ. Mechanisms of fluoroquinolone resistance: an update 1994-1998. Drugs.
1999;58(Suppl 2):11–8.
12. Prystowsky J, Siddiqui F, Chosay J, Shinabarger DL, Millichap J, Peterson LR, et al. Resistance
to linezolid: characterization of mutations in rRNA and comparison of their occurrences in
vancomycin-resistant enterococci. Antimicrob Agents Chemother. 2001;45(7):2154–6.
328 P. A. Bradford

13. Tsiodras S, Gold HS, Sakoulas G, Eliopoulos GM, Wennersten C, Venkataraman

L, et al. Linezolid resistance in a clinical isolate of Staphylococcus aureus. Lancet.
2001;358(9277):207–8.
14. Hall RM. Integrons and gene cassettes: hotspots of diversity in bacterial genomes. Ann N Y
Acad Sci. 2012;1267:71–8.
15. Toulouse JL, Edens TJ, Alejaldre L, Manges AR, Pelletier JN. Integron-associated DfrB4, a
previously uncharacterized member of the trimethoprim-resistant dihydrofolate reductase B
family, is a clinically identified emergent source of antibiotic resistance. Antimicrob Agents
Chemother. 2017;61(5)
16. Lee JJ, Kim M-N, Park KS, Lee JH, Karim AM, Park M, et al. Complex class 1 integron
carrying qnrB62 and blaVIM-2 in a Citrobacter freundii clinical isolate. Antimicrob Agents
Chemother. 2016;60(11):6937–40.
17. KDO P, Campos JC, SCF S, Lezirovitz K, Seco BM, MDO P, et al. fosI is a new integron-­
associated gene cassette encoding reduced susceptibility to fosfomycin. Antimicrob Agents
Chemother. 2016;60(1):686–8.
18. Livermore DM. β-Lactamases in laboratory and clinical resistance. Clin Microbiol Rev.
1995;8(4):557–84.
19. Philippon A, Arlet G, Jacoby GA. Plasmid-determined AmpC-type β-lactamases. Antimicrob
Agents Chemother. 2002;46(1):1–11.
20. Bradford PA, Urban C, Mariano N, Projan SJ, Rahal JJ, Bush K. Imipenem resistance in
Klebsiella pneumoniae is associated with the combination of ACT-1, a plasmid-­mediated
AmpC β-lactamase, and the foss of an outer membrane protein. Antimicrob Agents
Chemother. 1997;41(3):563–9.
21. Oliver A, Perez-Diaz JC, Coque TM, Baquero F, Canton R. Nucleotide sequence and char-
acterization of a novel cefotaxime-hydrolyzing β-lactamase (CTX-M-10) isolated in Spain.
Antimicrob Agents Chemother. 2001;45(2):616–20.
22. Yoon EJ, Goussard S, Nemec A, Lambert T, Courvalin P, Grillot-Courvalin C. Origin in
Acinetobacter gyllenbergii and dissemination of aminoglycoside-modifying enzyme
AAC(6′)-Ih. J Antimicrob Chemother. 2016;71(3):601–6.
23. Yoon EJ, Goussard S, Touchon M, Krizova L, Cerqueira G, Murphy C, et al. Origin in
Acinetobacter guillouiae and dissemination of the aminoglycoside-modifying enzyme
Aph(3′)-VI. MBio. 2014;5(5):e01972-14.
24. Spratt BG. Resistance to antibiotics mediated by target alterations. Science.
1994;264(5157):388–93.
25. Appelbaum PC. Resistance among Streptococcus pneumoniae: implications for drug selec-
tion. Clin Infect Dis. 2002;34(12):1613–20.
26. Stefani S, Agodi A. Molecular epidemiology of antibiotic resistance. Int J Antimicrob Agents.
2000;13(3):143–53.
27. Musser JM. Molecular population genetic analysis of emerged bacterial pathogens: selected
insights. Emerg Infect Dis. 1996;2(1):1–17.
28. Struelens MJ. Consensus guidelines for appropriate use and evaluation of microbial epide-
miologic typing systems. Clin Microbiol Infect. 1996;2(1):2–11.
29. Tenover FC, Arbeit R, Archer G, Biddle J, Byrne S, Goering R, et al. Comparison of tradi-
tional and molecular methods of typing isolates of Staphylococcus aureus. J Clin Microbiol.
1994;32(2):407–15.
30. Dijkshoorn L, Aucken HM, Gerner-Smidt P, Kaufmann ME, Ursing J, Pitt TL. Correlation
of typing methods for Acinetobacter isolates from hospital outbreaks. J Clin Microbiol.
1993;31(3):702–5.
31. Struelens MJ, Rost F, Deplano A, Maas A, Schwam V, Serruys E, et al. Pseudomonas aeru-
ginosa and Enterobacteriaceae bacteremia after biliary endoscopy: an outbreak investigation
using DNA macrorestriction analysis. Am J Med. 1993;95(5):489–98.
32. Wenzler E, Goff DA, Humphries R, Goldstein EJC. Anticipating the unpredictable: a review of
antimicrobial stewardship and Acinetobacter infections. Infect Dis Ther. 2017;6(2):149–72.
10 Epidemiology of Bacterial Resistance 329

33. Gerner-Smidt P, Hyytiä-Trees E, Rota PA. Molecular epidemiology. In: Versalovic J, editor.
Manual of clinical microbiology, vol. 1. 10th ed. Washington, D.C.: American Society for
Microbiology Press; 2011.
34. Yoshida C, Franklin K, Konczy P, McQuiston JR, Fields PI, Nash JH, et al. Methodologies
towards the development of an oligonucleotide microarray for determination of Salmonella
serotypes. J Microbiol Methods. 2007;70(2):261–71.
35. Fitzgerald C, Collins M, van Duyne S, Mikoleit M, Brown T, Fields P. Multiplex, bead-based
suspension array for molecular determination of common Salmonella serogroups. J Clin
Microbiol. 2007;45(10):3323–34.
36. Neal S, Beall B, Ekelund K, Henriques-Normark B, Jasir A, Johnson D, et al. International
quality assurance study for characterization of Streptococcus pyogenes. J Clin Microbiol.
2007;45(4):1175–9.
37. Facklam RF, Martin DR, Lovgren M, Johnson DR, Efstratiou A, Thompson TA, et al.
Extension of the Lancefield classification for group A streptococci by addition of 22 new
M protein gene sequence types from clinical isolates: emm103 to emm124. Clin Infect Dis.
2002;34(1):28–38.
38. Pai R, Gertz RE, Beall B. Sequential multiplex PCR approach for determining capsular sero-
types of Streptococcus pneumoniae isolates. J Clin Microbiol. 2006;44(1):124–31.
39. Tenover FC, Arbeit RD, Goering RV, Mickelsen PA, Murray BE, Persing DH, et al.
Interpreting chromosomal DNA restriction patterns produced by pulsed-field gel electropho-
resis: criteria for bacterial strain typing. J Clin Microbiol. 1995;33(9):2233–9.
40. Wiener J, Quinn JP, Bradford PA, Goering RV, Nathan C, Bush K, et al. Multiple antibiotic-­
resistant Klebsiella and Escherichia coli in nursing homes. JAMA. 1999;281(6):517–23.
41. Mayer LW. Use of plasmid profiles in epidemiologic surveillance of disease outbreaks and in
tracing the transmission of antibiotic resistance. Clin Microbiol Rev. 1988;1(2):228–43.
42. Grimont F, Grimont PA. Ribosomal ribonucleic acid gene restriction patterns as potential
taxonomic tools. Ann Inst Pasteur Microbiol. 1986;137b(2):165–75.
43. Blumberg HM, Rimland D, Kiehlbauch JA, Terry PM, Wachsmuth IK. Epidemiologic typ-
ing of Staphylococcus aureus by DNA restriction fragment length polymorphisms of rRNA
genes: elucidation of the clonal nature of a group of bacteriophage-nontypeable, ciprofloxacin-­
resistant, methicillin-susceptible S. aureus isolates. J Clin Microbiol. 1992;30(2):362–9.
44. Popovic T, Bopp CA, Olsvik O, Kiehlbauch JA. Ribotyping in molecular epidemiology.
In: Persing DH, editor. Diagnostic molecular microbiology, principles and applications.
Washington, D.C.: American Society for Microbiology; 1993. p. 573–83.
45. Bingen EH, Denamur E, Elion J. Use of ribotyping in epidemiological surveillance of noso-
comial outbreaks. Clin Microbiol Rev. 1994;7(3):311–27.
46. Jones CH, Tuckman M, Keeney D, Ruzin A, Bradford PA. Characterization and sequence
analysis of extended-spectrum-β-lactamase-encoding genes from Escherichia coli, Klebsiella
pneumoniae, and Proteus mirabilis isolates collected during tigecycline phase 3 clinical tri-
als. Antimicrob Agents Chemother. 2009;53(2):465–75.
47. Adler A, Miller-Roll T, Bradenstein R, Block C, Mendelson B, Parizade M, et al. A national
survey of the molecular epidemiology of Clostridium difficile in Israel: the dissemination of
the ribotype 027 strain with reduced susceptibility to vancomycin and metronidazole. Diagn
Microbiol Infect Dis. 2015;83(1):21–4.
48. McAleese F, Murphy E, Babinchak T, Singh G, Said-Salim B, Kreiswirth B, et al. Use of
ribotyping to retrospectively identify methicillin-resistant Staphylococcus aureus isolates
from phase 3 clinical trials for tigecycline that are genotypically related to community-­
associated isolates. Antimicrob Agents Chemother. 2005;49(11):4521–9.
49. Bouchet V, Huot H, Goldstein R. Molecular genetic basis of ribotyping. Clin Microbiol Rev.
2008;21(2):262–73.
50. Fawley WN, Knetsch CW, MacCannell DR, Harmanus C, Du T, Mulvey MR, et al.
Development and validation of an internationally-standardized, high-resolution capillary
330 P. A. Bradford

gel-based electrophoresis PCR-ribotyping protocol for Clostridium difficile. PLoS One.

2015;10(2):e0118150.
51. van Belkum A. DNA fingerprinting of medically important microorganisms by use of
PCR. Clin Microbiol Rev. 1994;7(2):174–84.
52. Welsh J, McClelland M. Fingerprinting genomes using PCR with arbitrary primers. Nucleic
Acids Res. 1990;18(24):7213–8.
53. Williams JG, Kubelik AR, Livak KJ, Rafalski JA, Tingey SV. DNA polymorphisms amplified
by arbitrary primers are useful as genetic markers. Nucleic Acids Res. 1990;18(22):6531–5.
54. Bustamante-Rengifo JA, Matta AJ, Pazos AJ, Bravo LE. Effect of treatment failure on
the CagA EPIYA motif in Helicobacter pylori strains from Colombian subjects. World
J Gastroenterol. 2017;23(11):1980–9.
55. Miao J, Chen L, Wang J, Wang W, Chen D, Li L, et al. Evaluation and application of molecu-
lar genotyping on nosocomial pathogen-methicillin-resistant Staphylococcus aureus isolates
in Guangzhou representative of Southern China. Microb Pathog. 2017;107:397–403.
56. Mobasherizadeh S, Shojaei H, Havaei SA, Mostafavizadeh K, Davoodabadi F, Khorvash F,
et al. Application of the random amplified polymorphic DNA (RAPD) fingerprinting to ana-
lyze genetic variation in community associated-methicillin resistant Staphylococcus aureus
(CA-MRSA) isolates in Iran. Global J Health Sci. 2016;8(8):53822.
57. Stern MJ, Ames GF, Smith NH, Robinson EC, Higgins CF. Repetitive extragenic palindromic
sequences: a major component of the bacterial genome. Cell. 1984;37(3):1015–26.
58. Versalovic J, Koeuth T, Lupski JR. Distribution of repetitive DNA sequences in eubacteria and
application to fingerprinting of bacterial genomes. Nucleic Acids Res. 1991;19(24):6823–31.
59. Healy M, Huong J, Bittner T, Lising M, Frye S, Raza S, et al. Microbial DNA typing by
automated repetitive-sequence-based PCR. J Clin Microbiol. 2005;43(1):199–207.
60. Ahmadi A, Salimizand H. Delayed identification of Acinetobacter baumannii during
an outbreak owing to disrupted blaOXA-51-like by ISAba19. Int J Antimicrob Agents.
2017;50(1):119–22.
61. Zarfel G, Lipp M, Gurtl E, Folli B, Baumert R, Kittinger C. Troubled water under the
bridge: screening of River Mur water reveals dominance of CTX-M harboring Escherichia
coli and for the first time an environmental VIM-1 producer in Austria. Sci Total Environ.
2017;593-594:399–405.
62. van Belkum A, Scherer S, van Alphen L, Verbrugh H. Short-sequence DNA repeats in pro-
karyotic genomes. Microbiol Mol Biol Rev 1998;62(2):275–293.
63. Hoffmaster AR, Fitzgerald CC, Ribot E, Mayer LW, Popovic T. Molecular subtyping of
Bacillus anthracis and the 2001 bioterrorism-associated anthrax outbreak, United States.
Emerg Infect Dis. 2002;8(10):1111–6.
64. Jagielski T, Minias A, van Ingen J, Rastogi N, Brzostek A, Żaczek A, et al. Methodological
and clinical aspects of the molecular epidemiology of Mycobacterium tuberculosis and other
mycobacteria. Clin Microbiol Rev. 2016;29(2):239–90.
65. Armas F, Camperio C, Coltella L, Selvaggini S, Boniotti MB, Pacciarini ML, et al.
Comparison of semi-automated commercial rep-PCR fingerprinting, spoligotyping, 12-locus
MIRU-VNTR typing and single nucleotide polymorphism analysis of the embB gene as
molecular typing tools for Mycobacterium bovis. J Med Microbiol. 2017;
66. Gawryszewska I, Żabicka D, Hryniewicz W, Sadowy E. Linezolid-resistant enterococci
in Polish hospitals: species, clonality and determinants of linezolid resistance. Eur J Clin
Microbiol Infect Dis. 2017;36(7):1279–86.
67. Schwartz DC, Cantor CR. Separation of yeast chromosome-sized DNAs by pulsed field gra-
dient gel electrophoresis. Cell. 1984;37(1):67–75.
68. Swaminathan B, Barrett TJ, Hunter SB, Tauxe RV. PulseNet: the molecular subtyping
network for foodborne bacterial disease surveillance, United States. Emerg Infect Dis.
2001;7(3):382–9.
10 Epidemiology of Bacterial Resistance 331

69. Zheng B, Dai Y, Liu Y, Shi W, Dai E, Han Y, et al. Molecular epidemiology and risk factors of
Carbapenem-resistant Klebsiella pneumoniae infections in Eastern China. Front Microbiol.
2017;8:1061.
70. Adler A, Glick R, Lifshitz Z, Carmeli Y. Does Acinetobacter baumannii serve as a source
for blaNDM dissemination into Enterobacteriaceae in hospitalized patients? Microb Drug
Resist. 2017;
71. Sit PS, Teh CS, Idris N, Sam IC, Syed Omar SF, Sulaiman H, et al. Prevalence of methicillin-­
resistant Staphylococcus aureus (MRSA) infection and the molecular characteristics of
MRSA bacteraemia over a two-year period in a tertiary teaching hospital in Malaysia. BMC
Infect Dis. 2017;17(1):274.
72. Feil EJ, Li BC, Aanensen DM, Hanage WP, Spratt BG. eBURST: inferring patterns of evo-
lutionary descent among clusters of related bacterial genotypes from multilocus sequence
typing data. J Bacteriol. 2004;186(5):1518–30.
73. Nadon C, Van Walle I, Gerner-Smidt P, Campos J, Chinen I, Concepcion-Acevedo J, et al.
PulseNet International: Vision for the implementation of whole genome sequencing (WGS)
for global food-borne disease surveillance. Euro Surveill. 2017, 22;(23):pii: 30544.
74. Lytsy B, Engstrand L, Gustafsson A, Kaden R. Time to review the gold standard for genotyp-
ing vancomycin-resistant enterococci in epidemiology: comparing whole-genome sequenc-
ing with PFGE and MLST in three suspected outbreaks in Sweden during 2013-2015. Infect
Genet Evol. 2017;54:74–80.
75. Hall TA, Sampath R, Blyn LB, Ranken R, Ivy C, Melton R, et al. Rapid molecular genotyp-
ing and clonal complex assignment of Staphylococcus aureus isolates by PCR coupled to
electrospray ionization-mass spectrometry. J Clin Microbiol. 2009;47(6):1733–41.
76. Nicolas-Chanoine M-H, Bertrand X, Madec J-Y. Escherichia coli ST131, an intriguing clonal
group. Clin Microbiol Rev. 2014;27(3):543–74.
77. Peirano G, Pitout JDD. Molecular epidemiology of Escherichia coli producing CTX-M
β-lactamases: the worldwide emergence of clone ST131 O25:H4. Int J Antimicrob Agents.
2010;35(4):316–21.
78. Peirano G, Bradford PA, Kazmierczak KM, Badal RE, Hackel M, Hoban DJ, et al.
Global incidence of carbapenemase-producing Escherichia coli ST131. Emerg Infect Dis.
2014;20(11):1928–31.
79. Banerjee R, Johnson JR. A new clone sweeps clean: the enigmatic emergence of Escherichia
coli sequence type 131. Antimicrob Agents Chemother. 2014;58(9):4997–5004.
80. Mathers AJ, Peirano G, Pitout JDD. The role of epidemic resistance plasmids and inter-
national high-risk clones in the spread of multidrug-resistant Enterobacteriaceae. Clin
Microbiol Rev. 2015;28(3):565–91.
81. Peirano G, Bradford PA, Kazmierczak KM, Chen L, Kreiswirth BN, Pitout JD. Importance
of clonal complex 258 and IncFK2-like plasmids among a global collection of Klebsiella
pneumoniae with blaKPC. Antimicrob Agents Chemother. 2017;61(4)
82. Baraniak A, Izdebski R, Zabicka D, Bojarska K, Gorska S, Literacka E, et al. Multiregional
dissemination of KPC-producing Klebsiella pneumoniae ST258/ST512 genotypes in Poland,
2010-14. J Antimicrob Chemother. 2017;72(6):1610–6.
83. Kanamori H, Parobek CM, Juliano JJ, van Duin D, Cairns BA, Weber DJ, et al. A prolonged
outbreak of KPC-3-producing Enterobacter cloacae and Klebsiella pneumoniae driven by
multiple mechanisms of resistance transmission at a large academic burn center. Antimicrob
Agents Chemother. 2017;61(2):pii: e01516-16.
84. Ruppe E, Olearo F, Pires D, Baud D, Renzi G, Cherkaoui A, et al. Clonal or not clonal?
Investigating hospital outbreaks of KPC-producing Klebsiella pneumoniae with whole-­
genome sequencing. Clin Microbiol Infect. 2017;23(7):470–5.
85. Deng X, den Bakker HC, Hendriksen RS. Genomic epidemiology: whole-genome-­
sequencing-­powered surveillance and outbreak investigation of foodborne bacterial patho-
gens. Annu Rev Food Sci Technol. 2016;7:353–74.
332 P. A. Bradford

86. Humphries RM, Yang S, Kim S, Muthusamy VR, Russell D, Trout AM, et al. Duodenoscope-­
related outbreak of a carbapenem resistant Klebsiella pneumoniae identified using advanced
molecular diagnostics. Clin Infect Dis. 2017;65(7):1159–66.
87. Archibald L, Phillips L, Monnet D, JE MG Jr, Tenover F, Gaynes R. Antimicrobial resistance
in isolates from inpatients and outpatients in the United States: increasing importance of the
intensive care unit. Clin Infect Dis. 1997;24(2):211–5.
88. Hetem DJ, Derde LP, Empel J, Mroczkowska A, Orczykowska-Kotyna M, Kozinska A, et al.
Molecular epidemiology of MRSA in 13 ICUs from eight European countries. J Antimicrob
Chemother. 2016;71(1):45–52.
89. Sader HS, Mendes RE, Streit JM, Flamm RK. Antimicrobial susceptibility trends among
Staphylococcus aureus from United States hospitals: results from 7 years of the Ceftaroline
(AWARE) Surveillance Program (2010-2016). Antimicrob Agents Chemother. 2017;61(9):pii:
e01043-17.
90. Adam HJ, Baxter MR, Davidson RJ, Rubinstein E, Fanella S, Karlowsky JA, et al.
Comparison of pathogens and their antimicrobial resistance patterns in paediatric, adult and
elderly patients in Canadian hospitals. J Antimicrob Chemother. 2013;68(Suppl 1):i31–7.
91. ECDC. Surveillance atlas of infectious disease 2017. Available from: http://atlas.ecdc.europa.
eu/public/index.aspx.
92. Lee NY, Song JH, Kim S, Peck KR, Ahn KM, Lee SI, et al. Carriage of antibiotic-resistant
pneumococci among Asian children: a multinational surveillance by the Asian Network for
Surveillance of Resistant Pathogens (ANSORP). Clin Infect Dis. 2001;32(10):1463–9.
93. Di Martino P, Livrelli V, Sirot D, Joly B, Darfeuille-Michaud A. A new fimbrial antigen har-
bored by CAZ-5/SHV-4-producing Klebsiella pneumoniae strains involved in nosocomial
infections. Infect Immun 1996;64(6):2266–2273.
94. Jevons MP. “Celbenin” – resistant staphylococci. Br Med J. 1961;1(5219):124–5.
95. Chambers HF. Methicillin resistance in staphylococci: molecular and biochemical basis and
clinical implications. Clin Microbiol Rev. 1997;10(4):781–91.
96. Kosowska-Shick K, McGhee PL, Appelbaum PC. Affinity of ceftaroline and other β-lactams
for penicillin-binding proteins from Staphylococcus aureus and Streptococcus pneumoniae.
Antimicrob Agents Chemother. 2010;54(5):1670–7.
97. Deurenberg RH, Stobberingh EE. The evolution of Staphylococcus aureus. Infect Genet
Evol. 2008;8(6):747–63.
98. Zhang K, McClure J-A, Elsayed S, Conly JM. Novel Staphylococcal Cassette Chromosome
mec type, tentatively designated type VIII, harboring Class A mec and type 4 ccr gene
complexes in a Canadian epidemicxtrain of methicillin-resistant Staphylococcus aureus.
Antimicrob Agents Chemother. 2009;53(2):531–40.
99. Elements IWGCSCC. Classification of Staphylococcal Cassette Chromosome mec
(SCCmec): guidelines for reporting novel SCCmec elements. Antimicrob Agents Chemother.
2009;53(12):4961–7.
100. Chambers HF. The changing epidemiology of Staphylococcus aureus? Emerg Infect Dis.
2001;7(2):178–82.
101. Albrich WC, Harbarth S. Health-care workers: source, vector, or victim of MRSA? Lancet
Infect Dis. 2008;8(5):289–301.
102. Adcock PM, Pastor P, Medley F, Patterson JE, Murphy TV. Methicillin-resistant
Staphylococcus aureus in two child care centers. J Infect Dis. 1998;178(2):577–80.
103. Katayama Y, Robinson DA, Enright MC, Chambers HF. Genetic background affects stability
of mecA in Staphylococcus aureus. J Clin Microbiol. 2005;43(5):2380–3.
104. Paterson GK, Harrison EM, Holmes MA. The emergence of mecC methicillin-resistant
Staphylococcus aureus. Trends Microbiol. 2014;22(1):42–7.
105. Musser JM, Kapur V. Clonal analysis of methicillin-resistant Staphylococcus aureus strains
from intercontinental sources: association of the mec gene with divergent phylogenetic lin-
eages implies dissemination by horizontal transfer and recombination. J Clin Microbiol.
1992;30(8):2058–63.
10 Epidemiology of Bacterial Resistance 333

106. Reacher MH, Shah A, Livermore DM, Wale MCJ, Graham C, Johnson AP, et al. Bacteraemia
and antibiotic resistance of its pathogens reported in England and Wales between 1990 and
1998: trend analysis. BMJ. 2000;320(7229):213–6.
107. Johnson AP, Aucken HM, Cavendish S, Ganner M, Wale MCJ, Warner M, et al. Dominance
of EMRSA-15 and -16 among MRSA causing nosocomial bacteraemia in the UK: analy-
sis of isolates from the European Antimicrobial Resistance Surveillance System (EARSS).
J Antimicrob Chemother. 2001;48(1):143–4.
108. Nakamura A, Miyake K, Misawa S, Kuno Y, Horii T, Hori S, et al. Association between anti-
microbial consumption and clinical isolates of methicillin-resistant Staphylococcus aureus: a
14-year study. J Infect Chemother. 2012;18(1):90–5.
109. Ravensbergen SJ, Berends M, Stienstra Y, Ott A. High prevalence of MRSA and ESBL
among asylum seekers in the Netherlands. PLoS One. 2017;12(4):e0176481.
110. Saravolatz LD, Markowitz N, Arking L, Pohlod D, Fisher E. Methicillin-resistant
Staphylococcus aureus. Epidemiologic observations during a community-acquired outbreak.
Ann Intern Med. 1982;96(1):11–6.
111. CDC. Four pediatric deaths from community-acquired methicillin-resistant Staphylococcus
aureus – Minnesota and North Dakota, 1997-1999. MMWR Morb Mortal Wkly Rep
1999;48(32):707–710.
112. Ugarte Torres A, Chu A, Read R, MacDonald J, Gregson D, Louie T, et al. The epidemiology
of Staphylococcus aureus carriage in patients attending inner city sexually transmitted infec-
tions and community clinics in Calgary, Canada. PLoS One. 2017;12(5):e0178557.
113. Akpaka PE, Roberts R, Monecke S. Molecular characterization of antimicrobial resistance
genes against Staphylococcus aureus isolates from Trinidad and Tobago. J Infect Public
Health. 2017;10(3):316–23.
114. Zuma AV, Lima DF, Assef AP, Marques EA, Leao RS. Molecular characterization of
methicillin-­resistant Staphylococcus aureus isolated from blood in Rio de Janeiro displaying
susceptibility profiles to non-beta-lactam antibiotics. Braz J Microbiol. 2017;48(2):237–41.
115. Lynch JP, Zhanel GG. Streptococcus pneumoniae: epidemiology, risk factors, and strategies
for prevention. Semin Respir Crit Care. 2009;30(2):189–209.
116. Hansman D, Bullen M. A resistant pneumococcus. Lancet. 1967;290(7509):264–5.
117. Michel J, Dickman D, Greenberg Z, Bergner-Rabinowitz S. Serotype distribution of
penicillin-­resistant pneumococci and their susceptibilities to seven antimicrobial agents.
Antimicrob Agents Chemother. 1983;23(3):397–401.
118. Kyaw MH, Lynfield R, Schaffner W, Craig AS, Hadler J, Reingold A, et al. Effect of introduc-
tion of the pneumococcal conjugate vaccine on drug-resistant Streptococcus pneumoniae. N
Engl J Med. 2006;354(14):1455–63.
119. Jabes D, Nachman S, Tomasz A. Penicillin-binding protein families: evidence for the
clonal nature of penicillin resistance in clinical isolates of pneumococci. J Infect Dis.
1989;159(1):16–25.
120. Dowson CG, Coffey TJ, Kell C, Whiley RA. Evolution of penicillin resistance in Streptococcus
pneumoniae; the role of Streptococcus mitis in the formation of a low affinity PBP2B in S.
pneumoniae. Mol Microbiol. 1993;9(3):635–43.
121. Sibold C, Wang J, Henrichsen J, Hakenbeck R. Genetic relationships of penicillin-­susceptible
and -resistant Streptococcus pneumoniae strains isolated on different continents. Infect
Immun. 1992;60(10):4119–26.
122. Kristinsson KG, Hjalmarsdottir MA, Steingrimsson O. Increasing penicillin resistance in
pneumococci in Iceland. Lancet. 1992;339(8809):1606–7.
123. Soares S, Kristinsson KG, Musser JM, Tomasz A. Evidence for the introduction of a multi-
resistant clone of serotype 6B Streptococcus pneumoniae from Spain to Iceland in the late
1980s. J Infect Dis. 1993;168(1):158–63.
124. Hjalmarsdottir MA, Kristinsson KG. Epidemiology of penicillin-non-susceptible pneumo-
cocci in Iceland, 1995-2010. J Antimicrob Chemother. 2014;69(4):940–6.
334 P. A. Bradford

125. Kim L, McGee L, Tomczyk S, Beall B. Biological and epidemiological features of antibiotic-­
resistant Streptococcus pneumoniae in pre- and post-conjugate vaccine eras: a United States
perspective. Clin Microbiol Rev. 2016;29(3):525–52.
126. Prevention CfDCa. Active Bacterial Core Surveillance (ABCs) Report Emerging Infections
Program Network Streptococcus pneumoniae, 2015. Available from: https://www.cdc.gov/
abcs/reports-findings/survreports/spneu-types.html.
127. Leclercq R, Derlot E, Duval J, Courvalin P. Plasmid-mediated resistance to vancomycin and
teicoplanin in Enterococcus faecium. N Engl J Med. 1988;319(3):157–61.
128. Sahm DF, Kissinger J, Gilmore MS, Murray PR, Mulder R, Solliday J, et al. In vitro suscepti-
bility studies of vancomycin-resistant Enterococcus faecalis. Antimicrob Agents Chemother.
1989;33(9):1588–91.
129. Acar J, Casewell M, Freeman J, Friis C, Goossens H. Avoparcin and virginiamycin as
animal growth promoters: a plea for science in decision-making. Clin Microbiol Infect.
2000;6(9):477–82.
130. Kirst HA, Thompson DG, Nicas TI. Historical yearly usage of vancomycin. Antimicrob
Agents Chemother. 1998;42(5):1303–4.
131. Frieden TR, Munsiff SS, Low DE, Willey BM, Williams G, Faur Y, et al. Emergence of
vancomycin-resistant enterococci in New York City. Lancet. 1993;342(8863):76–9.
132. Deshpande LM, Fritsche TR, Moet GJ, Biedenbach DJ, Jones RN. Antimicrobial resistance
and molecular epidemiology of vancomycin-resistant enterococci from North America and
Europe: a report from the SENTRY antimicrobial surveillance program. Diagn Microbiol
Infect Dis. 2007;58(2):163–70.
133. O'Driscoll T, Crank CW. Vancomycin-resistant enterococcal infections: epidemiology, clini-
cal manifestations, and optimal management. Infect Drug Resist. 2015;8:217–30.
134. Baden LR, Critchley IA, Sahm DF, So W, Gedde M, Porter S, et al. Molecular characterization
of vancomycin-resistant Enterococci repopulating the gastrointestinal tract following treat-
ment with a novel glycolipodepsipeptide, ramoplanin. J Clin Microbiol. 2002;40(4):1160–3.
135. Snyder GM, Thom KA, Furuno JP, Perencevich EN, Roghmann MC, Strauss SM, et al.
Detection of methicillin-resistant Staphylococcus aureus and vancomycin-resistant entero-
cocci on the gowns and gloves of healthcare workers. Infect Control Hosp Epidemiol.
2008;29(7):583–9.
136. Ghanem G, Hachem R, Jiang Y, Chemaly RF, Raad I. Outcomes for and risk factors associated
with vancomycin-resistant Enterococcus faecalis and vancomycin-resistant Enterococcus
faecium bacteremia in cancer patients. Infect Control Hosp Epidemiol. 2007;28(9):1054–9.
137. Hidron AI, Edwards JR, Patel J, Horan TC, Sievert DM, Pollock DA, et al. NHSN annual
update: antimicrobial-resistant pathogens associated with healthcare-associated infec-
tions: annual summary of data reported to the National Healthcare Safety Network at the
Centers for Disease Control and Prevention, 2006-2007. Infect Control Hosp Epidemiol.
2008;29(11):996–1011.
138. Sievert DM, Ricks P, Edwards JR, Schneider A, Patel J, Srinivasan A, et al. Antimicrobial-­
resistant pathogens associated with healthcare-associated infections: summary of data
reported to the National Healthcare Safety Network at the Centers for Disease Control and
Prevention, 2009-2010. Infect Control Hosp Epidemiol. 2013;34:1):1–14.
139. Zhanel GG, Adam HJ, Baxter MR, Fuller J, Nichol KA, Denisuik AJ, et al. Antimicrobial
susceptibility of 22746 pathogens from Canadian hospitals: results of the CANWARD 2007-­
11 study. J Antimicrob Chemother. 2013;68(Suppl 1):i7–22.
140. ECDC. European Centre for Disease Prevention and Control. Antimicrobial resistance
surveillance in Europe 2015. Annual report of the European Antimicrobial Resistance
Surveillance Network (EARS-Net). 2017.
141. Yigit H, Queenan AM, Anderson GJ, Domenech-Sanchez A, Biddle JW, Steward CD, et al.
Novel carbapenem-hydrolyzing β-lactamase, KPC-1, from a carbapenem-resistant strain of
Klebsiella pneumoniae. Antimicrob Agents Chemother. 2001;45(4):1151–61.
10 Epidemiology of Bacterial Resistance 335

142. Bradford PA, Bratu S, Urban C, Visalli M, Mariano N, Landman D, et al. Emergence of
carbapenem-resistant Klebsiella species possessing the class A carbapenem-hydrolyzing
KPC-2 and inhibitor-resistant TEM-30 β-lactamases in New York City. Clin Infect Dis.
2004;39(1):55–60.
143. Navon-Venezia S, Leavitt A, Schwaber MJ, Rasheed JK, Srinivasan A, Patel JB, et al. First
report on a hyperepidemic clone of KPC-3-producing Klebsiella pneumoniae in Israel geneti-
cally related to a strain causing outbreaks in the United States. Antimicrob Agents Chemother.
2009;53(2):818–20.
144. Kazmierczak KM, Biedenbach DJ, Hackel M, Rabine S, de Jonge BLM, Bouchillon SK,
et al. Global dissemination of blaKPC into bacterial species beyond Klebsiella pneumoniae
and in vitro susceptibility to ceftazidime-avibactam and Aztreonam-avibactam. Antimicrob
Agents Chemother. 2016;60(8):4490–500.
145. Y-y H, D-x G, J-c C, H-w Z, Zhang R. Emergence of KPC-2-producing Pseudomonas aeru-
ginosa sequence type 463 isolates in Hangzhou, China. Antimicrob Agents Chemother.
2015;59(5):2914–7.
146. Liang Y, Yin X, Zeng L, Chen S. Clonal replacement of epidemic KPC-producing Klebsiella
pneumoniae in a hospital in China. BMC Infect Dis. 2017;17(1):363.
147. Kim JO, Song SA, Yoon EJ, Shin JH, Lee H, Jeong SH, et al. Outbreak of KPC-2-producing
Enterobacteriaceae caused by clonal dissemination of Klebsiella pneumoniae ST307 car-
rying an IncX3-type plasmid harboring a truncated Tn4401a. Diagn Microbiol Infect Dis.
2017;87(4):343–8.
148. Vubil D, Figueiredo R, Reis T, Canha C, Boaventura L, GJ DAS. Outbreak of KPC-3-­
producing ST15 and ST348 Klebsiella pneumoniae in a Portuguese hospital. Epidemiol
Infect. 2017;145(3):595–9.
149. Snitkin ES, Zelazny AM, Thomas PJ, Stock F, Henderson DK, Palmore TN, et al. Tracking
a hospital outbreak of carbapenem-resistant Klebsiella pneumoniae with whole-genome
sequencing. Sci Transl Med. 2012;4(148):148ra16-ra16.
150. Abdallah M, Olafisoye O, Cortes C, Urban C, Landman D, Ghitan M, et al. Rise and fall
of KPC-producing Klebsiella pneumoniae in New York City. J Antimicrob Chemother.
2016;71(10):2945–8.
151. Walther-Rasmussen J, Høiby N. Class A carbapenemases. J Antimicrob Chemother.
2007;60(3):470–82.
152. Hecker SJ, Reddy KR, Totrov M, Hirst GC, Lomovskaya O, Griffith DC, et al. Discovery
of a cyclic boronic acid beta-lactamase inhibitor (RPX7009) with utility vs class A serine
carbapenemases. J Med Chem. 2015;58(9):3682–92.
153. Munoz-Price LS, Poirel L, Bonomo RA, Schwaber MJ, Daikos GL, Cormican M, et al.
Clinical epidemiology of the global expansion of Klebsiella pneumoniae carbapenemases.
Lancet Infect Dis. 2013;13(9):785–96.
154. Pitout JDD, Nordmann P, Poirel L. Carbapenemase-producing Klebsiella pneumoniae,
a key pathogen set for global nosocomial dominance. Antimicrob Agents Chemother.
2015;59(10):5873–84.
155. Chen L, Chavda KD, DeLeo FR, Bryant KA, Jacobs MR, Bonomo RA, et al. Genome
sequence of a Klebsiella pneumoniae sequence type 258 isolate with prophage-encoded K.
pneumoniae carbapenemase. Genome Announc. 2015;3(3)
156. Yong D, Toleman MA, Giske CG, Cho HS, Sundman K, Lee K, et al. Characterization of
a new metallo-β-lactamase gene, blaNDM-1, and a novel erythromycin esterase gene car-
ried on a unique genetic structure in Klebsiella pneumoniae sequence type 14 from India.
Antimicrob Agents Chemother. 2009;53(12):5046–54.
157. Nordmann P, Poirel L. The difficult-to-control spread of carbapenemase producers among
Enterobacteriaceae worldwide. Clin Microbiol Infect. 2014;20(9):821–30.
158. Kumarasamy KK, Toleman MA, Walsh TR, Bagaria J, Butt F, Balakrishnan R, et al.
Emergence of a new antibiotic resistance mechanism in India, Pakistan, and the UK: a molec-
ular, biological, and epidemiological study. Lancet Infect Dis. 2010;10(9):597–602.
336 P. A. Bradford

159. Kazmierczak KM, Rabine S, Hackel M, McLaughlin RE, Biedenbach DJ, Bouchillon SK,
et al. Multiyear, multinational survey of the incidence and global distribution of metallo-β-­
lactamase-producing Enterobacteriaceae and Pseudomonas aeruginosa. Antimicrob Agents
Chemother. 2016;60(2):1067–78.
160. Bocanegra-Ibarias P, Garza-González E, Morfín-Otero R, Barrios H, Villarreal-Treviño L,
Rodríguez-Noriega E, et al. Molecular and microbiological report of a hospital outbreak of
NDM-1-carrying Enterobacteriaceae in Mexico. PLoS One. 2017;12(6):e0179651.
161. Bosch T, Lutgens SPM, Hermans MHA, Wever PC, Schneeberger PM, Renders NHM, et al.
An outbreak of NDM-1 producing Klebsiella pneumoniae in a Dutch hospital with interspe-
cies transfer of the resistance plasmid and unexpected occurrence in unrelated healthcare
centers. J Clin Microbiol. 2017;
162. Yu J, Wang Y, Chen Z, Zhu X, Tian L, Li L, et al. Outbreak of nosocomial NDM-1-­
producing Klebsiella pneumoniae ST1419 in a neonatal unit. J Global Antimicrobial Resist.
2017;8:135–9.
163. Lin J-N, Chang L-L, Lai C-H, Huang Y-H, Chen W-F, Yang C-H, et al. High prevalence of
fluoroquinolone-nonsusceptible Streptococcus pyogenes emm12 in Taiwan. Diagn Microbiol
Infect Dis. 2015;83(2):187–92.
164. McCurdy SP, Jones RN, Mendes RE, Puttagunta S, Dunne MW. In vitro activity of
Dalbavancin against drug-resistant Staphylococcus aureus isolates from a global surveillance
program. Antimicrob Agents Chemother. 2015;59(8):5007–9.
165. Weiner LM, Webb AK, Limbago B, Dudeck MA, Patel J, Kallen AJ, et al. Antimicrobial-­
resistant pathogens associated with healthcare-associated infections: summary of data
reported to the National Healthcare Safety Network at the Centers for Disease Control and
Prevention, 2011–2014. Infect Control Hosp Epidemiol. 2016;37(11):1288–301.
166. Bidell MR, Palchak M, Mohr J, Lodise TP. Fluoroquinolone and third-generation-­
cephalosporin resistance among hospitalized patients with urinary tract infections due to
Escherichia coli: do rates vary by hospital characteristics and geographic region? Antimicrob
Agents Chemother. 2016;60(5):3170–3.
167. Strahilevitz J, Jacoby GA, Hooper DC, Robicsek A. Plasmid-mediated quinolone resistance:
a multifaceted threat. Clin Microbiol Rev. 2009;22(4):664–89.
168. Wu JJ, Ko WC, Wu HM, Yan JJ. Prevalence of Qnr determinants among bloodstream iso-
lates of Escherichia coli and Klebsiella pneumoniae in a Taiwanese hospital, 1999-2005.
J Antimicrob Chemother. 2008;61(6):1234–9.
169. Strahilevitz J, Engelstein D, Adler A, Temper V, Moses AE, Block C, et al. Changes in
qnr prevalence and fluoroquinolone resistance in clinical isolates of Klebsiella pneu-
moniae and Enterobacter spp. collected from 1990 to 2005. Antimicrob Agents Chemother.
2007;51(8):3001–3.
170. Castanheira M, Costello SE, Jones RN, Mendes RE, editors. Prevalence of aminoglycoside
resistance genes among contemporary Gram-negative resistant isolates collected worldwide.
25th European Congress of Clinical Microbiology and Infectious Diseases (ECCMID); 2015
April 25–28; Copenhagen.
171. Miro E, Grunbaum F, Gomez L, Rivera A, Mirelis B, Coll P, et al. Characterization of
aminoglycoside-­modifying enzymes in Enterobacteriaceae clinical strains and characteriza-
tion of the plasmids implicated in their diffusion. Microb Drug Resist. 2013;19(2):94–9.
172. Almaghrabi R, Clancy CJ, Doi Y, Hao B, Chen L, Shields RK, et al. Carbapenem-resistant
Klebsiella pneumoniae strains exhibit diversity in aminoglycoside-modifying enzymes,
which exert differing effects on Plazomicin and other agents. Antimicrob Agents Chemother.
2014;58(8):4443–51.
173. Bell J, Andersson P, Jones RN, Turnidge J, editors. 16S rRNA methylase containing
Enterobacteriaceae in the SENTRY Asia-Pacific region frequently harbour plasmid-mediated
quinolone resistance and CTX-M types. European Congress of Clinical Microbiology and
Infectious Diseases (ECCMID); 2010 April 10–13; Vienna.
10 Epidemiology of Bacterial Resistance 337

174. Wachino J, Arakawa Y. Exogenously acquired 16S rRNA methyltransferases found in

aminoglycoside-­resistant pathogenic Gram-negative bacteria: an update. Drug Resist Updat.
2012;15(3):133–48.
175. Bercot B, Poirel L, Nordmann P. Updated multiplex polymerase chain reaction for detection
of 16S rRNA methylases: high prevalence among NDM-1 producers. Diagn Microbiol Infect
Dis. 2011;71(4):442–5.
176. Poirel L, Savov E, Nazli A, Trifonova A, Todorova I, Gergova I, et al. Outbreak caused by
NDM-1- and RmtB-producing Escherichia coli in Bulgaria. Antimicrob Agents Chemother.
2014;58(4):2472–4.
177. Biswas S, Brunel J-M, Dubus J-C, Reynaud-Gaubert M, Rolain J-M. Colistin: an update on
the antibiotic of the 21st century. Expert Rev Anti-Infect Ther. 2012;10(8):917–34.
178. Bradford PA, Kazmierczak KM, Biedenbach DJ, Wise MG, Hackel M, Sahm DF. Correlation
of β-lactamase production and Colistin resistance among Enterobacteriaceae isolates from a
global surveillance program. Antimicrob Agents Chemother. 2016;60(3):1385–92.
179. Ah Y-M, Kim A-J, Lee J-Y. Colistin resistance in Klebsiella pneumoniae. Int J Antimicrob
Agents. 2014;44(1):8–15.
180. Liu Y-Y, Wang Y, Walsh TR, Yi L-X, Zhang R, Spencer J, et al. Emergence of plasmid-­
mediated colistin resistance mechanism MCR-1 in animals and human beings in China: a
microbiological and molecular biological study. Lancet Infect Dis. 2016;16(2):161–8.
181. Cui M, Zhang J, Gu Z, Li R, EW-c C, Yan M, et al. Prevalence and molecular characterization
of mcr-1-positive Salmonella strains recovered from clinical specimens in China. Antimicrob
Agents Chemother. 2017;61(5)
182. Huang TD, Bogaerts P, Berhin C, Hoebeke M, Bauraing C, Glupczynski Y. Increasing propor-
tion of carbapenemase-producing Enterobacteriaceae and emergence of a MCR-1 ­producer
through a multicentric study among hospital-based and private laboratories in Belgium from
September to November 2015. Euro Surveill. 2017;22(19).
183. Mohsin J, Pal T, Petersen JE, Darwish D, Ghazawi A, Ashraf T, et al. Plasmid-mediated
Colistin resistance gene mcr-1 in an Escherichia coli ST10 bloodstream isolate in the
Sultanate of Oman. Microb Drug Resist. 2017;
184. Newton-Foot M, Snyman Y, Maloba MRB, Whitelaw AC. Plasmid-mediated mcr-1 colistin
resistance in Escherichia coli and Klebsiella spp. clinical isolates from the Western Cape
region of South Africa. Antimicrob Resist Infect Control. 2017;6:78.
185. McGann P, Snesrud E, Maybank R, Corey B, Ong AC, Clifford R, et al. Escherichia coli
Harboring mcr-1 and blaCTX-M on a novel IncF plasmid: first report of mcr-1 in the United
States. Antimicrob Agents Chemother. 2016;60(7):4420–1.
186. Castanheira M, Griffin MA, Deshpande LM, Mendes RE, Jones RN, Flamm RK. Detection
of mcr-1 among Escherichia coli clinical isolates collected worldwide as part of the SENTRY
antimicrobial surveillance program in 2014 and 2015. Antimicrob Agents Chemother.
2016;60(9):5623–4.
187. Terveer EM, Nijhuis RHT, Crobach MJT, Knetsch CW, Veldkamp KE, Gooskens J, et al.
Prevalence of colistin resistance gene (mcr-1) containing Enterobacteriaceae in feces of
patients attending a tertiary care hospital and detection of a mcr-1 containing, colistin sus-
ceptible E. coli. PLoS One. 2017;12(6):e0178598.
188. Chiou C-S, Chen Y-T, Wang Y-W, Liu Y-Y, Kuo H-C, Tu Y-H, et al. Dissemination of mcr-­
1-­carrying plasmids among Colistin-resistant Salmonella strains from humans and food-­
producing animals in Taiwan. Antimicrob Agents Chemother. 2017;61(7)
189. El Garch F, Sauget M, Hocquet D, LeChaudee D, Woehrle F, Bertrand X. mcr-1 is borne by
highly diverse Escherichia coli isolates since 2004 in food-producing animals in Europe. Clin
Microbiol Infect 2017;23(1):51.e1–51.e4.
190. Roschanski N, Falgenhauer L, Grobbel M, Guenther S, Kreienbrock L, Imirzalioglu C, et al.
Retrospective survey of mcr-1 and mcr-2 in German pig-fattening farms, 2011-2012. Int
J Antimicrob Agents. 2017;50(2):266–71.
338 P. A. Bradford

191. Zhou H-W, Zhang T, Ma J-H, Fang Y, Wang H-Y, Huang Z-X, et al. Occurrence of plas-
mid- and chromosome-carried mcr-1 in waterborne Enterobacteriaceae in China. Antimicrob
Agents Chemother. 2017;61(8)
192. Hembach N, Schmid F, Alexander J, Hiller C, Rogall ET, Schwartz T. Occurrence of the
mcr-1 colistin resistance gene and other clinically relevant antibiotic resistance genes in
microbial populations at different municipal wastewater treatment plants in Germany. Front
Microbiol. 2017;8:1282.
193. Ovejero CM, Delgado-Blas JF, Calero-Caceres W, Muniesa M, Gonzalez-Zorn B. Spread of
mcr-1-carrying Enterobacteriaceae in sewage water from Spain. J Antimicrob Chemother.
2017;72(4):1050–3.
194. Kizny Gordon AE, Mathers AJ, Cheong EYL, Gottlieb T, Kotay S, Walker AS, et al.
The hospital water environment as a reservoir for carbapenem-resistant organisms caus-
ing hospital-acquired infections-a systematic review of the literature. Clin Infect Dis.
2017;64(10):1435–44.
195. Hashida K, Shiomori T, Hohchi N, Ohkubo J, Ohbuchi T, Mori T, et al. Nasopharyngeal
Streptococcus pneumoniae carriage in Japanese children attending day-care centers. Int
J Pediatr Otorhinolaryngol. 2011;75(5):664–9.
196. Dueger EL, Asturias EJ, Matheu J, Gordillo R, Torres O, Halsey N. Increasing penicillin
and trimethoprim-sulfamethoxazole resistance in nasopharyngeal Streptococcus pneumoniae
isolates from Guatemalan children, 2001-2006. Int J Infect Dis. 2008;12(3):289–97.
197. Braga EDV, Aguiar-Alves F, de Freitas MFN, de e Silva MO, Correa TV, Snyder RE, et al.
High prevalence of Staphylococcus aureus and methicillin-resistant S. aureus colonization
among healthy children attending public daycare centers in informal settlements in a large
urban center in Brazil. BMC Infect Dis. 2014;14
198. Rasmussen BA, Bradford PA, Quinn JP, Wiener J, Weinstein RA, Bush K. Genetically
diverse ceftazidime-resistant isolates from a single center: biochemical and genetic charac-
terization of TEM-10 β-lactamases encoded by different nucleotide sequences. Antimicrob
Agents Chemother. 1993;37(9):1989–92.
199. Bradford PA, Urban C, Jaiswal A, Mariano N, Rasmussen BA, Projan SJ, et al. SHV-7, a
novel cefotaxime-hydrolyzing β-lactamase, identified in Escherichia coli isolates from hos-
pitalized nursing home patients. Antimicrob Agents Chemother. 1995;39(4):899–905.
200. Valenza G, Nickel S, Pfeifer Y, Pietsch M, Voigtlander E, Lehner-Reindl V, et al. Prevalence
and genetic diversity of extended-spectrum beta-lactamase (ESBL)-producing Escherichia
coli in nursing homes in Bavaria, Germany. Vet Microbiol. 2017;200:138–41.
201. Trick WE, Weinstein RA, DeMarais PL, Kuehnert MJ, Tomaska W, Nathan C, et al.
Colonization of skilled-care facility residents with antimicrobial-resistant pathogens. J Am
Geriatr Soc. 2001;49(3):270–6.
202. Jiménez-Truque N, Saye EJ, Soper N, Saville BR, Thomsen I, Edwards KM, et al.
Longitudinal assessment of colonization with Staphylococcus aureus in healthy collegiate
athletes. J Pediatric Infect Dis Soc. 2016;5(2):105–13.
203. Lindenmayer JM, Schoenfeld S, O'Grady R, Carney JK. Methicillin-resistant Staphylococcus
aureus in a high school wrestling team and the surrounding community. Arch Intern Med.
1998;158(8):895–9.
204. Begier EM, Frenette K, Barrett NL, Mshar P, Petit S, Boxrud DJ, et al. A high-morbidity out-
break of methicillin-resistant Staphylococcus aureus among players on a college football team,
facilitated by cosmetic body shaving and turf burns. Clin Infect Dis. 2004;39(10):1446–53.
205. CDC. Methicillin-resistant Staphylococcus aureus infections among competitive sports par-
ticipants–Colorado, Indiana, Pennsylvania, and Los Angeles County, 2000–2003. MMWR.
2003;52:793–5.
206. Young LM, Motz VA, Markey ER, Young SC, Beaschler RE. Recommendations for best
disinfectant practices to reduce the spread of infection via wrestling mats. J Athl Train.
2017;52(2):82–8.
10 Epidemiology of Bacterial Resistance 339

207. Dao MN, Laurene M, Elizabeth B. Recurring methicillin-resistant Staphylococcus aureus

infections in a football team. Emerg Infect Dis J. 2005;11(4):526.
208. Buss BF, Mueller SW, Theis M, Keyser A, Safranek TJ. Population-based estimates of
methicillin-­resistant Staphylococcus aureus (MRSA) infections among high school athletes –
Nebraska, 2006-2008. J Sch Nurs. 2009;25(4):282–91.
209. McKenna M. National Geographic 2015. Available from: http://phenomena.nationalgeo-
graphic.com/2015/10/15/mrsa-football/.
210. Perez AJ. Giants TE. Daniel Fells to have 10th surgery to treat MRSA USA Today
2015. Available from: https://www.usatoday.com/story/sports/nfl/giants/2015/12/07/
daniel-fells-mrsa-surgery-new-york-giants-nfl/76915784/.
211. Kazakova SV, Hageman JC, Matava M, Srinivasan A, Phelan L, Garfinkel B, et al. A clone
of methicillin-resistant Staphylococcus aureus among professional football players. N Engl
J Med. 2005;352(5):468–75.
212. Kantele A, Laaveri T, Mero S, Vilkman K, Pakkanen SH, Ollgren J, et al. Antimicrobials
increase travelers’ risk of colonization by extended-spectrum betalactamase-producing
Enterobacteriaceae. Clin Infect Dis. 2015;60(6):837–46.
213. Leangapichart T, Rolain JM, Memish ZA, Al-Tawfiq JA, Gautret P. Emergence of drug resis-
tant bacteria at the Hajj: a systematic review. Travel Med Infect Dis. 2017;
214. Katz AR, Komeya AY, Kirkcaldy RD, Whelen AC, Soge OO, Papp JR, et al. Cluster of
Neisseria gonorrhoeae isolates with high-level azithromycin resistance and decreased ceftri-
axone susceptibility, Hawaii, 2016. Clin Infect Dis. 2017;
215. Kim JS, Kim JJ, Kim SJ, Jeon S-E, Seo KY, Choi J-K, et al. Outbreak of ciprofloxacin-­
resistant Shigella sonnei associated with travel to Vietnam, Republic of Korea. Emerg Infect
Dis. 2015;21(7):1247–50.
216. Post A, Martiny D, van Waterschoot N, Hallin M, Maniewski U, Bottieau E, et al. Antibiotic
susceptibility profiles among Campylobacter isolates obtained from international travelers
between 2007 and 2014. Eur J Clin Microbiol Infect Dis. 2017;
217. Dave J, Warburton F, Freedman J, de Pinna E, Grant K, Sefton A, et al. What were the risk
factors and trends in antimicrobial resistance for enteric fever in London 2005-2012? J Med
Microbiol. 2017;
218. Reinheimer C, Kempf VAJ, Jozsa K, Wichelhaus TA, Hogardt M, O’Rourke F, et al.
Prevalence of multidrug-resistant organisms in refugee patients, medical tourists and domes-
tic patients admitted to a German university hospital. BMC Infect Dis. 2017;17:17.
Chapter 11
Transmissible Antibiotic Resistance

George A. Jacoby

11.1 Introduction

Transmissible antibiotic resistance was discovered by Japanese investigators in

1959 and publicized to the rest of the world in a review in English by Watanabe in
1963 [1]. Its discovery was motivated by a puzzling increase in resistance of clini-
cal Shigella isolates in Japan following World War II with many strains co-resis-
tant to chloramphenicol, streptomycin, sulfonamide, and tetracycline. All four
resistances proved transmissible to Escherichia coli by conjugation and were car-
ried on an element separate from the bacterial chromosome and able to replicate
autonomously that came to be called a plasmid [2]. Plasmid-mediated penicillin
resistance was inferred in Staphylococcus aureus in 1963 [3], and transmissible
resistance to ampicillin was demonstrated in E. coli in 1965 [4]. Since then plas-
mid-mediated resistance to a particular antibiotic has followed its clinical use by
from 2 to 58 years (Table 11.1) with the longer latencies generally associated with
low use of the drug. For example, colistin was approved in 1958 but little used
because of toxicity concerns until it became a drug of last resort in the 2000s and
also widely used in animal husbandry before plasmid-mediated colistin resistance
was reported in 2016 [9].
This chapter will address the biochemical and molecular mechanisms of resis-
tance, the sources of resistance genes, and how they are acquired by transmissible
elements. Resistance mediated by plasmids, insertion sequences, transposons, and
related transmissible elements will be included along with a few considerations
about what can be done to reduce or circumvent such spread.

G. A. Jacoby (*)
Lahey Hospital and Medical Center, Burlington, MA, USA

© Springer International Publishing AG, part of Springer Nature 2018 341

I. W. Fong et al. (eds.), Antimicrobial Resistance in the 21st Century, Emerging Infectious
Diseases of the 21st Century, https://doi.org/10.1007/978-3-319-78538-7_11
342 G. A. Jacoby

Table 11.1 Timetable of antibiotic approval and appearance of transmissible resistance

Approved for Transmissible resistance
Antibiotic human use reported Organism Reference
Sulfonamide 1936 1959 Shigella sp. [1]
Penicillin G 1942 1963 S. aureus [3]
Streptomycin 1944 1959 Shigella sp. [1]
Bacitracina 1945 2004 E. faecalis [5]
Chlortetracycline 1948 1959 Shigella sp. [1]
Oxytetracycline 1950
Tetracycline 1955
Chloramphenicol 1949 1959 Shigella sp. [1]
Erythromycin 1952 1963 S. aureus [6]
Nitrofurantoin 1953 1983 E. coli [7]
Vancomycin 1955 1988 E. faecium [8]
Colistin 1958 2016 E. coli [9]
Metronidazole 1960 1989 B. fragilis [10]
Ampicillin 1961 1965 E. coli [4]
Trimethoprim 1961 1972 E. coli [11]
K. aerogenes
Fusidic acid 1962 1966 S. aureus [12]
Gentamicin 1964 1971 K. [13]
pneumoniae
Rifampin 1967 1998 P. fluorescens [14]
Nalidixic acid 1967 1998 K. [15]
Norfloxacin 1983 pneumoniae
Ciprofloxacin 1987
Lincomycin 1967 Previouslyb
Clindamycin 1970
Fosfomycin 1973 1980 S. [16]
marcescens
Cefoxitin 1977 1989 K. [17]
pneumoniae
Cefotaxime 1980 1983 K. [18]
pneumoniae
Imipenem 1985 1991 P. aeruginosa [19]
Mupirocina 1985 1987 S. aureus [20]
Quinupristin/ 1999 Previouslyb
Dalfopristin 1977 S. aureus [21]
Linezolid 2000 2006 S. aureus [22]
Retapamulina 2007 2006 S. aureus [22]
a
Topical
b
The plasmid-mediated Erm 23S rRNA methyltransferase that confers resistance to lincomycin,
clindamycin, and quinupristin as well as erythromycin was discovered before these non-macrolide
antibiotics were approved for clinical use
11 Transmissible Antibiotic Resistance 343

11.2 Molecular Mechanisms of Resistance

Transmissible resistance utilizes the same three mechanisms involved in chromo-

somally mediated resistance: target alteration (including both protection and
replacement), antibiotic modification, and drug exclusion (Table 11.2). For some
antimicrobial agents all three mechanisms are utilized, with sometimes more than
one determined by the same plasmid. Some resistances mechanisms are agent spe-
cific, other class specific, and a few affect susceptibility to several classes of antibi-
otics. A variety of antibiotic resistance gene databases are available on line and have
been critically reviewed [23].

11.2.1 Aminoglycoside Resistance

More than a hundred plasmid-mediated enzymes are known that modify aminogly-
cosides by transferring acetyl, phospho, or adenyl groups to essential –OH or –NH2
groups on the aminoglycoside core (Fig 11.1). Such modifications reduce the bind-
ing of the drug to its 16S rRNA target in the 30S ribosome. Two nomenclatures are
in use. In one three letters identify the enzyme activity (AAC, APH, ANT) followed
by the site of modification in parenthesis, a roman numeral for the resistance profile,
and a lower case letter as an individual identifier [25, 26]. In the other system, the
genes are labeled aac, aph, and aad followed by a letter or a number that identifies
the site of modification and another number as a unique identifier [2]. Thus,
aac(6′)-Ib and aacA4 represent the same gene for an aminoglycoside
N-acetyltransferase modifying position 6′ and giving a Ib pattern of resistance, a
potentially confusing situation. A table of equivalent nomenclature is available [25].
The aminoglycoside acetyltransferase group is the largest. There are four sub-
classes named according to the position of the amino group that is modified. AAC(1)
and AAC(3) target groups at positions 1 and 3 of the 2-deoxystreptamine ring, while
AAC(2′) and AAC(6′) target groups found at the 2′ or 6′ position of the
2,6-­dideoxy-2,6-diaminoglucose ring. The AAC(6′) group can be further subdi-
vided into type I giving resistance to amikacin but not gentamicin and type II con-
ferring resistance to gentamicin but not amikacin. Both AAC(6′) groups modify
tobramycin and netilmicin and have many subgroups. AAC(6′) also exists as a
bifunctional hybrid with APH(2″) in Gram-positive bacteria. The second largest is
the phosphotransferase group, which targets 3′, 4′, and 2″ hydroxyls and provides
resistance to kanamycin, to neomycin, and for some enzymes to additional amino-
glycosides including amikacin. APH(3″) and APH(6) that modify streptomycin also
belong to this group. The final group of nucleotidyltransferases attacks hydroxyls at
the 2″, 3″, 4′, 6, and 9 positions. Clinically important members include ANT(2″)
active on 4,6-disubstituted aminoglycosides with a deoxystreptamine ring including
gentamicins, kanamycins, sisomicin, and tobramycin, ANT (3″) modifying the 3″
position of streptomycin and the 9-hydroxyl group of spectinomycin, and ANT(4′)
344 G. A. Jacoby

Table 11.2 Representative transmissible resistance genes

Antimicrobial agent Modification Target alteration Efflux
Aminoglycoside aac(1), aac(2′), aac(3), aac(6′), armA, npmA,
aac(6′)-aph(2′), ant(2′), ant(3′), rmtA, rmtB,
ant (4′), ant(6), ant(9), aph(2′), rmtC, rmtD,
aph(3′), aph(3′), aph(4′), aph(6) rmtE, rmtF,
rmtG, rmtH
Bacitracin bcrAB
β-lactam blaTEM, blaSHV, blaCMY, blaCTX-M,
blaKPC, blaNDM, blaOXA, blaPER,
blaVEB, blaVIM, blaZ, many others
Colistin mcr
Fluoroquinolone aac(6′)-Ib-cr qnrA, qnrB, qepA, oqxAB
qnrC, qnrD,
qnrE, qnrS,
qnrVC
Fosfomycin fosA, fosB, fosX, fosK
Fusidic acid catI fusA, fusB
Glycopeptide vanA, vanB,
vanD, vanE,
vanG, vanL,
vanM, vanN
Macrolide/ ere(A), ere(B), vgb(A), vgb(B), erm(A), erm(B), lsa(B), lsa(E),
lincosamide/ lnu(A), lnu(B), lnu(C), lnu(D), erm(C), erm(F), mef(A), msr(A),
streptogramin B lnu(E), lnu(F), lnu(G), vat(A), erm(T), erm(Y), msr(E), vga(A),
(MLSa group) vat(B), vat(C), vat(D), vat(E), cfr(A), cfr(B) vga(B), vga(C),
vat(F), vat(G), vat(H), mph(A), cfr(C) vga(E), optrA
mph(B), mph(C), mph(E), mph(F),
mph(G)
Mupirocin mupA mupB
Nitrofurantoin oqxAB
Nitroimidazole nimA, nimB, nimC, nimD, nimE,
nimF, nimG, nimH, nimI, nimJ
Oxazolidinone cfr(A), cfr(B), optrA
cfr(C)
Phenicol catA, catB cfr(A), cfr(B), cmlA, cmlB1,
cfr(C) cmr, cmx, fexA,
fexB, floR,
optrA, oqxAB
Rifampin arr-2, arr-3, arr-4, arr-5, arr-7,
arr-8
Sulfonamide sul1, sul2, sul3
(continued)
11 Transmissible Antibiotic Resistance 345

Table 11.2 (continued)

Antimicrobial agent Modification Target alteration Efflux
Tetracycline tet(X) tet(M), tet(O), tet(A), tet(B),
tet(Q), tet(S), tet(C), tet(D),
tet(W), tet(36) tet(E), tet(G),
tet(H), tet(J)
tet(K), tet(L),
tet(39), tet(42)
Trimethoprim dfrA, dfrB, dfrD, oqxAB
dfrG, dfrK

Fig. 11.1 Kanamycin B AAC(6’)

and sites of modification
by representative
aminoglycoside ANT(4’) NH2
acetyltransferase (AAC),
AAC(3)
aminoglycoside HO O
nucleotidyltransferase HO
APH(3’) H2N H2N
(ANT), and
O
aminoglycoside HO NH2
phosphotransferase (APH) O
enzymes. (Adapted from
[24]) O OH
HO OH
NH2

ANT(2) APH(2)

which targets kanamycin A, B, and C, gentamicin A, amikacin, tobramycin, and

neomycin B and C [27]. ANT(6) adenylates the 6 position of streptomycin, while
ANT(9) attacks the 9-hydroxyl of spectinomycin. The detailed substrate spectra and
properties of these and other aminoglycoside-modifying enzymes have been cata-
loged in reviews [25, 26]. The resistance spectrum of clinically important
aminoglycoside-­modifying enzyme is shown in Table 4.6.
Target modification leading to high-level and broad aminoglycoside resistance
due to S-adenosyl-L-methionine-dependent methylation of 16S rRNA was reported
in 2003 [28]. Several gene families are involved, and either position 1405 or 1408
within the aminoacyl or A-site of 16 S rRNA can be modified. ArmA and RmtA,
RmtB1, RmtB2, RmtC, RmtD, RmtD2, RmtE, RmtF, RmtG, and RmtH methylate
guanine 1405 and confer high-level resistance to gentamicin, tobramycin, and ami-
kacin but not to streptomycin, neomycin, or apramycin [29]. NpmA methylates
adenine 1408 and provides resistance to all these aminoglycosides [30]. Initially
these methyltransferases were rare, but they have subsequently been found on plas-
mids linked to common ESBLs and carbapenemases and are becoming more
prevalent.
346 G. A. Jacoby

11.2.2 Bacitracin

Bacitracin acts by binding to and sequestering an undecaprenol pyrophosphate car-

rier in the cell membrane that transports peptidoglycan monomers for cell wall bio-
synthesis. During transport, undecaprenol monophosphate is phosphorylated to the
pyrophosphate, which must be converted back to the monophosphate to allow fur-
ther transport. Plasmid-mediated bacitracin resistance in E. faecalis is produced by
an ATP-binding cassette (ABC) transporter encoded by the bcrABD operon under
the control of bcrR, a membrane-bound bacitracin sensor that binds to the operon
[31]. BcrA is an ATP-binding domain that together with the BcrB membrane-­
spanning domain makes up the homodimeric bacitracin transporter. BcrD is an
undecaprenol kinase that converts undecaprenol to undecaprenol monophosphate
and could thus counter the block in monophosphate synthesis by bacitracin but
seems to play no role in resistance [5, 32]. bcrAB genes have been found in a variety
of enterococcal species. Some strains lack bcrR but still express high-level (MIC
>256 μg/ml) bacitracin resistance [32].

11.2.3 β-Lactam Resistance

The plasmid-mediated mechanism for β-lactam resistance is drug inactivation by

β-lactamase enzymes. More than a thousand individual β-lactamases have been
described and can be classified into classes A, B, C, and D by structure [33] or into
16 or more groups based on such functional properties as substrate spectrum and
response to inhibitors [34, 35]. Class A, C, and D β-lactamases have serine at the
active site and form acyl intermediates between the β-lactam and the serine hydroxyl
in the course of hydrolyzing the β-lactam ring. Class B enzymes are metallo-β-­
lactamases with one or two Zn++ ions required to activate water for a direct attack on
the β-lactam core.
The clinically important β-lactamases in Gram-positive organisms are found in
Staphylococcus spp. About 90% of S. aureus make the class A (group 2a) BlaZ (also
known as PC1) β-lactamase. Four immunotypes of the enzyme are known of which
at least three can be distinguished by substrate profile [36] and by sequence [37].
The type A staphylococcal enzyme has also been found occasionally in E. faecalis
on enterococcal-specific plasmids [38]. Unlike most plasmid-mediated β-lactamases,
blaZ is inducible.
In Gram-negative bacteria, the commonest β-lactamase is TEM-1, the first one
reported in 1965 [4]. Development of isoelectric focusing allowed β-lactamases to
be differentiated further. TEM-2, which has similar biochemical properties, differs
from TEM-1 by a single amino acid that changes the isoelectric point of the enzyme.
Advances in DNA technology made sequencing a prime mode of characterization.
Naturally occurring enzymes that vary from each other by from one amino acid to
about 10% of their sequence constitute a β-lactamase family, many of which are
11 Transmissible Antibiotic Resistance 347

described in an online database [39]. TEM-1 and TEM-2 belong to class A or group
2b as does SHV-1, another common plasmid-determined β-lactamase, differing in
sequence from TEM-1 by 36%. Group 2b enzymes hydrolyze penicillins such as
ampicillin and early cephalosporins such as cephalothin or cefazolin and are inhib-
ited by clavulanic acid, sulbactam, or tazobactam (Table 11.3).
Pharmaceutical chemists devised oxyimino-β-lactams such as cefotaxime,
ceftazidime, ceftriaxone, and the monobactam aztreonam that are effective against
TEM-1-, TEM-2-, or SHV-1-producing bacteria. Within a few years, transmissible
resistance to oxyimino-β-lactams appeared and proved to be due to mutations in
TEM-1, TEM-2, and SHV-1 that changed the charge and configuration of the active
site allowing access to β-lactams with the oxyimino side chain. Many such extended-­
spectrum β-lactamases or ESBLs in the TEM (TEM-3, TEM-10, TEM-12, TEM-26,
etc.) and SHV (SHV-2, SHV-4, SHV-5, etc.) families are known and belong to
group 2be. Group 2be also includes a large family of CTX-M enzymes and smaller
ones of PER and VEB β-lactamases.
Combinations of clavulanic acid, sulbactam, or tazobactam with β-lactamase
susceptible drugs were also developed. Inhibitor-resistant TEM and SHV varieties
followed (group 2br) such as TEM-30 or SHV-10 as well as a few enzymes that
combined the extended-spectrum and inhibitor-resistant phenotypes (group 2ber),
including TEM-50 and TEM-68. β-lactamases pay a price for broader spectrum or
inhibitor resistance in terms of efficiency, so that compensatory promoter mutations
increasing gene expression often accompany such mutations [40] as well as changes
in the bacterial host to decrease β-lactam accumulation.
Carbenicillin and ticarcillin were developed to treat infections caused by P. aeru-
ginosa. β-lactamases with high activity toward carbenicillin (group 2c) were found
and, since they were at first thought to be pseudomonas-specific, were given names
like PSE-1 (for Pseudomonas-specific enzyme) or, for carbenicillinase activity,
CARB-3. A related variety able to hydrolyze cefepime as well as carbenicillin
(group 2ce) has also evolved.
Cephamycins, such as cefoxitin and cefotetan, were also developed as agents
effective against class A β-lactamase-producing bacteria and others. Class C (group
1) enzymes can hydrolyze cephamycins, but these enzymes were determined by
chromosomal ampC genes and hence restricted to bacteria expressing them until
plasmid-mediated class 3 enzymes (CMY-1, MIR-1, ACT-1, DHA-1, FOX-1,
MOX-1, and more in each family) were discovered [41]. These β-lactamases could
also hydrolyze oxyimino-β-lactams and were naturally resistant to the first genera-
tion of β-lactamase inhibitors. Group 1e enzymes have short deletions or amino acid
substitutions in the R2 loop that enhance activity toward particular substrates, espe-
cially ceftazidime. With host porin and efflux pump changes to reduce drug accu-
mulation, AmpC enzymes can even provide carbapenem resistance [42].
Carbapenems such as imipenem, meropenem, doripenem, and ertapenem came
to be widely used against ESBL and AmpC β-lactamase-producing bacteria with the
all too predictable appearance of carbapenemases belonging to class B (group 3a:
VIM-1, IMP-1, NDM-1) and class A (group 2f: KPC-2, IMI-1) β-lactamases, which
have spread worldwide [43] (see Chaps. 10 and 12 for further details). In the United
 348

Table 11.3 Classification of transmissible β-lactamases

Inhibited by
Group Class Distinctive substrates Aa Bb Cc Examples
1 C Cephalosporins No Yes No ACC-1, ACT-1, DHA-1, FOX-1, CMY-1,
MIR-1, MOX-1
1e C Cephalosporins No Yes No CMY-10, CMY-19, CMY-37
2a A Penicillins Yes Yes No BlaZ (PC1)
2b A Penicillins, early cephalosporins Yes Yes No TEM-1, TEM-2, SHV-1
2be A Extended-­spectrum cephalosporins, monobactam Yes Yes No TEM-3, SHV-2, CTX-M-15, GES-1,
PER-1, VEB-1
2br A Penicillins, early cephalosporins No Yes No TEM-30, SHV-10
2ber A Extended-­spectrum cephalosporins, monobactam No Yes No TEM-50, TEM-68
2c A Carbenicillin Yes Yes No PSE-1, CARB-3
2ce A Carbenicillin, cefepime Yes Yes No RTG-4
2d D Oxacillin Variable Variable No OXA-1, OXA-10
2de D Extended-­spectrum cephalosporins Variable Variable No OXA-11, OXA-15
2df D Carbapenems Variable Variable No OXA-23, OXA-48
2f A Carbapenems No Yes No KPC-2, IMI-1
3a B Carbapenems No No Yes NDM-1, VIM-1, IMP-1
Modified from [35]
a
Clavulanic acid, sulbactam, or tazobactam
b
Avibactam and related diazabicyclooctane derivatives
c
EDTA
G. A. Jacoby
11 Transmissible Antibiotic Resistance 349

States, most carbapenem resistance in Enterobacteriaceae is due to KPC-­type

enzymes, which efficiently hydrolyze penicillins, oxyimino-, and other cephalospo-
rins including cephamycins and aztreonam as well as carbapenems. blaKPC is carried
by transposon Tn4401 allowing its spread to a variety of plasmid types many with
other resistance genes. Worldwide, the most successful carbapenemase is NDM-1,
which hydrolyzes all β-lactams except the monobactam aztreonam and for which no
clinically useful inhibitor is yet available. In A. baumannii, NDM-1 is carried by
composite transposon Tn125 with two copies of ISAba125 and downstream the
bleMBL gene for resistance to the antitumor agent bleomycin [44]. In
Enterobacteriaceae blaNDM-1 is located on plasmids belonging to several incompat-
ibility groups linked to single or truncated copies of ISAba125 and bleMBL (for fur-
ther details of the epidemiology of NDM spread see Chaps. 10 and 12).
OXA β-lactamases (class D, group 2d) were originally distinguished by their
ability to hydrolyze oxacillin or cloxacillin at a rate at least 50% of that for benzyl-
penicillin [34]. A few are susceptible to inhibition by clavulanic acid, sulbactam,
and tazobactam but most are resistant. Response to avibactam is also variable. Class
D β-lactamases are a large (more than 500) and diverse group, some members of
which are ESBLs able to attack oxyimino-β-lactams (group 2de: OXA-11, OXA-­
15) or even act as carbapenemases (group 2df: OXA-23, OXA-48) [45]. 2df enzymes
have been most frequently identified in Acinetobacter spp. but are becoming more
common in Enterobacteriaceae. Spread of OXA-48 is mainly due to a specific plas-
mid in which the blaOXA-48 gene is part of composite transposon Tn1999.
β-lactams and β-lactamases have thus evolved together with the introduction of
derivatives with increased enzymatic stability followed by the appearance of
β-lactamases with broader spectrum able to hydrolyze the new agents [46]. The
evolution is ongoing: TEM and SHV ESBLs have been largely replaced by ESBLs
in the CTX-M family, especially CTX-M-15 and CTX-M-14 [47]. The CTX-M suc-
cess story seems due not so much to superior properties of the enzyme as to capture
of the genes by mobile genetic elements and their association with plasmids carry-
ing other effective resistance determinants in highly successful sequence type
strains [48, 49].

11.2.4 Colistin Resistance

Long after colistin (polymyxin E) was first used clinically but within a few years of
its extensive use in animal husbandry, plasmid-mediated colistin resistance was
reported in 2016 in an isolate from an intensive pig farm in China. The plasmid gene
mcr-1 encodes a membrane-bound phosphoethanolamine transferase, a zinc metal-
loprotein, that modifies lipid A decreasing the binding affinity of polymyxins and
thus conferring a colistin MIC of 4–8 μg/ml [9]. Within a year of its discovery, mcr-­
1 was found on a wide variety of plasmids in a diversity of Enterobacteriaceae in
countries on five continents and in isolates from as early as the 1980s [50]. ISApl1
is often found upstream from the mcr-1 gene [51]. mcr-1 has been found in 10% of
350 G. A. Jacoby

NDM-1-producing Enterobacteriaceae in the United Kingdom and in KPC-­

producing pathogens in Italy and Greece so it is already compromising colistin use
as a last resort antibiotic. Several alleles, including mcr-1.2, mcr-1.6, mcr-2, and
mcr-3, have been reported [52].

11.2.5 Fluoroquinolone Resistance

Plasmid-mediated quinolone resistance was reported in 1998, 31 years after nali-

dixic acid began to be used clinically and 15 years after modern fluoroquinolones
were approved for use [15]. The first transmissible quinolone resistance was discov-
ered in a 1994 urinary isolate of K. pneumoniae from Alabama that could transfer
low-level ciprofloxacin resistance on a multiresistance plasmid to a variety of Gram-­
negative bacteria. In E. coli, the plasmid caused an 8-fold decrease in susceptibility
to nalidixic acid and a 30-fold decrease in susceptibility to all fluoroquinolones
tested. When the responsible gene, named qnr and later qnrA, was cloned and
sequenced facilitating its identification by PCR, qnrA was soon found at low fre-
quency on plasmids in Gram-negative isolates around the world with the strain car-
rying the earliest known plasmid-mediated qnr collected in 1988. Further searches
led to the discovery of plasmid-mediated qnrS, qnrB, qnrC, qnrD, and qnrE [53,
54]. The qnrVC gene from Vibrio cholerae can also be located on a plasmid or an
integrated conjugative element. These new qnr genes generally differed by 35% or
more from qnrA and each other. Allelic varieties that differ by 10% or less have
been described in almost all families; currently, there are 8 qnrA, 85 qnrB, 1 qnrC,
2 qnrD, 1 qnrE, 9 qnrS, and 6 qnrVC alleles. Qnr genes are also present on the
chromosome of a variety of Gram-negative and Gram-positive bacteria from both
clinical and environmental sources. Their utility prior to the clinical use of quino-
lones is not known.
Qnr proteins are composed of tandemly repeating five amino acid units. Two Qnr
molecules dimerize carboxyl terminus to carboxyl terminus and fold into a right-­
handed quadrilateral β helix with size, shape, and charge mimicking the B-form of
DNA. Loops of 8 and 12 amino acids project from their otherwise rod-like structure.
Deletion of the larger loop or even one amino acid in this loop compromises protec-
tive activity, suggesting that the loops are essential for proper positioning of Qnr on
topoisomerase.
In vitro purified Qnr proteins protect DNA gyrase or topoisomerase IV from
quinolone inhibition [55]. As the quinolone concentration is increased, more Qnr is
needed for protection suggesting that they both compete for a site on the topoisom-
erase. In a gel-displacement assay or bacterial two-hybrid system [56], Qnr binds to
both topoisomerase holoenzymes and their individual subunits. Subinhibitory con-
centrations of ciprofloxacin, however, reduce binding to GyrA but not to GyrB,
suggesting that Qnr protects gyrase by blocking quinolone access to GyrA sites
essential for its lethal action. The molecular details of this interaction are still being
debated [57].
11 Transmissible Antibiotic Resistance 351

Many naturally occurring antibiotics and synthetic agents target DNA gyrase
besides quinolones. Qnr protects against compounds with a somewhat quinolone-­
like structure, for example, 2-pyridone, quinazoline-2,4-dione, or spiropyrimidine-
trione, so it is not strictly quinolone specific. Qnr, however, does not block agents
acting on the GyrB subunit [58].
Expression of qnrB, qnrD, and qnrE is regulated by components of the bacterial
SOS system in which DNA damage induced by quinolones activates RecA protease
to cleave LexA protein which otherwise blocks expression by binding to specific
DNA sequence upstream from these qnr genes. Expression of qnrS1 is also induced
by quinolone and requires DNA sequence upstream from the gene, but qnrS1 regu-
lation is independent of the SOS system. Expression of qnrA in the aquatic organ-
ism Shewanella algae, on the other hand, is induced not by quinolone but by cold
shock.
Qnr plasmids have been found around the world in a variety of Enterobacteriaceae,
especially E. coli, E. cloacae, K. pneumoniae, and S. enterica but rarely in non-­
fermenting bacteria such as P. aeruginosa or A. baumannii. Plasmids carrying qnr
genes vary in size and incompatibility specificity, indicating that the spread of mul-
tiple plasmids has been responsible for their dissemination and that plasmid acquisi-
tion has occurred multiple times. A mobile or transposable element is almost
invariably associated with plasmid-mediated qnr genes, especially insertion
sequences ISCR1, ISExp1, and IS26. The complex is often inserted into a sul1-type
integron. qnrVC is so far the only qnr gene located in a cassette with a linked attC
site ready by itself for integron capture. Because of such linkage, qnr genes are
often found on plasmids with genes for other resistance determinants such as ESBLs
and carbapenemases. Qnr prevalence seems to be increasing and has reached as
high as 39% in a sample of E. cloacae isolates at one hospital in China [59].
The second plasmid-mediated quinolone resistance mechanism to be discovered
involves modification. AAC(6′)-Ib-cr is an acetyltransferase providing resistance to
kanamycin, tobramycin, and amikacin that by two amino acid substitutions has
broadened its spectrum to acetylate quinolones with an amino nitrogen on the piper-
azinyl ring, such as ciprofloxacin and norfloxacin [60]. Acetylation decreases the
antibacterial potency raising the ciprofloxacin MIC eightfold. Both the substitutions
Trp102Arg and Asp79Tyr are required for quinolone-acetylating activity. Seven
nucleotide variants of aac(6′)-Ib-cr are known, one with a 26 amino acid insertion
changing the position of the amino acid changes essential for anti-quinolone activ-
ity [61]. The aac(6’)-Ib-cr gene has been found worldwide in a variety of
Enterobacteriaceae (especially E. coli) and is often more common in surveys than
qnr alleles. It is usually found in a cassette as part of a class 1 integron in a multire-
sistance plasmid, which may contain other plasmid-mediated quinolone resistance
genes. Association with CTX-M-15 is particularly common.
The third mechanism of plasmid-mediated quinolone resistance involves efflux
pumps. QepA is a plasmid-acquired efflux pump in the major facilitator family that
decreases susceptibility to hydrophilic fluoroquinolones. qepA has often been found
on plasmids together with aminoglycoside ribosomal methyltransferase rmtB. At
least three qepA variants are known. OqxAB is an efflux pump in the
352 G. A. Jacoby

r­ esistance-­nodulation-­division family that was originally recognized as responsible

for resistance to olaquindox used for growth enhancement in pigs. OqxAB can
efflux other antimicrobial agents including chloramphenicol, trimethoprim, and
nitrofurantoin [62]. oqxAB has been found on plasmids in clinical isolates of E. coli
and K. pneumoniae and on plasmids (especially IncHI2 plasmids) and the chromo-
some of S. enteritis, flanked in both locations by IS26-like elements.
By themselves none of the plasmid-mediated quinolone resistance determinants
raises the MIC above the CLSI designated susceptibility breakpoint, but they raise
mutant prevention concentration (MPC) [63] as well as the MIC and facilitate the
selection of a variety of additional chromosomal mutations of higher MIC via acti-
vation of efflux pumps and altered membrane LPS [64].

11.2.6 Fosfomycin Resistance

Plasmid-mediated resistance to fosfomycin involves enzymes that open the oxirane

ring of the antibiotic inactivating it [65]. FosA, FosB, and FosX are members of the
same metalloenzyme superfamily but differ in their preferred cofactor. FosA utilizes
glutathione, while FosB requires bacillithiol or L-cysteine and FosX uses water.
Subtypes of each gene have been described: fosA1–6 and fosB1–6.
fosA has been reported on plasmids in Enterobacteriaceae, Pseudomonas spp.,
and Acinetobacter spp. fosB has only been described in Gram-positive strains: on
small plasmids in S. aureus and other staphylococcal species, on the chromosome
in B. anthracis, and on transferable circular intermediates in E. faecium [66]. fos-
XCC is transferrable between strains of Campylobacter but is located in a multi-
drug resistance genomic island rather than a plasmid [67]. Another allele, fosK,
has been found in an integron in a highly fosfomycin-resistant isolate of
Acinetobacter soli [68].

11.2.7 Fusidic Acid Resistance

Transmissible resistance to fusidic acid in staphylococci is produced by fusA and

fusB genes that encode proteins that bind and protect elongation factor G (EF-G),
the antibiotic target, from inhibition [69–71]. Related gene fusC and fusD are chro-
mosomal, and fusB can also be found on the chromosome in a genomic island [72].
In countries where fusidic acid is used to treat S. aureus infections, resistance is
slowly increasing [73], but it remains very low in US isolates [74].
In addition to the fus genes, some type A chloramphenicol acetyltransferases
bind fusidic acid conferring resistance. For example, in fusidate-sensitive E. coli
DB10, a cloned catI gene increased the chloramphenicol MIC from 2 to 50 μg/ml
and the Na fusidate MIC from 5 to 80 μg/ml [75]. The two structurally unrelated
antibiotics compete for binding to the enzyme at the same site [76].
11 Transmissible Antibiotic Resistance 353

11.2.8 Glycopeptide

The glycopeptide vancomycin, the lipoglycopeptide teicoplanin (approved for use

in Europe), and the semisynthetic lipoglycopeptides telavancin, oritavancin, and
dalbavancin bind to the D-alanyl-D-alanine (D-Ala-D-Ala) terminus of intermedi-
ates in peptidoglycan synthesis, inhibiting bacterial cell wall cross-linking.
Glycopeptide resistance involves bypass or target modification by enzymes that
substitute D-lactate (D-Lac) or D-serine (D-Ser) for the terminal D-alanine, thus
reducing glycopeptide binding. These enzymes act together with others that elimi-
nate the normal C-terminal D-alanine precursors [77]. Nine such gene clusters are
currently associated with vancomycin resistance in enterococci [78] and are distin-
guished by the sequence of the structural gene for the resistance ligase: vanA, vanB,
vanD, and vanM create a precursor ending in D-Ala-D-Lac with a 1000-fold
decrease in vancomycin binding, while vanC, vanE, vanG, vanL, and vanN produce
intermediates terminating in D-Ala-D-Ser with only a sevenfold binding decrement.
The vanC cluster is intrinsic to Enterococcus gallinarum, Enterococcus casselifla-
vus, and Enterococcus flavescens; the rest are acquired with vanA, vanB, vanD,
vanM, and vanN located either on plasmids or the chromosome and vanG, vanE,
and vanL only chromosomal. The vanA and vanD clusters provide high and moder-
ate levels of resistance, respectively, to both vancomycin and teicoplanin. vanB and
the van clusters producing D-Ala-D-Ser confer only vancomycin resistance.
Figure 11.2 details the mechanism of VanA-type resistance. The vanA cluster is
typically located in 11-kb transposon Tn1546 or a related element on transferable or
nontransferable plasmids or the host chromosome. It includes nine genes. Two
genes are involved in transposition. VanS and VanR encode a two-component regu-
latory system that modulates transcription of the cluster. VanS is a membrane-bound
kinase containing a histidine residue that is phosphorylated in the presence of gly-
copeptide. The phosphoryl group is transferred to an aspartate residue on VanR,
which can then activate transcription from promotors PH and PR upstream from
vanH and vanR [80]. The dehydrogenase VanH converts pyruvate to D-Lac, and the
ligase VanA combines D-Lac and D-Ala to make the depsipeptide D-Ala-D-Lac,
which is incorporated into peptidoglycan precursors in place of D-Ala-D-Ala. VanX
is a D,D-dipeptidase that hydrolyzes the dipeptide D-Ala-D-Ala formed by the
endogenous D-Ala-D-Ala:ligase thus reducing vancomycin susceptible targets.
VanY, not essential for resistance, is a D,D-carboxypeptidase that cleaves the termi-
nal D-Ala of any pentapeptide precursors synthesized from D-Ala-D-Ala that
escaped VanX hydrolysis. VanZ contributes to teicoplanin resistance but its function
is not known in detail.
The organization of the VanB cluster is similar to that of VanA but differs in its
regulation. While VanA strains are inducible to high levels by either vancomycin or
teicoplanin, VanB strains have variable levels of inducibility only by vancomycin.
The VanB operon is found in Tn1547 and other transposons and includes genes
encoding a dehydrogenase, a ligase, and a dipeptidase with 67–76% identity to the
corresponding VanA enzymes and a two-component vanRBSB set of regulatory genes
354 G. A. Jacoby

Ddl
a
L-Ala D-Ala futile D-Ala- D-Ala Penta Penta
cycle
racemase
VanX
adding VanY Tetra
Tri
enzyme

VanH VanA
Pyruvate D-Lac D-Ala-D-Lac Pentadepsi Pentadepsi

b regulation required for glycopeptide resistance accessory proteins

PR PH
VanR VanS VanH VanA VanX VanY VanZ

regulator sensor dehydrogenase ligase D,D-dipeptidase D,D-carboxy- unknown

peptidase TeR

Fig. 11.2 VanA-type glycopeptide resistance. a: Organization of the vanA operon in Tn1546.
Open arrows represent coding sequences and indicate the direction of transcription. IRL and IRR,
inverted repeat left and right, respectively. b: Synthesis of peptidoglycan precursors in a VanA-­
type-­resistant strain after induction with glycopeptide; Ddl, D-Ala:D-Ala ligase [79]

only distantly (34% and 24%) related to those of VanA. A vanW gene of unknown
function is included but no vanZ.
The VanC cluster produces peptidoglycan precursors ending in D-Ala-D-Ser and
is chromosomal and not transferable. VanT is a racemase converting L-Ser to D-Ser
which is joined to D-Ala by VanC ligase and incorporated into peptidoglycan pre-
cursors in place of D-Ala-D-Ala. The two enzymes of VanA or VanB that eliminate
D-Ala-D-Ala targets are combined in VanC in a single VanXYC D,D-peptidase.
Expression of vanC, vanXYC, and vanT is again controlled by a two-component
regulatory system VanRC and VanSC. The VanE cluster has a similar organization
and 41–60% identity to VanC [81].
VanA and VanB are the most frequent causes of vancomycin resistance in
Enterococcus spp. and have occasionally been detected in coryneform bacteria and
other streptococci. VanA has also appeared in S. aureus causing high-level resis-
tance but fortunately only rarely.
In a sample of enterococci causing bloodstream infections between 2001 and
2014 from the United States, the frequency of vancomycin resistance in E. faecalis
was 6.5% in 2008 declining to 1.9% in 2014 but reached 80.7% in E. faecium in
2010 and 68.4% in 2014. In Europe, the corresponding vancomycin resistance fre-
quencies were 2.6% in E. faecalis and peaked at 27.3% in E. faecium in 2012. All
three lipoglycopeptides are active against VanB enterococci, while oritavancin and
telavancin have activity as well against VanA strains but of unproven clinical signifi-
cance [82].
11 Transmissible Antibiotic Resistance 355

11.3  acrolide/Lincosamide/Streptogramin B (MLSB

M
Group)

Macrolides in clinical use include erythromycin and clarithromycin with

14-­membered lactone rings, azithromycin with a 15-membered ring, and josamy-
cin, spiramycin, and tylosin with 16-membered rings. Lincosamides include linco-
mycin and clindamycin, and for veterinary use pirlimycin. The streptogramin B
group includes pristinamycin, virginiamycin, and quinupristin, which is combined
with dalfopristin, a streptogramin A agent, in Synercid®. This group of antibiotics
is considered together because they share overlapping binding sites on the bacterial
50S ribosomal subunit and often have resistance mechanisms in common.
Inactivating enzymes, target-modifying rRNA methylases, and efflux pumps are all
involved. Coresistance with pleuromutilins (retapamulin, valnemulin, and tiamulin)
and phenicols may also be seen. Resistance to this group of antibiotics has recently
been reviewed [83].
Inactivation of these antimicrobials involves esterases, hydrolases, nucleotidyl-
transferases, acetyltransferases, and phosphotransferases that are more or less spe-
cific for structurally related antibiotics. ere(A) and ere(B) coding for erythromycin
esterases are found in transposons and integrons on plasmids in Gram-negative and
only occasionally in Gram-positive bacteria. Both inactivate drugs through hydroly-
sis of the macrolactone ring. Ere(A) is specific for erythromycin, while Ere(B) inac-
tivates azithromycin as well. Neither attacks the ketolide telithromycin or the
lincosamides [84]. vgb(A) and vgb(B) (virginiamycin B lyase) encode enzymes
catalyzing the linearization of cyclic streptogramin B-type antibiotics (such as qui-
nupristin) via a cleavage that requires a divalent metal ion [85]. Several varieties
have been found in staphylococci and enterococci. The lnu family encodes
lincosamide-­specific nucleotidyltransferase on plasmids and transposons in staphy-
lococci, streptococci, enterococci, anaerobes, Salmonella spp., and E. coli. Lnu(A)
modifies a hydroxyl group of clindamycin and lincomycin at positions 3 and 4,
respectively, while Lnu(B) attacks a hydroxyl at position 3 in both clindamycin and
lincomycin, although only resistance to lincomycin is observed with both enzymes
in Gram-positive organisms [86]. Nucleotidyltransferases Lnu(C), Lnu(D), Lnu(E),
Lnu(F), and Lnu(G) differing in size and origin have also been described. The vat
family encodes acetyltransferases on plasmids in Staphylococcus and Enterococcus
spp. that inactivate streptogramin A-type antibiotics (such as dalfopristin). Vat(A),
Vat(B), Vat(C), Vat(D), Vat(E), Vat(F), Vat(G), and Vat(H) are currently known. mph
genes encode macrolide-2′-phosphotransferases that preferentially inactivate 14-
and 15-membered macrolides [mph(A)] or 14- and 16-membered ones [mph(B)]
[87]. mph(A) and mph(B) have been found in Gram-negative organisms, while
mph(C) has been described in staphylococci, corynebacteria, and Stenotrophomonas
maltophilia. Plasmid-mediated mph(E), mph(F), and mph(G) have also been
reported.
The MLSB antibiotics bind to 23S rRNA of the 50S ribosomal subunit at the
peptidyltransferase center blocking protein synthesis. More than 40 erm methylase
356 G. A. Jacoby

genes have been distinguished based on >20% difference in sequence. Many are
determined by plasmids, transposons, or integrative and conjugative elements [88].
Erm proteins add one or two methyl groups to adenine 2058 in domain V of 23S
rRNA preventing MLSB antibiotic attachment. Resistance is produced to 14-, 15-,
and 16-membered macrolides, ketolides, lincosamides, and streptogramin B antibi-
otics. erm(A), erm(B), and erm(C) are typically found in staphylococci. erm(B) and
a subclass of erm(A) [erm(TR)] are widespread in enterococci and streptococci.
erm(F) has been found in anaerobes and H. influenzae. erm(A) is part of transposon
Tn554 or its close relative Tn6133, while erm(B) is part of transposons Tn917 and
Tn551. erm(C) is often located on small plasmids in staphylococci and erm(T) on
larger ones. erm(33) is the result of in vivo recombination between erm(A) and
erm(C). erm expression may be inducible or constitutive. Erythromycin and other
14- and 15-membered macrolides tend to be good inducers via a mechanism that
involves ribosome stalling while translating an upstream leader peptide with conse-
quent changes in the structure of erm mRNA that allows it to be translated. Ketolides
and lincosamides are usually not inducers but may become so by deletions, inser-
tions, and point mutations in this attenuator system [87, 89, 90]. Hence a staphylo-
coccal strain with erm(A) or erm(C) may appear erythromycin resistant but
clindamycin susceptible, but if exposed to clindamycin, it can mutate to resistance
to both agents [91].
A different methyl transferase is encoded by the cfr gene and confers resistance
to lincosamides, streptogramins A, phenicols, oxazolidinones, and pleuromutilins
and decreased susceptibility to such 16-membered macrolides as josamycin and
spiramycin. It adds a methyl group from S-adenosyl-L-methionine to the C8 posi-
tion of adenine 2503 at the peptidyltransferase center in domain V of 23 rRNA by a
two-­step mechanism involving intermediate methylation of a Cys residue on the
enzyme [92]. The cfr gene has been found worldwide in Staphylococcus spp.,
Enterococcus spp., other Gram-positive organisms, and P. vulgaris and E. coli on
plasmids or together with insertion sequences. Cfr(B) with 74.9% amino acid iden-
tity to Cfr(A) has been described in E. faecium [93] and Cfr(C) with 55.1% identity
in Campylobacter spp. [94].
Plasmid-mediated efflux genes are also involved in MLSB resistance. Msr(A) in
the ABC transporter family confers resistance to 14- and 15-membered macrolides,
and streptogramin B antibiotics and low-level resistance to ketolides, while Mef(A)
in the major facilitator superfamily provides resistance to most 14- and 15-­membered
macrolides but not 16-membered macrolides, lincosamides, or streptogramin B
[87]. Msr(A) has mainly been found in Staphylococcus spp. but also in Streptococci,
Corynebacterium, and Pseudomonas [95]. Mef(A) has been detected in strepto-
cocci including pneumococci and Group A and D organisms and also in Gram-­
negative bacteria. Plasmid- or transposon-borne vga(A), vga(AV), vga(A)LC, vga(B),
vga(C), vga(E), and vga(E)V encode ABC transporters that export streptogramin A
antibiotics, while vga(A), vga(C), and vga(E) export lincosamides and pleuromuti-
lins as well. lsa(B) found on a plasmid in S. sciuri encodes an ABC transporter
active on clindamycin but probably not streptogramins [96]. lsa(E) on the other
hand confers resistance to lincosamides, streptogramins A, and pleuromutilins and
11 Transmissible Antibiotic Resistance 357

has been found in S. aureus and several species of Enterococcus [97]. OptrA in the
ABC transporter family confers resistance to oxazolidinones and phenicols and has
been found in E. faecium, E. faecalis, Staphylococcus sciuri, and Streptococcus suis
[98, 99]. Insertion sequence IS1216E has been implicated in the spread of optrA
among enterococcal plasmids and to the streptococcal chromosome [100, 101].

11.3.1 Mupirocin

High-level resistance to the topical anti-staphylococcal agent mupirocin involves

mup genes determining mupirocin-resistant isoleucyl-tRNA synthetases [102].
mupA (also known as ileS2) is determined by readily transmissible plasmids, while
the 65.5% identical mupB (ileS3) gene is located on a nonconjugative plasmid in the
single strain studied [103]. In a recent investigation of 358 S. aureus isolates cul-
tured from children attending a Dermatology Clinic in New York City, 96 of 112
mupirocin-resistant isolates had high-level resistance typical of the plasmid-­
determined mechanism [104].

11.3.2 Nitrofuran

Nitrofurantoin resistance transferable from clinical E. coli to laboratory strains of E.

coli was reported in 1983. The nitrofurantoin MIC rose from 5 to 50–70 μg/ml.
Plasmids were not demonstrated physically, and the mechanism of resistance was
not established, but resistance was cured by rifampin treatment and transmissible by
conjugation. More than 30 years later, plasmids discovered for resistance to olaqui-
ndox and later fluoroquinolone were found also to confer nitrofurantoin resistance
of the same degree via the resistance-nodulation-division family efflux pump
OqxAB [62].

11.3.3 Nitroimidazole

Plasmid and chromosomally located nim genes (A through J) [105, 106] encode
nitroimidazole reductases that convert 5-nitroimidazole to 5-aminoimidazole thus
blocking formation of toxic nitroso derivatives that are essential for bactericidal
activity by metronidazole and tinidazole [107]. The nim genes are usually tran-
scribed from promotors located within different insertion elements: IS1168 for
nimA-nimB, IS1169 for nimD, and IS1170 for nimC. nim plasmids characterized in
Bacteroides spp. have been nonconjugative (7.2 to 10–kb in size) but mobilizable by
larger plasmids or transferable by electroporation [10, 108]. Metronidazole resis-
tance in the B. fragilis group has been quite rare in the United States [109].
358 G. A. Jacoby

11.3.4 Oxazolidinone

Linezolid targets the peptidyltransferase center of the bacterial 50S ribosomal sub-
unit and, like other drugs with the same target, is blocked by the Cfr 23S rRNA
methyltransferase with an increase in the MIC of S. aureus from 0.5 to 8–16 μg/ml
[22]. The plasmid-mediated ABC transporter OptrA also exports linezolid with
typical MIC increases in S. aureus or enterococci of 2 to 8–16 μg/ml. Susceptibility
of tedizolid is affected to a lesser extent by OptrA [98] and not at all by Cfr [110].

11.3.5 Phenicol

Chloramphenicol resistance is most often due to chloramphenicol acetyltransferase

(CAT), which transfers an acetyl group from acetyl-CoA to the C3 position of the
antibiotic. The acetyl groups then shifts to the C1 position making chloramphenicol
available for diacetylation. The fluorine group in florfenicol (licensed for use only
in animals) occupies the C3 position making florfenicol resistant to inactivation by
CAT. There are two main types of transmissible CAT enzymes with many sub-
groups [111]. They are found in plasmids, transposons, integrons, and integrative
and conjugative elements in both Gram-negative and Gram-positive pathogens.
Some cat genes on plasmids in S. aureus are induced by chloramphenicol via
upstream translation attenuators but most are expressed constitutively. Both CAT
types have a trimeric structure composed of three identical monomers.
The cfr gene was first recognized by its production of combined chlorampheni-
col and florfenicol resistance in S. sciuri [112] and later appreciated to provide
resistance to lincosamides, streptogramins A, oxazolidinones, pleuromutilins, and
16-membered macrolides as well as via methylation of 23S rRNA at the peptidyl-
transferase center.
In addition to drug and target modification, a number of plasmid and transposon-­
mediated phenicol exporters have been described. They belong to the major facilita-
tor superfamily and include chloramphenicol-specific cmr, cmx, cmlA, and cmlB1
genes and floR and floRV genes that export florfenicol as well [113]. cmlA and cmlB1
are expressed inducibly via translational attenuation. In Gram-positive organisms,
fexA and fexB encode chloramphenicol/florfenicol exporters that are plasmid-­
mediated. Both phenicols are also exported by the OptrA pump that also transports
oxazolidinones [98], and at least chloramphenicol is effluxed by the multidrug
OqxAB pump [114].
11 Transmissible Antibiotic Resistance 359

11.3.6 Rifamycin

In 1998 transfer of rifampin resistance on a multiresistance plasmid from P. fluore-

scens to E. coli or P. putida was reported. Accumulation of rifampin was blocked by
the plasmid and relieved by the energy uncoupler potassium cyanide, suggesting
that an efflux pump was involved [14]. The gene was not named or sequenced nor
have subsequent reports elaborated on the evidence for its nature.
The next year, a gene was found in a class 1 integron in P. aeruginosa related to
the rifampin ADP-ribosylating transferase responsible for rifampin resistance in
Mycobacterium smegmatis [115]. It was named arr-2 and has been subsequently
found in integrons on plasmids in K. pneumoniae, E. coli, and species of Enterobacter
and Acinetobacter. Additional alleles arr-3, arr-4, arr-5, arr-7, and arr-8 have been
reported in integrons, some associated with carbapenemases KPC-2 or NDM-1
[116, 117].

11.3.7 Sulfonamide

Plasmid-mediated sulfonamide resistance adopts the simple solution of providing a

resistant dihydropteroate synthase to substitute for this usually sulfonamide-­
sensitive enzyme in the pathway to folic acid. Sulfonamides are structural analogs
of p-aminobenzoic acid with which they compete in the synthesis of dihydropteroic
acid. The resistant enzymes efficiently distinguish between its normal substrate
dihydropteroic acid and the inhibitor. There are three sul genes encoding this resis-
tance mechanism on plasmids in Gram-negative organisms: sul1 usually found with
other resistances in a Tn21-type integron, sul2 found on small plasmids in the IncQ
or pBP1 families or larger conjugative plasmids of several Inc groups, and, least
common, sul3 located in a composite transposon [118].

11.3.8 Tetracycline

More than 60 genes conferring resistance to tetracycline are known, most associated
with mobile elements that allow for gene exchange. Most common are genes encod-
ing energy-dependent efflux proteins. Others code for ribosomal protection proteins
or inactivating enzymes. Further details can be found in the web site maintained by
M. Roberts [88].
tet(X) encoding a NADP-dependent monooxygenase that requires oxygen to
degrade tetracycline was originally discovered as part of a conjugative transposon
in Bacteroides sp. where, lacking oxygen, it does not confer tetracycline resistance
on its host. Subsequently, it has been found in E. cloacae, K. pneumoniae, and other
360 G. A. Jacoby

Enterobacteriaceae [119] and deserves attention since it inactivates all tetracyclines

including tigecycline, a derivative designed to overcome resistance.
Ribosomal protection involves proteins that displace tetracycline from its ribo-
somal binding site allowing protein synthesis to proceed [120]. Further details of
the mechanism can be found in Sect. 4.2.4.2. Resistance is conferred to tetracycline,
doxycycline, and minocycline but not to tigecycline. The ribosomal protection pro-
teins have sequence similarity to ribosomal elongation factors EF-G and EF-Tu and
like them are GTPases. The tet(M), tet(O), tet(Q), tet(S), tet(W), and tet(36) ribo-
somal protection genes have been found in both Gram-negative and Gram-positive
organisms. Tet(M), tet(Q), and tet(W) are usually associated with conjugative trans-
posons, while tet(O) and tet(S) have been found on conjugative and nonconjugative
plasmids. A subgroup of ribosomal protection genes are mosaics, made up of seg-
ments of two or three different known tet genes [121].
The tet efflux genes belong to the major facilitator superfamily and encode
membrane-­associated proteins that exchange a proton for the tetracycline cation
thus reducing the intracellular concentration of the antibiotic. Most export tetracy-
cline and doxycycline but not minocycline or tigecycline. Tet(B), however, exports
minocycline as well. tet(A) and presumably most other efflux genes are regulated
by a divergently transcribed repressor gene that produces a protein that binds to the
tet operator. Tetracycline complexed with Mg++ binds to the repressor spreading its
binding domains apart so that they no longer interact with the operator thus allow-
ing transcription to take place [122]. Widely distributed tet efflux genes include
tet(A), tet(B), tet(C), tet(D), tet(E), tet(G), tet(H), and tet(J) found in Gram-negative
bacteria and tet(K), tet(L), tet(39), and tet(42) found in both Gram-negative and
Gram-positive organisms. Some may be found integrated into the host chromosome
as well as on plasmids.
Overexpression of plasmid-mediated tet(M) ribosomal protection protein
and tet(L) encoded efflux has been associated with tigecycline resistance in E.
faecium [123].

11.3.9 Trimethoprim

Like that for sulfonamide, the strategy for plasmid-mediated trimethoprim resis-
tance is a resistant substitute for the trimethoprim target, dihydrofolate reductase. A
genetically diverse set of more than 30 dfr genes are known, mostly located in
Gram-negative organisms in integron cassettes or associated with ISCR elements.
The dfrA genes encode dimeric dihydrofolate reductases and include at least 26
alleles with dfrA1 and dfrA17 the most common [124]. dfrB genes encode smaller
trimeric enzymes of seven varieties [125, 126]. In addition, several dfr genes confer-
ring high-level trimethoprim resistance are known in Gram-positive organisms:
dfrA located on transposon Tn4003 in S. aureus and other Staphylococcus spp.,
dfrD found on small plasmids in staphylococci and Listeria monocytogenes [127],
11 Transmissible Antibiotic Resistance 361

dfrG which seems to be the most common variety in S. aureus [128], and dfrK found
on small plasmids in species of Staphylococcus [129].
The multidrug OqxAB pump effluxes trimethoprim [114].

11.4 Source of Resistance Genes

Resistance genes existed long before the antibiotic era and have been found, for
example, in ancient permafrost, at the bottom of isolated caves, and in the gut
microbiome of a pre-Columbian mummy [130]. An origin in the organisms that
produce antibiotics with hence a need to be protected from their action is plausible
[131, 132]. Alternatively, organisms living in the same environment as antibiotic
producers, often soil, also need resistance genes to be able to compete. In other
cases, a housekeeping gene playing no apparent role in antibiotic production or
defense could have been adapted to this new use [133]. Sophisticated metagenomic
studies have found sequences in DNA from soil, oceans, and human feces 100%
identical to genes for resistance to aminoglycosides, β-lactams, glycopeptides,
phenicols, tetracycline, and other agents supporting an environmental source for
exchange of resistance genes, although the direction of this transfer is not always
obvious [134, 135].
Some examples of potential sources are listed in Table 11.4. Streptomyces gri-
seus, producer of streptomycin, makes phosphotransferases modifying the same-
­OH groups as APH enzymes in pathogens. Streptomyces fradiae, producer of
neomycin, has acetyltransferases with the same specificity as the plasmid-mediated
AACs. Micromonospora purpurea, producer of gentamicin, has methylases that
modify 16S rRNA like ArmA and Rmt enzymes. Streptoalloteichus tenebrarius,
producer of tobramycin, has an rRNA methylase similar to the acquired NpmA
methylase. Bacillus licheniformis, producer of bacitracin, protects itself with a
BcrABC transporter similar to that determined by plasmids in E. faecalis.
Streptococcus (Saccharopolyspora) erythreus, producer of erythromycin, has 23S
rRNA methylases similar to acquired Erm methylases. Amycolatopsis orientalis
and Streptomyces toyocaensis, glycopeptide producers, have Van-like systems for
self-protection. Pseudomonas fluorescens, producer of mupirocin, has a resistant
isoleucyl-tRNA synthetase like the acquired Mup enzyme. Streptomyces rimosus,
an oxytetracycline producer, has ribosome protecting Tet(M) and Tet(O)-like pro-
teins. In each case, although the mechanisms are the same, the amino acid identity
between producer and plasmid-mediated resistance protein is too low to accommo-
date a direct transfer. Both could have a common ancestor, but convergent evolution
has not been ruled out.
Candidates with much closer sequence identity have been found for other resis-
tance genes. AAC(6′)-Ih, so far found only on plasmids in Acinetobacter spp., and
the more broadly distributed APH(3′)-VI are 99–100% identical to chromosomal
enzymes from particular species of Acinetobacter. QnrA, QnrC, and QnrS have
97–99% identical analogues in aquatic bacteria such as Shewanella and Vibrio spp.,
362 G. A. Jacoby

Table 11.4 Suggested sources of transmissible resistance genes

Antibiotic Resistance gene Representative source Identity %a Reference
Aminoglycoside aph(3″) Streptomyces griseus 50 [132]
aph(6) Streptomyces griseus 34 [132]
aph(3′)-VI Acinetobacter guillouiae 99 [136]
aac(3)-IIa Streptomyces fradiae 37 [132]
aac(6′)-Ih Acinetobacter gyllenbergii 100 [137]
armA Micromonospora purpurea 27
rmt Micromonospora purpurea 33
npmA Streptoalloteichus tenebrarius 28
Bacitracin bcrA Bacillus licheniformis 52 [5]
β-lactam blaSHV Klebsiella pneumoniae 100 [138]
blaCTX-M Kluyvera spp. 99 [139]
blaOXA-48 Shewanella oneidensis 92 [140]
blaOXA-51 Acinetobacter baumannii 97 [141]
blaACT-1 Enterobacter asburiae 98 [142]
blaCMY-1 Aeromonas sp. 95 [143]
blaCMY-2 Citrobacter freundii 96 [144]
blaMIR-1 Enterobacter cloacae 99 [145]
blaMOX-1 Aeromonas hydrophila 94 [146]
blaFOX-1 Aeromonas caviae 99 [147]
blaDHA-1 Morganella morganii 99 [148]
blaACC-1 Hafnia alvei 99 [149]
blaKPC Chromobacterium piscinae 76 [150]
blaNDM-1 Erythrobacter litoralis 55 [151]
Chloramphenicol catA Streptomyces albus 36 [152]
cfr(A) Bacillus amyloliquefaciens 74 [153]
Colistin mcr-1 Moraxella porci 63 [154]
Fluoroquinolone qnrA1 Shewanella algae 99 [155]
qnrB Citrobacter freundii complex 99 [156]
qnrC Vibrio parahaemolyticus 97 [157]
qnrE Enterobacter spp. 96 [158]
qnrS1 Vibrio parahaemolyticus 97 [157]
oqxAB Klebsiella pneumoniae 100 [159]
qepA Pseudorhodoferax sp. 84
Fosfomycin fosA3 Klebsiella pneumoniae 80 [160]
fosA6 Klebsiella pneumoniae 99 [161]
fosC2 Achromobacter xylosoxidans 28 [160]
Erythromycin ermA Streptococcus erythreus 21 [162]
ermB Streptococcus erythreus 24 [162]
ermC Streptococcus erythreus 24 [162]
Glycopeptide vanA Amycolatopsis orientalis 62 [163]
Mupirocin mup Pseudomonas fluorescens 35 [164]

(continued)
11 Transmissible Antibiotic Resistance 363

Table 11.4 (continued)

Antibiotic Resistance gene Representative source Identity %a Reference
Rifampin arr Mycobacterium sp. 63
Tetracycline tet(M) Streptomyces rimosus 33 [132]
tet(O) Streptomyces rimosus 36 [120]
Calculated from data available in GenBank May 2017
a

while QnrB originates from Citrobacter and QnrE from Enterobacter spp. The
QepA efflux pump is related to ones in the order Burkholderiales such as species of
Pseudorhodoferax, while the OqxAB pump has close relatives in K. pneumoniae,
an organism that is also the likely source of fosA genes. Plasmid-mediated fosA3,
fosA5, and fosA6 are surrounded by truncated genes that also delimit the fosA gene
on the chromosome of strains of K. pneumoniae.
Many β-lactamases also have a clear pedigree. The origin of TEM-1 is not
known, but blaSHV-1 is a chromosomal as well as a plasmid gene in K. pneumoniae
and has been mobilized onto plasmids at least twice [165]. Close homologues of
blaCTX-M genes can be found on the chromosome of rarely pathogenic Kluyvera
species with blaCTX-M groups 1 and 2 related to genes of K. ascorbata; blaCTX-M
groups 8, 9, and 25 related to genes of K. georgiana; and blaCTX-M-37 related to
genes of K. cryocrescens [48]. Several plasmid-mediated OXA-type carbapene-
mases are close enough in sequence to chromosomal genes in Acinetobacter or
Shewanella spp. to make them likely progenitors. Plasmid-mediated AmpC-type
β-lactamases have close homologues in chromosomally determined enzymes of
various species. Enzymes in Chromobacterium spp. are as much as 76% identical
in amino acid sequence to KPC β-lactamase. NDM β-lactamase appears to be a
chimera formed, probably in A. baumannii [166], between the aminoglycoside
resistance gene aphA6 and a metallo-β-lactamase such as ElBla2 from Erythrobacter
litoralis [151].
Aminoglycoside nucleotidyltransferases are missing among aminoglycoside pro-
ducers. Several ANTs, however, share structural similarity and catalytic mechanism
with housekeeping enzymes such as DNA polymerase β, which has a similar rela-
tionship with lincosamide nucleotidyltransferases LnuA and LinB [167].
Chloramphenicol acetyl transferase has been found in species of Streptomyces, such
as S. albus, but not in Streptomyces venezuelae, the organism known to produce it
[152]. The cfr methyltransferase gene has homologues in Bacillus spp. Moraxella
spp. contain chromosomal mcr-like genes and also ISApl1 that is often associated
with them [154]. Species of Aeromonas have also been suggested for the origin of
mcr-like genes [52]. Mycobacterium sp. has a rifampin ribosyltransferase 63% iden-
tical to the plasmid Arr-2 enzyme. The origin of acquired sul and dfr genes is not
known.
364 G. A. Jacoby

11.5 Plasmids and Other Mobile Genetic Elements

Plasmids vary in size from a few to more than 500-kb. Core plasmid functions
include systems for maintenance (replication, stability, and copy control), for parti-
tioning between daughter cells at the time of bacterial division, and for mobility
(mobilizability and conjugal transfer). Small plasmids usually exist in multiple cop-
ies within the cell, while replication of larger ones is limited to a few copies.
Plasmids in Gram-negative organisms smaller than about 25-kb lack space for the
genetic machinery involved in mating pair formation but may be mobilized by a
conjugative helper plasmid. In Gram-positive organisms, many plasmids rely on
chemical signals mediated by oligopeptides for mating pair formation [168]. A third
group of plasmids, found across the size spectrum, is neither conjugative nor mobi-
lizable and is thought to rely on transformation or transduction for transfer [169].
Several plasmid classification schemes have been developed but can be compro-
mised by plasmid plasticity and recombination [170]. Historically, plasmids were
classified into Inc or incompatibility groups based on whether two plasmids were
unable to coexist stably in the same bacterial host, a property based on replication
specificity and copy control. Inc grouping is now tested with specific primers by
PCR-based replicon typing. In Enterobacteriaceae as of 2014, PCR-based replicon
typing could identify 24 distinct plasmid replicons with IncFII, IncA/C, IncL/M, and
IncI1 being the most common groups among typed resistance plasmids [171, 172].
Since the system was based on established Inc groups, it relates directly to the older
classifications. In Acinetobacter baumannii, plasmids have been subdivided into 19
GR types based on replicon sequences [173], and in Enterococcus and Staphylococcus
spp., more than 25 rep families have been defined [174, 175]. Alternatively, MOB
classification is based on variations in relaxase, an enzyme in the plasmid mobiliza-
tion system that nicks DNA at a specific site to produce a single-­stranded substrate
for transfer. Use of degenerate primers recognizing the conserved N-terminal portion
of the relaxase gene allows five MOB types to be distinguished for plasmids of
γ-Proteobacteria [176]. Advantages of the MOB scheme include broad applicability
to plasmids of Acinetobacter and Pseudomonas spp. as well as Enterobacteriaceae,
and inclusion as well of integrative conjugative or mobilizable elements (ICE and
IME). A disadvantage is the limited resolution inherent in the current number of
MOB types. For some plasmid groups, a multilocus sequence typing (pMLST) sys-
tem based on 2–6 core plasmid genes is available for subtyping [172, 177]. With
neither replicon nor MOB typing, however, can all plasmids be classified at present,
so further evolution of plasmid taxonomy can be anticipated.
Plasmids vary in host range, due mainly to specificity of replication rather than
requirements of the conjugative system itself. In liquid culture, little if any mating
occurs between Gram-negative and Gram-positive bacteria or between strict anaer-
obes and facultative organisms. Plasmids found in Enterobacteriaceae are usually
transferable within that family, but only those belonging to a few Inc groups are
transferable to P. aeruginosa, which has its own set of plasmids transferable to other
Pseudomonas spp. [178]. Similarly, among plasmids in Gram-positive organisms,
11 Transmissible Antibiotic Resistance 365

some are species specific, while others have a broad host range and can be found in
both Enterococcus and Staphylococcus spp. [174].
Plasmids carry accessory functions besides antimicrobial resistance such as met-
abolic pathways, colonization and virulence factors, sex factor activity, bacteriocin
production and resistance, restriction/modification systems, biocide resistance, and
heavy metal ion resistance. Hughes and Datta examined over 400 enterobacterial
isolates collected in the “pre-antibiotic era” between 1917 and 1954 and found that
24% contained conjugative plasmids, many in the same Inc groups as contemporary
resistance plasmids, but none carried antibiotic resistance genes [179, 180]. How
then did naked plasmids and other mobile genetic elements acquire the genes for
resistance?
Figure 11.3 shows tools that bacteria have used to capture genes and incorporate
them into plasmids [181]. An insertion sequence (IS) is a 700 to 2500-bp DNA seg-
ment usually bounded by short, identical, sometimes imperfect inverted repeats (IRL
and IRR) and containing one or two transposase (tnp) genes that code for enzymes
that recognize the IRs and catalyze movement to another DNA site where integra-
tion generates direct repeats of 2 to14-bp depending on the IS [182]. As originally
defined, classical IS did not carry resistance genes but may locate to provide an
active promoter to activate an adjacent gene. Two copies of the same IS or related
ones can, however, surround a resistance gene creating a composite transposon that
can now move as a unit to another plasmid or chromosomal location. In particular,
820-bp IS26 is very common in multiresistance regions of plasmids and is a fre-
quent flanking element in composite transposons.
A few IS are unusual in that a single copy of the element can capture and move
an adjacent resistance gene. For example, 1656-bp ISEcp1 is bounded by 14-bp IRs
but on moving can utilize IRL and a new IRR distal to an adjacent gene which con-
sequently becomes part of the mobile unit. ISEcp1 has been implicated in the mobi-
lization of blaCTX-M, qnr, rmt, and other resistance genes [183]. ISCR elements differ
in moving by rolling circle replication and can incorporate larger segments of DNA
than ISEcp1. They are bounded not by IRs but by a downstream origin (oriIS) and
an upstream terminus (terIS) and do not create DR. Failure to recognize terIS allows
replication to continue into an adjacent gene which is thus mobilized. More than 20
ISCR elements have been distinguished based on the sequence of their transposases.
ISCR have been involved in the mobilization of virtually every class of antibiotic
resistance genes in Gram-negative organisms [184] .
An integron is an even more sophisticated system for capturing resistance genes
packaged in cassettes. A cassette contains the gene, often preceded by a ribosome
binding site but usually not a promoter, and an attC recombination site. The inte-
gron is made up of an intI gene encoding an integrase of the tyrosine recombinase
family, an attI recombination site, and a Pc promoter. The integrase catalyzes site-­
specific recombination between the attI and attC sites capturing or releasing gene
cassettes which can be lined up in tandem, all under the control of the Pc promoter.
One hundred or so different cassettes are known carrying antibiotic resistance genes
and three main groups of integrons, classified on the basis of intI sequences. Class
1 integrons are the most common [185, 186].
366 G. A. Jacoby

IRL IRR
Insertion
(DR) tnp (DR) (DR) tnp tnp (DR)
sequence
Composite transposon
IRL IRR IRalt
DR DR
ISEcp1 (5) tnp (5) Terminal inverted repeat (IR)

Transposition unit Antibiotic resistance gene

ter lS ori lS
ISCR rcr

Gene
cassettes attC attC
attl
Integron
intl Cassette array
IRtnp IR
res
Tn3-subgroup DR DR
(5) tnpA tnpR (5)
transposon

IRtnp IR
res
Tn21-subgroup DR DR
tnpA tnpR mer or
transposon (5) (5)

IR IRt
DR
res
MIC in parallel DR
with ISCR (5) tniR tniQ tniB tniA (5)
intl1/attlt1
or mer

Fig. 11.3 Mobile elements involved in the capture and mobilization of antibiotic resistance genes
in Gram-negative bacteria. DR, direct repeat; tnp, tni, transposition functions; IRL and IRR, left and
right inverted repeats; IRalt, alternative IR; rcr, rolling circle replicase; oriIS, origin of ISCR ele-
ments; terIS, terminus of ISCR elements; attC, cassette recombination site; attI, integron recombi-
nation site; res, resolvase site. Elements that create DR are indicated and the DR length given,
except for IS, where the DR length varies for different elements. Tn21-subfamily transposons may
carry resistance genes as part of class 1 integrons inserted in or near the res site. MIC mobile inser-
tion cassette. (Adapted from [181])

The first moveable units on plasmids to be described were complex or unit trans-
posons, such as 4957-bp Tn3 encoding TEM-1 β-lactamase. Members of the Tn3
family are bigger than IS and include a transposase gene (tnpA), a resolvase gene
(tnpR), and a resolution (res) site as well as one or more resistance genes all bounded
by 38-bp IR. Movement is replicative and involves formation of a cointegrate
11 Transmissible Antibiotic Resistance 367

i­ntermediate consisting of two copies of the transposon linking donor and recipient
molecules. Tn21 (19,671-bp) and Tn21-like transposons contain the same tnpA,
tnpR, and res genes in different orientations and often include mer genes for resis-
tance to Hg++. Integrons are often found within transposons and ISCR elements
within integrons.
Another transposable element termed mic for mobile insertion cassette is com-
posed of a resistance gene bracketed by IR but lacking an integrase/transposase,
which must be supplied in trans for the unit to move [187].
Integrative and conjugative elements (ICE) (also known as conjugative transpo-
sons) encode a phage-like integrase (int) that catalyzes recombination between an
attP site on the unit and the host chromosome. The chromosomal integration site is
typically specific for a particular ICE family. They are bounded by IRs, and most
ICE also encode an excisionase (xis) that removes the ICE from the chromosome as
a circular molecule. Transfer of the circular form to a new host requires plasmid-like
genes that control DNA transfer and genes that form a mating pair between donor
and recipient. Lack of the latter function produces an integrative mobilizable ele-
ment (IME) that requires the missing functions in trans for transfer by conjugation.
Both ICE and IME can contain transposons, ISs, and integrons. ICE and IME thus
share many of the functions of conjugative and nonconjugative plasmids except for
their preference for a chromosomal location. Surveys of prokaryotic genomes indi-
cate that ICE are more common than plasmids and mobilizable elements outnumber
self-conjugative ones [188]. They occur in Gram-positive, Gram-negative, and
strictly anaerobic organisms.
Genomic islands are gene clusters, some very large, fixed in the chromosome
with features that suggest a foreign origin. In one strain of A. baumannii, an 86-kb
resistance island containing a variety of ISs, transposons, integrases, and 45 resis-
tance genes has been identified and obviously allows for rapid development of pan-­
resistance [189]. Genomic islands are also important in the evolution of
multiresistance in P. aeruginosa [190].
Phage particles carrying resistant genes (blaTEM, blaCTX-M, qnrA, qnrS, armA)
have been identified in wastewater or the human gut and constitute another class of
mobile elements [191, 192].
These elements can interact in various and complex ways. For example, a plas-
mid in K. pneumoniae carrying genes for both carbapenem, aminoglycoside, and
quinolone resistance was found to contain a complex transposon incorporating
blaKPC-3 inserted into a Tn3-family complex transposon with aminoglycoside resis-
tance genes and blaTEM-1 that also contained qnrB19 mobilized by ISEcp1 [193].
Because these mobile elements are built in modules, they can exchange, rearrange,
insert, delete, and recombine to generate remarkable diversity [194]. They have also
been doing this for a long time. Plasmid NR1 (also known as R100), one of the
original transmissible elements discovered in Japan in the 1950s, already contained
both integrons and transposons [195].
368 G. A. Jacoby

11.6 Overcoming Transmissible Antibiotic Resistance

Despite detailed investigation, no clinically useful direct attack on plasmid replica-

tion, stability, or mobility has been discovered. The most successful application of
knowledge about resistance has been the development of antibiotics that escape
resistance mechanisms and of inhibitors that restore effectiveness to agents that
would otherwise be inactivated. Table 11.5 lists examples of successful antibiotic
modification or discovery. β-lactam antibiotics provide many examples. The
7-α-methoxy group of the cephamycins cefoxitin and cefotetan allow activity
against many class A β-lactamase-producing bacteria as well as enhanced activity
against anaerobes. The oxyimino group of cefotaxime, ceftazidime, ceftriaxone,
cefepime, and aztreonam gave these antibiotics an even broader activity against
class A β-lactamases, and the carbapenems imipenem, meropenem, doripenem,
ertapenem, and others have the broadest activity spectrum of all. Success was met
with counterattack in the form of plasmid-mediated AmpC enzymes active against
cephamycins, extended-spectrum β-lactamases active against oxyimino-β-lactams,
and carbapenemases of classes A and D as well as class B [196].
Other successful antibiotic modifications include amikacin, a semisynthetic
derivative of kanamycin, with a 2-hydroxy-4 aminobutyric acid side chain that
makes it less susceptible to many aminoglycoside-modifying enzymes, and florfeni-
col, a fluorinated derivative of thiamphenicol, that is insensitive to CAT enzymes
and some chloramphenicol efflux pumps. The oxazolidinone tedizolid is more
potent than linezolid particularly against strains with the Cfr methyltransferase
because of facilitated binding to methylated 23 S rRNA [197], and the minocycline
derivative tigecycline is active against most organisms with transmissible tetracy-
cline resistance, although it can be overcome by a combination of tet(L) and tet(M)
[123]. See also Sect. 4.2.4.3 for other mechanisms of emerging tigecycline resis-
tance. Finally, the semisynthetic lipoglycopeptides telavancin, oritavancin, and
­telavancin are active against vancomycin-resistant enterococci containing the VanB
gene cluster, and oritavancin and telavancin are also active against VanA strains [82]
with the caution that resistance may emerge if used as monotherapy [198].
Clavulanic acid, sulbactam, and tazobactam are β-lactamase inhibiting β-lactams
that have been combined with otherwise enzyme-susceptible agents (amoxicillin,
ampicillin, ticarcillin, piperacillin) to expand their spectrum of action. Problems
with their use are that many β-lactamases are intrinsically resistant to inhibition
(Table 11.3) and that initially sensitive enzymes can develop inhibitor resistance by
mutation, as happened first for TEM and SHV-type β-lactamases and recently with
a CTX-M variety [199]. A new group of diazabicyclooctane compounds (avibac-
tam, relebactam, zidebactam, and others) with a broader spectrum of inhibition is
currently undergoing evaluation (see Sect. 4.2.2.3 for further details). Several have
direct antibacterial activity and attack organisms producing metallo-­carbapenemases
11 Transmissible Antibiotic Resistance 369

Table 11.5 Antibiotics with improved resistance properties

Parent antibiotic Improved derivative Escapes resistance from
Benzylpenicillin Cefoxitin, cefotetan Many class A β-lactamases
Ampicillin Cefotaxime, ceftazidime, ceftriaxone, Most class A β-lactamases
Cephalothin and cefepime, aztreonam
other Imipenem, meropenem, doripenem Most class A, C, and D
1st generation β-lactamases
β-lactams
Chloramphenicol Florfenicol CAT, some chloramphenicol
efflux pumps
Kanamycin Amikacin Many aminoglycoside-­
modifying enzymes
Linezolid Tedizolid Cfr methyltransferase
Tetracycline Tigecycline Tet efflux and ribosomal
protection agents
Vancomycin Telavancin, oritavancin, telavancin VanB,? VanA

as well as acting as β-lactamase inhibitors [200]. The ceftazidime-avibactam com-

bination has been approved for clinical use, and already inhibitor-resistant KPC-3
mutations have been reported in patients treated for K. pneumoniae infections pro-
ducing the carbapenemase [201].
Much can also be done to reduce the selective pressure for developing and main-
taining resistance. More than half of the antibiotics produced commercially are used
for other than human medicine. For example, streptomycin was sprayed for years on
apple and pear trees to prevent a destructive bacterial disease known as fire blight
until the responsible organism Erwinia amylovora became streptomycin resistant,
and its use had to be abandoned. Other nonhuman uses that contribute to resistance
development include antibiotic use for livestock growth promotion; use for pest
control; use for therapy of household pets; use for treatment of cows, pigs, chickens,
fish, and other animals produced for food; use as biocide in toiletries, skin care
creams, and cleaning products; and use in research and industry [202]. For example,
the glycopeptide avoparcin was used as an animal growth promoter with selection
of glycopeptide-resistant enterococci that were shown to be similar to those from
human infections [203] leading to a ban of avoparcin use in Europe. Attention needs
to be paid also to agents other than antibiotics that can select for resistance, such as
use of olaquindox and carbadox as feed additives for pigs that led to the spread of
the plasmid-mediated OqxAB multidrug efflux pump.
More than 40 years ago, studies in hospitals showed that more than half the anti-
biotics used clinically were not needed, were given inappropriately, or were dosed
incorrectly [204]. Recent studies indicate little if any improvement, but up-to-date
guidelines for antibiotic stewardship in human medicine are available [205], and
their implementation is now a requirement for hospital and nursing care center
approval by the Joint Commission that accredits healthcare organizations in the
United States.
370 G. A. Jacoby

Major Points
Our adversaries turn out to be cleverer than we thought with an abundant reservoir of
resistance genes and a toolkit of efficient genetic devices to mobilize, incorporate,
and share them. Resistance is increasing, and one by one agents that we thought
could still be counted on have become less reliable. Knowledge of resistance mecha-
nisms has allowed the development of antibiotics and combinations effective for a
time against resistant pathogens, but bacteria will continue to evolve resistance.
Speedier diagnostic tests will facilitate choice of effective agents, but new antibiotics
and new ideas to combat resistance are urgently needed.

References

1. Watanabe T. Infective heredity of multiple drug resistance in bacteria. Bacteriol Rev.

1963;27:87–115.
2. Novick RP, Clowes RC, Cohen SN, Curtiss R 3rd, Datta N, Falkow S. Uniform nomenclature
for bacterial plasmids: a proposal. Bacteriol Rev. 1976;40(1):168–89.
3. Novick RP. Analysis by transduction of mutations affecting penicillinase formation in
Staphylococcus aureus. J Gen Microbiol. 1963;33:121–36.
4. Datta N, Kontomichalou P. Penicillinase synthesis controlled by infectious R factors in
Enterobacteriaceae. Nature. 1965;208(5007):239–41.
5. Manson JM, Keis S, Smith JM, Cook GM. Acquired bacitracin resistance in
Enterococcus faecalis is mediated by an ABC transporter and a novel regulatory pro-
tein, BcrR. Antimicrob Agents Chemother. 2004;48(10):3743–8. https://doi.org/10.1128/
AAC.48.10.3743-3748.2004.
6. Mitsuhashi S, Morimura M, Kono K, Oshima H. Elimination of drug resistance of
Staphylococcus aureus by treatment with acriflavine. J Bacteriol. 1963;86:162–4.
7. Breeze AS, Obaseiki-Ebor EE. Transferable nitrofuran resistance conferred by R-plasmids in
clinical isolates of Escherichia coli. J Antimicrob Chemother. 1983;12(5):459–67.
8. Leclercq R, Derlot E, Duval J, Courvalin P. Plasmid-mediated resistance to vancomycin
and teicoplanin in Enterococcus faecium. N Engl J Med. 1988;319(3):157–61. https://doi.
org/10.1056/NEJM198807213190307.
9. Liu YY, Wang Y, Walsh TR, Yi LX, Zhang R, Spencer J, Doi Y, Tian G, Dong B, Huang X,
Yu LF, Gu D, Ren H, Chen X, Lv L, He D, Zhou H, Liang Z, Liu JH, Shen J. Emergence
of plasmid-mediated colistin resistance mechanism MCR-1 in animals and human beings in
China: a microbiological and molecular biological study. Lancet Infect Dis. 2016;16(2):161–
8. https://doi.org/10.1016/S1473-3099(15)00424-7.
10. Breuil J, Dublanchet A, Truffaut N, Sebald M. Transferable 5-nitroimidazole resistance in the
Bacteroides fragilis group. Plasmid. 1989;21(2):151–4.
11. Fleming MP, Datta N, Gruneberg RN. Trimethoprim resistance determined by R factors. Br
Med J. 1972;1(5802):726–8.
12. Evans RJ, Waterworth PM. Naturally-occurring fusidic acid resistance in staphylococci and
its linkage to other resistances. J Clin Pathol. 1966;19(6):555–60.
13. Benveniste R, Davies J. R-factor mediated gentamicin resistance: a new enzyme which modi-
fies aminoglycoside antibiotics. FEBS Lett. 1971;14(5):293–6.
14. Chandrasekaran S, Lalithakumari D. Plasmid-mediated rifampicin resistance in
Pseudomonas fluorescens. J Med Microbiol. 1998;47(3):197–200. https://doi.
org/10.1099/00222615-47-3-197.
11 Transmissible Antibiotic Resistance 371

15. Martínez-Martínez L, Pascual A, Jacoby GA. Quinolone resistance from a transferable plas-
mid. Lancet. 1998;351(9105):797–9.
16. Mendoza C, Garcia JM, Llaneza J, Mendez FJ, Hardisson C, Ortiz JM. Plasmid-determined
resistance to fosfomycin in Serratia marcescens. Antimicrob Agents Chemother.
1980;18(2):215–9.
17. Bauernfeind A, Chong Y, Schweighart S. Extended broad spectrum ß-lactamase in Klebsiella
pneumoniae including resistance to cephamycins. Infection. 1989;17(5):316–21.
18. Knothe H, Shah P, Krcmery V, Antal M, Mitsuhashi S. Transferable resistance to cefotaxime,
cefoxitin, cefamandole and cefuroxime in clinical isolates of Klebsiella pneumoniae and
Serratia marcescens. Infection. 1983;11(6):315–7.
19. Watanabe M, Iyobe S, Inoue M, Mitsuhashi S. Transferable imipenem resistance in
Pseudomonas aeruginosa. Antimicrob Agents Chemother. 1991;35(1):147–51.
20. Rahman M, Noble WC, Cookson B. Mupirocin-resistant Staphylococcus aureus. Lancet.
1987;ii:387.
21. Le Goffic F, Capmau ML, Bonnet D, Cerceau C, Soussy C, Dublanchet A, Duval J. Plasmid-­
mediated pristinamycin resistance. PAC IIA: a new enzyme which modifies pristinamycin
IIA. J Antibiot. 1977;30(8):665–9.
22. Long KS, Poehlsgaard J, Kehrenberg C, Schwarz S, Vester B. The Cfr rRNA methyltrans-
ferase confers resistance to phenicols, lincosamides, oxazolidinones, pleuromutilins, and
streptogramin A antibiotics. Antimicrob Agents Chemother. 2006;50(7):2500–5. https://doi.
org/10.1128/AAC.00131-06.
23. Xavier BB, Das AJ, Cochrane G, De Ganck S, Kumar-Singh S, Aarestrup FM, Goossens H,
Malhotra-Kumar S. Consolidating and exploring antibiotic resistance gene data resources.
J Clin Microbiol. 2016;54(4):851–9. https://doi.org/10.1128/JCM.02717-15.
24. Armstrong ES, Kostrub CF, Cass RT, Moser HE, Serio AW, Miller GH. Aminoglycosides.
In: Dougherty TJ, Pucci MJ, editors. Antibiotic discovery and development, vol. I. 1st ed.
New York/Dordrecht/Heidelberg/London: Springer; 2012. p. 229–69.
25. Shaw KJ, Rather PN, Hare RS, Miller GH. Molecular genetics of aminoglycoside resistance
genes and familial relationships of the aminoglycoside-modifying enzymes. Microbiol Rev.
1993;57(1):138–63.
26. Ramirez MS, Tolmasky ME. Aminoglycoside modifying enzymes. Drug Resist Updat.
2010;13(6):151–71. https://doi.org/10.1016/j.drup.2010.08.003.
27. Krause KM, Serio AW, Kane TR, Connolly LE. Aminoglycosides: an overview. Cold Spring
Harb Perspect Med. 2016;6(6). https://doi.org/10.1101/cshperspect.a027029.
28. Yokoyama K, Doi Y, Yamane K, Kurokawa H, Shibata N, Shibayama K, Yagi T, Kato H,
Arakawa Y. Acquisition of 16S rRNA methylase gene in Pseudomonas aeruginosa. Lancet.
2003;362(9399):1888–93. https://doi.org/10.1016/S0140-6736(03)14959-8.
29. Doi Y, Arakawa Y. 16S ribosomal RNA methylation: emerging resistance mechanism against
aminoglycosides. Clin Infect Dis. 2007;45(1):88–94. https://doi.org/10.1086/518605.
30. Wachino J, Shibayama K, Kurokawa H, Kimura K, Yamane K, Suzuki S, Shibata N, Ike
Y, Arakawa Y. Novel plasmid-mediated 16S rRNA m1A1408 methyltransferase, NpmA,
found in a clinically isolated Escherichia coli strain resistant to structurally diverse ami-
noglycosides. Antimicrob Agents Chemother. 2007;51(12):4401–9. https://doi.org/10.1128/
AAC.00926-07. AAC.00926-07 [pii]
31. Gauntlett JC, Gebhard S, Keis S, Manson JM, Pos KM, Cook GM. Molecular analysis of
BcrR, a membrane-bound bacitracin sensor and DNA-binding protein from Enterococcus
faecalis. J Biol Chem. 2008;283(13):8591–600. https://doi.org/10.1074/jbc.M709503200.
32. Matos R, Pinto VV, Ruivo M, Lopes Mde F. Study on the dissemination of the bcrABDR clus-
ter in Enterococcus spp. reveals that the BcrAB transporter is sufficient to confer high-level
bacitracin resistance. Int J Antimicrob Agents. 2009;34(2):142–7. https://doi.org/10.1016/j.
ijantimicag.2009.02.008.
33. Ambler RP, Coulson AFW, Frère J-M, Ghuysen J-M, Joris B, Forsman M, Levesque RC,
Tiraby G, Waley SG. A standard numbering scheme for the class A ß-lactamases. Biochem
J. 1991;276:269–70.
372 G. A. Jacoby

34. Bush K, Jacoby GA, Medeiros AA. A functional classification scheme for ß-lactamases and
its correlation with molecular structure. Antimicrob Agents Chemother. 1995;39(6):1211–33.
35. Bush K, Jacoby GA. Updated functional classification of ß-lactamases. Antimicrob Agents
Chemother. 2010;54(3):969–76. https://doi.org/10.1128/AAC.01009-09. AAC.01009-09
[pii]
36. Kernodle DS, Stratton CW, McMurray LW, Chipley JR, McGraw PA. Differentiation of
ß-lactamase variants of Staphylococcus aureus by substrate hydrolysis profiles. J Infect Dis.
1989;159(1):103–8.
37. East AK, Dyke KG. Cloning and sequence determination of six Staphylococcus aureus
β-lactamases and their expression in Escherichia coli and Staphylococcus aureus. J Gen
Microbiol. 1989;135(4):1001–15. https://doi.org/10.1099/00221287-135-4-1001.
38. Murray BE. β-lactamase-producing enterococci. Antimicrob Agents Chemother.
1992;36(11):2355–9.
39. ß-Lactamase classification and amino acid sequences for TEM, SHV and OXA extended-­
spectrum and inhibitor resistant enzymes http:// www.lahey.org/studies.
40. Lartigue MF, Leflon-Guibout V, Poirel L, Nordmann P, Nicolas-Chanoine MH. Promoters
P3, Pa/Pb, P4, and P5 upstream from blaTEM genes and their relationship to ß-lactam resis-
tance. Antimicrob Agents Chemother. 2002;46(12):4035–7.
41. Philippon A, Arlet G, Jacoby GA. Plasmid-determined AmpC-type ß-lactamases. Antimicrob
Agents Chemother. 2002;46(1):1–11.
42. Jacoby GA. AmpC ß-lactamases. Clin Microbiol Rev. 2009;22(1):161–82. https://doi.
org/10.1128/CMR.00036-08. 22/1/161 [pii]
43. Nordmann P, Poirel L. The difficult-to-control spread of carbapenemase produc-
ers in Enterobacteriaceae worldwide. Clin Microbiol Infect. 2014. https://doi.
org/10.1111/1469-0691.12719.
44. Bontron S, Nordmann P, Poirel L. Transposition of Tn125 encoding the NDM-1 carbapen-
emase in Acinetobacter baumannii. Antimicrob Agents Chemother. 2016;60(12):7245–51.
https://doi.org/10.1128/AAC.01755-16.
45. Poirel L, Naas T, Nordmann P. Diversity, epidemiology, and genetics of class D ß-lactamases.
Antimicrob Agents Chemother. 2010;54(1):24–38. https://doi.org/10.1128/AAC.01512-08.
doi:AAC.01512-08 [pii]
46. Medeiros AA. Evolution and dissemination of ß-lactamases accelerated by generations of
ß-lactam antibiotics. Clin Infect Dis. 1997;24(Suppl 1):S19–45.
47. Castanheira M, Mendes RE, Jones RN, Sader HS. Changes in the frequencies of β-lactamase
genes among Enterobacteriaceae isolates in U.S. hospitals, 2012 to 2014: activity of
ceftazidime-avibactam tested against β-lactamase-producing isolates. Antimicrob Agents
Chemother. 2016;60(8):4770–7. https://doi.org/10.1128/AAC.00540-16.
48. D'Andrea MM, Arena F, Pallecchi L, Rossolini GM. CTX-M-type β-lactamases: a success-
ful story of antibiotic resistance. Int J Med Microbiol. 2013;303(6–7):305–17. https://doi.
org/10.1016/j.ijmm.2013.02.008.
49. Bevan ER, Jones AM, Hawkey PM. Global epidemiology of CTX-M β-lactamases: temporal
and geographical shifts in genotype. J Antimicrob Chemother. 2017:2145–55. https://doi.
org/10.1093/jac/dkx146.
50. Schwarz S, Johnson AP. Transferable resistance to colistin: a new but old threat. J Antimicrob
Chemother. 2016;71(8):2066–70. https://doi.org/10.1093/jac/dkw274.
51. Snesrud E, Ong AC, Corey B, Kwak YI, Clifford R, Gleeson T, Wood S, Whitman TJ,
Lesho EP, Hinkle M, McGann P. Analysis of serial isolates of mcr-1-positive Escherichia
coli reveals a highly active ISApl1 transposon. Antimicrob Agents Chemother. 2017;61(5).
https://doi.org/10.1128/AAC.00056-17.
52. Yin W, Li H, Shen Y, Liu Z, Wang S, Shen Z, Zhang R, Walsh TR, Shen J, Wang Y. Novel
plasmid-mediated colistin resistance gene mcr-3 in Escherichia coli. MBio. 2017;8(3).
https://doi.org/10.1128/mBio.00543-17.
11 Transmissible Antibiotic Resistance 373

53. Jacoby G, Strahilevitz J, Hooper D. Plasmid-mediated quinolone resistance. In: Tolmasky

M, Alonso JA, editors. Plasmids – biology and impact in biotechnology and discovery.
Washington, D.C.: ASM Press; 2015.
54. Rodriguez-Martinez JM, Machuca J, Cano ME, Calvo J, Martinez-Martinez L, Pascual
A. Plasmid-mediated quinolone resistance: two decades on. Drug Resist Updat. 2016;29:13–
29. https://doi.org/10.1016/j.drup.2016.09.001.
55. Tran JH, Jacoby GA. Mechanism of plasmid-mediated quinolone resistance. Proc Natl Acad
Sci U S A. 2002;99(8):5638–42. https://doi.org/10.1073/pnas.082092899.
56. Kim ES, Chen C, Braun M, Kim HY, Okumura R, Wang Y, Jacoby GA, Hooper
DC. Interactions between QnrB, QnrB mutants, and DNA gyrase. Antimicrob Agents
Chemother. 2015;59(9):5413–9. https://doi.org/10.1128/AAC.00771-15.
57. Shah S, Heddle JG. Squaring up to DNA: pentapeptide repeat proteins and DNA mimicry. Appl
Microbiol Biotechnol. 2014;98(23):9545–60. https://doi.org/10.1007/s00253-014-6151-3.
58. Jacoby GA, Corcoran MA, Hooper DC. Protective effect of Qnr on agents other than quino-
lones that target DNA gyrase. Antimicrob Agents Chemother. 2015;59(11):6689–95. https://
doi.org/10.1128/AAC.01292-15.
59. Zhao X, Xu X, Zhu D, Ye X, Wang M. Decreased quinolone susceptibility in high percentage
of Enterobacter cloacae clinical isolates caused only by Qnr determinants. Diagn Microbiol
Infect Dis. 2010;67:110–3. https://doi.org/10.1016/j.diagmicrobio.2009.12.018. S0732-
8893(09)00509-4 [pii]
60. Robicsek A, Strahilevitz J, Jacoby GA, Macielag M, Abbanat D, Park CH, Bush K, Hooper
DC. Fluoroquinolone-modifying enzyme: a new adaptation of a common aminoglycoside
acetyltransferase. Nat Med. 2006;12(1):83–8.
61. Quiroga MP, Orman B, Errecalde L, Kaufman S, Centron D. Characterization of Tn6238 with
a new allele of aac(6′)-Ib-cr. Antimicrob Agents Chemother. 2015;59(5):2893–7. https://doi.
org/10.1128/AAC.03213-14.
62. Ho P, Ng K, Lo W, Law P, Lai EL, Wang Y, Chow K. Plasmid-mediated OqxAB is an
important mechanism for nitrofurantoin resistance in Escherichia coli. Antimicrob Agents
Chemother. 2016;60(1):537–43.
63. Drlica K. The mutant selection window and antimicrobial resistance. J Antimicrob
Chemother. 2003;52(1):11–7.
64. Vinué L, Corcoran MA, Hooper DC, Jacoby GA. Mutations that enhance the ciprofloxacin
resistance of Escherichia coli with qnrA1. Antimicrob Agents Chemother. 2016;60(3):1537–
45. https://doi.org/10.1128/AAC.02167-15.
65. Castaneda-Garcia A, Blazquez J, Rodriguez-Rojas A. Molecular mechanisms and clini-
cal impact of acquired and intrinsic fosfomycin resistance. Antibiotics. 2013;2(2):217–36.
https://doi.org/10.3390/antibiotics2020217.
66. Fu Z, Liu Y, Chen C, Guo Y, Ma Y, Yang Y, Hu F, Xu X, Wang M. Characterization of fos-
fomycin resistance gene, fosB, in methicillin-resistant Staphylococcus aureus isolates. PLoS
One. 2016;11(5):e0154829. https://doi.org/10.1371/journal.pone.0154829.
67. Wang Y, Yao H, Deng F, Liu D, Zhang Y, Shen Z. Identification of a novel fosXCC gene confer-
ring fosfomycin resistance in Campylobacter. J Antimicrob Chemother. 2015;70(4):1261–3.
https://doi.org/10.1093/jac/dku488.
68. Kitanaka H, Wachino J, Jin W, Yokoyama S, Sasano MA, Hori M, Yamada K, Kimura K,
Arakawa Y. Novel integron-mediated fosfomycin resistance gene fosK. Antimicrob Agents
Chemother. 2014;58(8):4978–9. https://doi.org/10.1128/AAC.03131-14.
69. O'Neill AJ, McLaws F, Kahlmeter G, Henriksen AS, Chopra I. Genetic basis of resistance to
fusidic acid in staphylococci. Antimicrob Agents Chemother. 2007;51(5):1737–40. https://
doi.org/10.1128/AAC.01542-06.
70. Farrell DJ, Castanheira M, Chopra I. Characterization of global patterns and the genetics of
fusidic acid resistance. Clin Infect Dis. 2011;52(Suppl 7):S487–92. https://doi.org/10.1093/
cid/cir164.
374 G. A. Jacoby

71. Fernandes P. Fusidic acid: a bacterial elongation factor inhibitor for the oral treatment of acute
and chronic staphylococcal infections. Cold Spring Harb Perspect Med. 2016;6(1):a025437.
https://doi.org/10.1101/cshperspect.a025437.
72. O'Neill AJ, Larsen AR, Skov R, Henriksen AS, Chopra I. Characterization of the epidemic
European fusidic acid-resistant impetigo clone of Staphylococcus aureus. J Clin Microbiol.
2007;45(5):1505–10. https://doi.org/10.1128/JCM.01984-06.
73. Wang JL, Tang HJ, Hsieh PH, Chiu FY, Chen YH, Chang MC, Huang CT, Liu CP, Lau
YJ, Hwang KP, Ko WC, Wang CT, Liu CY, Liu CL, Hsueh PR. Fusidic acid for the treat-
ment of bone and joint infections caused by methicillin-resistant Staphylococcus aureus. Int
J Antimicrob Agents. 2012;40(2):103–7. https://doi.org/10.1016/j.ijantimicag.2012.03.010.
74. Farrell DJ, Mendes RE, Castanheira M, Jones RN. Activity of fusidic acid tested against
staphylococci isolated from patients in U.S. medical centers in 2014. Antimicrob Agents
Chemother. 2016;60(6):3827–31. https://doi.org/10.1128/AAC.00238-16.
75. Bennett AD, Shaw WV. Resistance to fusidic acid in Escherichia coli mediated by the type I
variant of chloramphenicol acetyltransferase. A plasmid-encoded mechanism involving anti-
biotic binding. Biochem J. 1983;215(1):29–38.
76. Murray IA, Cann PA, Day PJ, Derrick JP, Sutcliffe MJ, Shaw WV, Leslie AG. Steroid rec-
ognition by chloramphenicol acetyltransferase: engineering and structural analysis of a high
affinity fusidic acid binding site. J Mol Biol. 1995;254(5):993–1005. https://doi.org/10.1006/
jmbi.1995.0671.
77. Courvalin P. Vancomycin resistance in gram-positive cocci. Clin Infect Dis. 2006;42(Suppl
1):S25–34. https://doi.org/10.1086/491711.
78. Cattoir V, Leclercq R. Twenty-five years of shared life with vancomycin-resistant enterococci:
is it time to divorce? J Antimicrob Chemother. 2013;68(4):731–42. https://doi.org/10.1093/
jac/dks469.
79. Périchon B, Courvalin P. Glycopeptide resistance. In: Dougherty TJ, Pucci MJ, editors.
Antibiotic discovery and development, vol. 1. 1st ed. New York/Dordrecht/Heidelberg/
London: Springer; 2012. p. 515–42.
80. Holman TR, Wu Z, Wanner BL, Walsh CT. Identification of the DNA-binding site for the
phosphorylated VanR protein required for vancomycin resistance in Enterococcus faecium.
Biochemistry. 1994;33(15):4625–31.
81. Abadia Patino L, Courvalin P, Perichon B. vanE gene cluster of vancomycin-resistant
Enterococcus faecalis BM4405. J Bacteriol. 2002;184(23):6457–64.
82. Zeng D, Debabov D, Hartsell TL, Cano RJ, Adams S, Schuyler JA, McMillan R, Pace
JL. Approved glycopeptide antibacterial drugs: mechanism of action and resistance. Cold
Spring Harb Perspect Med. 2016. https://doi.org/10.1101/cshperspect.a026989.
83. Schwarz S, Shen J, Kadlec K, Wang Y, Michael GB, Fessler AT, Vester B. Lincosamides,
streptogramins, phenicols, and pleuromutilins: mode of action and mechanisms of resistance.
Cold Spring Harb Perspect Med. 2016. https://doi.org/10.1101/cshperspect.a027037.
84. Morar M, Pengelly K, Koteva K, Wright GD. Mechanism and diversity of the erythromy-
cin esterase family of enzymes. Biochemistry. 2012;51(8):1740–51. https://doi.org/10.1021/
bi201790u.
85. Korczynska M, Mukhtar TA, Wright GD, Berghuis AM. Structural basis for streptogramin
B resistance in Staphylococcus aureus by virginiamycin B lyase. Proc Natl Acad Sci U S A.
2007;104(25):10388–93. https://doi.org/10.1073/pnas.0701809104.
86. Bozdogan B, Berrezouga L, Kuo MS, Yurek DA, Farley KA, Stockman BJ, Leclercq R. A
new resistance gene, linB, conferring resistance to lincosamides by nucleotidylation in
Enterococcus faecium HM1025. Antimicrob Agents Chemother. 1999;43(4):925–9.
87. Fyfe C, Grossman TH, Kerstein K, Sutcliffe J. Resistance to macrolide antibiotics in public
health pathogens. Cold Spring Harb Perspect Med. 2016;6(10). https://doi.org/10.1101/csh-
perspect.a025395.
88. http://faculty.washington.edu/marilynr/.
89. Weisblum B. Insights into erythromycin action from studies of its activity as inducer of resis-
tance. Antimicrob Agents Chemother. 1995;39(4):797–805.
11 Transmissible Antibiotic Resistance 375

90. Dzyubak E, Yap MN. The expression of antibiotic resistance methyltransferase correlates
with mRNA stability independently of ribosome stalling. Antimicrob Agents Chemother.
2016;60(12):7178–88. https://doi.org/10.1128/AAC.01806-16.
91. Siberry GK, Tekle T, Carroll K, Dick J. Failure of clindamycin treatment of methicillin-­
resistant Staphylococcus aureus expressing inducible clindamycin resistance in vitro. Clin
Infect Dis. 2003;37(9):1257–60. https://doi.org/10.1086/377501.
92. Boal AK, Grove TL, McLaughlin MI, Yennawar NH, Booker SJ, Rosenzweig AC. Structural
basis for methyl transfer by a radical SAM enzyme. Science. 2011;332(6033):1089–92.
https://doi.org/10.1126/science.1205358.
93. Deshpande LM, Ashcraft DS, Kahn HP, Pankey G, Jones RN, Farrell DJ, Mendes
RE. Detection of a new cfr-like gene, cfr(B), in Enterococcus faecium isolates recovered from
human specimens in the United States as part of the SENTRY antimicrobial surveillance
program. Antimicrob Agents Chemother. 2015;59(10):6256–61. https://doi.org/10.1128/
AAC.01473-15.
94. Tang Y, Dai L, Sahin O, Wu Z, Liu M, Zhang Q. Emergence of a plasmid-borne multidrug
resistance gene cfr(C) in foodborne pathogen Campylobacter. J Antimicrob Chemother.
2017;72(6):1581–8. https://doi.org/10.1093/jac/dkx023.
95. Ojo KK, Striplin MJ, Ulep CC, Close NS, Zittle J, Luis H, Bernardo M, Leitao J, Roberts
MC. Staphylococcus efflux msr(a) gene characterized in Streptococcus, Enterococcus,
Corynebacterium, and Pseudomonas isolates. Antimicrob Agents Chemother.
2006;50(3):1089–91. https://doi.org/10.1128/AAC.50.3.1089-1091.2006.
96. Kehrenberg C, Ojo KK, Schwarz S. Nucleotide sequence and organization of the mul-
tiresistance plasmid pSCFS1 from Staphylococcus sciuri. J Antimicrob Chemother.
2004;54(5):936–9. https://doi.org/10.1093/jac/dkh457.
97. Li XS, Dong WC, Wang XM, Hu GZ, Wang YB, Cai BY, Wu CM, Wang Y, Du XD. Presence
and genetic environment of pleuromutilin-lincosamide-streptogramin a resistance gene lsa(E)
in enterococci of human and swine origin. J Antimicrob Chemother. 2014;69(5):1424–6.
https://doi.org/10.1093/jac/dkt502.
98. Wang Y, Lv Y, Cai J, Schwarz S, Cui L, Hu Z, Zhang R, Li J, Zhao Q, He T, Wang D, Wang Z,
Shen Y, Li Y, Fessler AT, Wu C, Yu H, Deng X, Xia X, Shen J. A novel gene, optrA, that con-
fers transferable resistance to oxazolidinones and phenicols and its presence in Enterococcus
faecalis and Enterococcus faecium of human and animal origin. J Antimicrob Chemother.
2015;70(8):2182–90. https://doi.org/10.1093/jac/dkv116.
99. Fan R, Li D, Wang Y, He T, Fessler AT, Schwarz S, Wu C. Presence of the optrA gene in
methicillin-resistant Staphylococcus sciuri of porcine origin. Antimicrob Agents Chemother.
2016;60(12):7200–5. https://doi.org/10.1128/AAC.01591-16.
100. He T, Shen Y, Schwarz S, Cai J, Lv Y, Li J, Fessler AT, Zhang R, Wu C, Shen J, Wang
Y. Genetic environment of the transferable oxazolidinone/phenicol resistance gene optrA
in Enterococcus faecalis isolates of human and animal origin. J Antimicrob Chemother.
2016;71(6):1466–73. https://doi.org/10.1093/jac/dkw016.
101. Huang J, Chen L, Wu Z, Wang L. Retrospective analysis of genome sequences revealed
the wide dissemination of optrA in Gram-positive bacteria. J Antimicrob Chemother.
2017;72(2):614–6. https://doi.org/10.1093/jac/dkw488.
102. Gilbart J, Perry CR, Slocombe B. High-level mupirocin resistance in Staphylococcus aureus:
evidence for two distinct isoleucyl-tRNA synthetases. Antimicrob Agents Chemother.
1993;37(1):32–8.
103. Seah C, Alexander DC, Louie L, Simor A, Low DE, Longtin J, Melano RG. MupB, a new
high-level mupirocin resistance mechanism in Staphylococcus aureus. Antimicrob Agents
Chemother. 2012;56(4):1916–20. https://doi.org/10.1128/AAC.05325-11.
104. Antonov NK, Garzon MC, Morel KD, Whittier S, Planet PJ, Lauren CT. High prevalence
of mupirocin resistance in Staphylococcus aureus isolates from a pediatric population.
Antimicrob Agents Chemother. 2015;59(6):3350–6. https://doi.org/10.1128/AAC.00079-15.
105. Trinh S, Reysset G. Detection by PCR of the nim genes encoding 5-nitroimidazole resistance
in Bacteroides spp. J Clin Microbiol. 1996;34(9):2078–84.
376 G. A. Jacoby

106. Husain F, Veeranagouda Y, Hsi J, Meggersee R, Abratt V, Wexler HM. Two multidrug-­
resistant clinical isolates of Bacteroides fragilis carry a novel metronidazole resistance nim
gene (nimJ). Antimicrob Agents Chemother. 2013;57(8):3767–74. https://doi.org/10.1128/
AAC.00386-13.
107. Carlier JP, Sellier N, Rager MN, Reysset G. Metabolism of a 5-nitroimidazole in suscep-
tible and resistant isogenic strains of Bacteroides fragilis. Antimicrob Agents Chemother.
1997;41(7):1495–9.
108. Reysset G, Haggoud A, Sebald M. Genetics of resistance of Bacteroides species to
5-­nitroimidazole. Clin Infect Dis. 1993;16(Suppl 4):S401–3.
109. Snydman DR, Jacobus NV, McDermott LA, Ruthazer R, Golan Y, Goldstein EJ, Finegold SM,
Harrell LJ, Hecht DW, Jenkins SG, Pierson C, Venezia R, Yu V, Rihs J, Gorbach SL. National
survey on the susceptibility of Bacteroides fragilis group: report and analysis of trends in
the United States from 1997 to 2004. Antimicrob Agents Chemother. 2007;51(5):1649–55.
https://doi.org/10.1128/AAC.01435-06.
110. Sahm DF, Deane J, Bien PA, Locke JB, Zuill DE, Shaw KJ, Bartizal KF. Results of the sur-
veillance of tedizolid activity and resistance program: in vitro susceptibility of gram-positive
pathogens collected in 2011 and 2012 from the United States and Europe. Diagn Microbiol
Infect Dis. 2015;81(2):112–8. https://doi.org/10.1016/j.diagmicrobio.2014.08.011.
111. Schwarz S, Kehrenberg C, Doublet B, Cloeckaert A. Molecular basis of bacterial resistance
to chloramphenicol and florfenicol. FEMS Microbiol Rev. 2004;28(5):519–42. https://doi.
org/10.1016/j.femsre.2004.04.001.
112. Schwarz S, Werckenthin C, Kehrenberg C. Identification of a plasmid-borne chloramphenicol-­
florfenicol resistance gene in Staphylococcus sciuri. Antimicrob Agents Chemother.
2000;44(9):2530–3.
113. Butaye P, Cloeckaert A, Schwarz S. Mobile genes coding for efflux-mediated antimicro-
bial resistance in Gram-positive and Gram-negative bacteria. Int J Antimicrob Agents.
2003;22(3):205–10.
114. Hansen LH, Jensen LB, Sorensen HI, Sorensen SJ. Substrate specificity of the OqxAB
multidrug resistance pump in Escherichia coli and selected enteric bacteria. J Antimicrob
Chemother. 2007;60(1):145–7.
115. Tribuddharat C, Fennewald M. Integron-mediated rifampin resistance in Pseudomonas aeru-
ginosa. Antimicrob Agents Chemother. 1999;43(4):960–2.
116. Almeida AC, Cavalcanti FL, Martins WM, Vilela MA, Gales AC, Morais Junior MA, Morais
MM. First description of KPC-2-producing Klebsiella oxytoca in Brazil. Antimicrob Agents
Chemother. 2013;57(8):4077–8. https://doi.org/10.1128/AAC.02376-12.
117. Shen P, Yi M, Fu Y, Ruan Z, Du X, Yu Y, Xie X. Detection of an Escherichia coli ST167 strain
with two tandem copies of blaNDM-1 encoded in the chromosome. J Clin Microbiol. 2016.
https://doi.org/10.1128/JCM.01581-16.
118. Bean DC, Livermore DM, Hall LM. Plasmids imparting sulfonamide resistance in Escherichia
coli: implications for persistence. Antimicrob Agents Chemother. 2009;53(3):1088–93.
https://doi.org/10.1128/AAC.00800-08.
119. Leski TA, Bangura U, Jimmy DH, Ansumana R, Lizewski SE, Stenger DA, Taitt CR, Vora
GJ. Multidrug-resistant tet(X)-containing hospital isolates in Sierra Leone. Int J Antimicrob
Agents. 2013;42(1):83–6. https://doi.org/10.1016/j.ijantimicag.2013.04.014.
120. Connell SR, Tracz DM, Nierhaus KH, Taylor DE. Ribosomal protection proteins and their
mechanism of tetracycline resistance. Antimicrob Agents Chemother. 2003;47(12):3675–81.
121. Warburton PJ, Amodeo N, Roberts AP. Mosaic tetracycline resistance genes encoding
ribosomal protection proteins. J Antimicrob Chemother. 2016;71(12):3333–9. https://doi.
org/10.1093/jac/dkw304.
122. Orth P, Schnappinger D, Hillen W, Saenger W, Hinrichs W. Structural basis of gene regulation
by the tetracycline inducible Tet repressor-operator system. Nat Struct Biol. 2000;7(3):215–
9. https://doi.org/10.1038/73324.
123. Fiedler S, Bender JK, Klare I, Halbedel S, Grohmann E, Szewzyk U, Werner G. Tigecycline
resistance in clinical isolates of Enterococcus faecium is mediated by an upregulation of
11 Transmissible Antibiotic Resistance 377

plasmid-encoded tetracycline determinants tet(L) and tet(M). J Antimicrob Chemother.

2016;71(4):871–81. https://doi.org/10.1093/jac/dkv420.
124. Brolund A, Sundqvist M, Kahlmeter G, Grape M. Molecular characterisation of trime-
thoprim resistance in Escherichia coli and Klebsiella pneumoniae during a two year inter-
vention on trimethoprim use. PLoS One. 2010;5(2):e9233. https://doi.org/10.1371/journal.
pone.0009233.
125. Huovinen P, Sundstrom L, Swedberg G, Skold O. Trimethoprim and sulfonamide resistance.
Antimicrob Agents Chemother. 1995;39(2):279–89.
126. Toulouse JL, Edens TJ, Alejaldre L, Manges AR, Pelletier JN. Integron-associated DfrB4, a
previously uncharacterized member of the trimethoprim-resistant dihydrofolate reductase B
family, is a clinically identified emergent source of antibiotic resistance. Antimicrob Agents
Chemother. 2017;61(5). https://doi.org/10.1128/AAC.02665-16.
127. Charpentier E, Courvalin P. Emergence of the trimethoprim resistance gene dfrD in Listeria
monocytogenes BM4293. Antimicrob Agents Chemother. 1997;41(5):1134–6.
128. Nurjadi D, Schafer J, Friedrich-Janicke B, Mueller A, Neumayr A, Calvo-Cano A, Goorhuis
A, Molhoek N, Lagler H, Kantele A, Van Genderen PJ, Gascon J, Grobusch MP, Caumes E,
Hatz C, Fleck R, Mockenhaupt FP, Zanger P (2015) Predominance of dfrG as determinant of
trimethoprim resistance in imported Staphylococcus aureus. Clin Microbiol Infect 2015;21
(12):1095.e5–1095.e9. doi:https://doi.org/10.1016/j.cmi.2015.08.021.
129. Kadlec K, Fessler AT, Couto N, Pomba CF, Schwarz S. Unusual small plasmids carrying
the novel resistance genes dfrK or apmA isolated from methicillin-resistant or -susceptible
staphylococci. J Antimicrob Chemother. 2012;67(10):2342–5. https://doi.org/10.1093/jac/
dks235.
130. Perry J, Waglechner N, Wright G. The prehistory of antibiotic resistance. Cold Spring Harb
Perspect Med. 2016;6(6). https://doi.org/10.1101/cshperspect.a025197.
131. Benveniste R, Davies J. Aminoglycoside antibiotic-inactivating enzymes in actinomycetes
similar to those present in clinical isolates of antibiotic-resistant bacteria. Proc Natl Acad Sci
U S A. 1973;70(8):2276–80.
132. Cundliffe E. How antibiotic-producing organisms avoid suicide. Annu Rev Microbiol.
1989;43:207–33. https://doi.org/10.1146/annurev.mi.43.100189.001231.
133. Morar M, Wright GD. The genomic enzymology of antibiotic resistance. Annu Rev Genet.
2010;44:25–51. https://doi.org/10.1146/annurev-genet-102209-163517.
134. Forsberg KJ, Reyes A, Wang B, Selleck EM, Sommer MO, Dantas G. The shared antibiotic
resistome of soil bacteria and human pathogens. Science. 2012;337(6098):1107–11. https://
doi.org/10.1126/science.1220761.
135. Nesme J, Cecillon S, Delmont TO, Monier JM, Vogel TM, Simonet P. Large-scale
metagenomic-­ based study of antibiotic resistance in the environment. Curr Biol.
2014;24(10):1096–100. https://doi.org/10.1016/j.cub.2014.03.036.
136. Yoon EJ, Goussard S, Touchon M, Krizova L, Cerqueira G, Murphy C, Lambert T, Grillot-­
Courvalin C, Nemec A, Courvalin P. Origin in Acinetobacter guillouiae and dissemination of
the aminoglycoside-modifying enzyme Aph(3′)-VI. MBio. 2014;5(5):e01972-01914. https://
doi.org/10.1128/mBio.01972-14.
137. Yoon EJ, Goussard S, Nemec A, Lambert T, Courvalin P, Grillot-Courvalin C. Origin in
Acinetobacter gyllenbergii and dissemination of aminoglycoside-modifying enzyme
AAC(6′)-Ih. J Antimicrob Chemother. 2016;71(3):601–6. https://doi.org/10.1093/jac/
dkv390.
138. Babini GS, Livermore DM. Are SHV ß-lactamases universal in Klebsiella pneumoniae?
Antimicrob Agents Chemother. 2000;44(8):2230.
139. Bonnet R. Growing group of extended-spectrum ß-lactamases: the CTX-M enzymes.
Antimicrob Agents Chemother. 2004;48:1–14.
140. Poirel L, Heritier C, Nordmann P. Chromosome-encoded Ambler class D β-lactamase of
Shewanella oneidensis as a progenitor of carbapenem-hydrolyzing oxacillinase. Antimicrob
Agents Chemother. 2004;48(1):348–51.
378 G. A. Jacoby

141. Brown S, Young HK, Amyes SG. Characterisation of OXA-51, a novel class D carbapen-
emase found in genetically unrelated clinical strains of Acinetobacter baumannii from
Argentina. Clin Microbiol Infect. 2005;11(1):15–23.
142. Rottman M, Benzerara Y, Hanau-Bercot B, Bizet C, Philippon A, Arlet G. Chromosomal
ampC genes in Enterobacter species other than Enterobacter cloacae, and ancestral asso-
ciation of the ACT-1 plasmid-encoded cephalosporinase to Enterobacter asburiae. FEMS
Microbiol Lett. 2002;210(1):87–92.
143. Bauernfeind A, Stemplinger I, Jungwirth R, Wilhelm R, Chong Y. Comparative characteriza-
tion of the cephamycinase blaCMY-1 gene and its relationship with other ß-lactamase genes.
Antimicrob Agents Chemother. 1996;40(8):1926–30.
144. Bauernfeind A, Stemplinger I, Jungwirth R, Ernst S, Casellas JM. Sequences of ß-lactamase
genes encoding CTX-M-1 (MEN-1) and CTX-M-2 and relationship of their amino acid
sequences with those of other ß-lactamases. Antimicrob Agents Chemother. 1996;40:509–13.
145. Papanicolaou GA, Medeiros AA, Jacoby GA. Antimicrob Agents Chemother. 1990;34:2200–9.
146. Horii T, Arakawa Y, Ohta M, Ichiyama S, Wacharotayankun R, Kato N. Plasmid-­
mediated AmpC-type ß-lactamase isolated from Klebsiella pneumoniae confers resis-
tance to broad-spectrum ß-lactams, including moxalactam. Antimicrob Agents Chemother.
1993;37(5):984–90.
147. Fosse T, Giraud-Morin C, Madinier I, Labia R. Sequence analysis and biochemical char-
acterisation of chromosomal CAV-1 (Aeromonas caviae), the parental cephalosporinase of
plasmid-mediated AmpC 'FOX' cluster. FEMS Microbiol Lett. 2003;222(1):93–8.
148. Verdet C, Arlet G, Barnaud G, Lagrange PH, Philippon A. A novel integron in Salmonella
enterica serovar Enteritidis, carrying the blaDHA-1 gene and its regulator gene ampR, origi-
nated from Morganella morganii. Antimicrob Agents Chemother. 2000;44(1):222–5.
149. Girlich D, Karim A, Spicq C, Nordmann P. Plasmid-mediated cephalosporinase ACC-1 in
clinical isolates of Proteus mirabilis and Escherichia coli. Eur J Clin Microbiol Infect Dis.
2000;19:893–5.
150. Gudeta DD, Bortolaia V, Jayol A, Poirel L, Nordmann P, Guardabassi L. Chromobacterium
spp. harbour Ambler class A β-lactamases showing high identity with KPC. J Antimicrob
Chemother. 2016;71(6):1493–6. https://doi.org/10.1093/jac/dkw020.
151. Zheng B, Tan S, Gao J, Han H, Liu J, Lu G, Liu D, Yi Y, Zhu B, Gao GF. An unexpected simi-
larity between antibiotic-resistant NDM-1 and beta-lactamase II from Erythrobacter litoralis.
Protein Cell. 2011;2(3):250–8. https://doi.org/10.1007/s13238-011-1027-0.
152. Shaw WV, Hopwood DA. Chloramphenicol acetylation in Streptomyces. J Gen Microbiol.
1976;94(1):159–66. https://doi.org/10.1099/00221287-94-1-159.
153. Hansen LH, Planellas MH, Long KS, Vester B. The order Bacillales hosts functional
homologs of the worrisome cfr antibiotic resistance gene. Antimicrob Agents Chemother.
2012;56(7):3563–7. https://doi.org/10.1128/AAC.00673-12.
154. Kieffer N, Nordmann P, Poirel L. Moraxella species as potential sources of MCR-like poly-
myxin resistance determinants. Antimicrob Agents Chemother. 2017;61(6). https://doi.
org/10.1128/AAC.00129-17.
155. Poirel L, Rodriguez-Martinez JM, Mammeri H, Liard A, Nordmann P. Origin of plasmid-­
mediated quinolone resistance determinant QnrA. Antimicrob Agents Chemother.
2005;49(8):3523–5.
156. Jacoby GA, Griffin CM, Hooper DC. Citrobacter spp. as a source of qnrB alleles. Antimicrob
Agents Chemother. 2011;55(11):4979–84. https://doi.org/10.1128/AAC.05187-11.
AAC.05187-11 [pii]
157. Hooper DC, Jacoby GA. Topoisomerase inhibitors: fluoroquinolone mechanisms of action
and resistance. Cold Spring Harb Perspect Med. 2016;6(9). https://doi.org/10.1101/cshper-
spect.a025320.
158. Albornoz E, Tijet N, De Belder D, Gomez S, Martino F, Corso A, Melano RG, Petroni A.
qnrE1, a member of a new family of plasmid-located quinolone resistance genes, originated
from the chromosome of Enterobacter species. Antimicrob Agents Chemother. 2017;61(5).
https://doi.org/10.1128/AAC.02555-16.
11 Transmissible Antibiotic Resistance 379

159. Yuan J, Xu X, Guo Q, Zhao X, Ye X, Guo Y, Wang M. Prevalence of the oqxAB gene complex
in Klebsiella pneumoniae and Escherichia coli clinical isolates. J Antimicrob Chemother.
2012;67(7):1655–9. https://doi.org/10.1093/jac/dks086.
160. Wachino J, Yamane K, Suzuki S, Kimura K, Arakawa Y. Prevalence of fosfomycin resistance
among CTX-M-producing Escherichia coli clinical isolates in Japan and identification of
novel plasmid-mediated fosfomycin-modifying enzymes. Antimicrob Agents Chemother.
2010;54(7):3061–4. https://doi.org/10.1128/AAC.01834-09.
161. Guo Q, Tomich AD, McElheny CL, Cooper VS, Stoesser N, Wang M, Sluis-Cremer N, Doi
Y. Glutathione-S-transferase FosA6 of Klebsiella pneumoniae origin conferring fosfomycin
resistance in ESBL-producing Escherichia coli. J Antimicrob Chemother. 2016;71(9):2460–
5. https://doi.org/10.1093/jac/dkw177.
162. Arthur M, Brisson-Noel A, Courvalin P. Origin and evolution of genes specifying resistance
to macrolide, lincosamide and streptogramin antibiotics: data and hypotheses. J Antimicrob
Chemother. 1987;20(6):783–802.
163. Marshall CG, Broadhead G, Leskiw BK, Wright GD. D-Ala-D-Ala ligases from glycopep-
tide antibiotic-producing organisms are highly homologous to the enterococcal vancomycin-­
resistance ligases VanA and VanB. Proc Natl Acad Sci U S A. 1997;94(12):6480–3.
164. Hughes J, Mellows G, Soughton S. How does Pseudomonas fluorescens, the producing organ-
ism of the antibiotic pseudomonic acid A, avoid suicide? FEBS Lett. 1980;122(2):322–4.
165. Ford PJ, Avison MB. Evolutionary mapping of the SHV ß-lactamase and evidence for two
separate IS26-dependent blaSHV mobilization events from the Klebsiella pneumoniae chro-
mosome. J Antimicrob Chemother. 2004;54(1):69–75.
166. Toleman MA, Spencer J, Jones L, Walsh TR. blaNDM-1 is a chimera likely constructed in
Acinetobacter baumannii. Antimicrob Agents Chemother. 2012;56(5):2773–6. https://doi.
org/10.1128/AAC.06297-11.
167. Cox G, Stogios PJ, Savchenko A, Wright GD (2015) Structural and molecular basis for
resistance to aminoglycoside antibiotics by the adenylyltransferase ANT(2″)-Ia. mBio 6 (1).
doi:https://doi.org/10.1128/mBio.02180-14.
168. Dunny GM, Berntsson RP. Enterococcal sex pheromones: evolutionary pathways to com-
plex, two-signal systems. J Bacteriol. 2016;198(11):1556–62. https://doi.org/10.1128/
JB.00128-16.
169. Garcillan-Barcia MP, Alvarado A, de la Cruz F. Identification of bacterial plasmids based on
mobility and plasmid population biology. FEMS Microbiol Rev. 2011;35(5):936–56. https://
doi.org/10.1111/j.1574-6976.2011.00291.x.
170. Orlek A, Stoesser N, Anjum MF, Doumith M, Ellington MJ, Peto T, Crook D, Woodford
N, Walker AS, Phan H, Sheppard AE. Plasmid classification in an era of whole-genome
sequencing: application in studies of antibiotic resistance epidemiology. Front Microbiol.
2017;8:182. https://doi.org/10.3389/fmicb.2017.00182.
171. Carattoli A. Resistance plasmid families in Enterobacteriaceae. Antimicrob Agents
Chemother. 2009;53(6):2227–38. https://doi.org/10.1128/AAC.01707-08. AAC.01707-08
[pii]
172. Carattoli A, Zankari E, Garcia-Fernandez A, Voldby Larsen M, Lund O, Villa L, Moller
Aarestrup F, Hasman H. In silico detection and typing of plasmids using PlasmidFinder and
plasmid multilocus sequence typing. Antimicrob Agents Chemother. 2014;58(7):3895–903.
https://doi.org/10.1128/AAC.02412-14.
173. Bertini A, Poirel L, Mugnier PD, Villa L, Nordmann P, Carattoli A. Characterization and
PCR-based replicon typing of resistance plasmids in Acinetobacter baumannii. Antimicrob
Agents Chemother. 2010;54(10):4168–77. https://doi.org/10.1128/AAC.00542-10.
174. Jensen LB, Garcia-Migura L, Valenzuela AJ, Lohr M, Hasman H, Aarestrup FM. A classifi-
cation system for plasmids from enterococci and other Gram-positive bacteria. J Microbiol
Methods. 2010;80(1):25–43. https://doi.org/10.1016/j.mimet.2009.10.012.
175. Lozano C, Garcia-Migura L, Aspiroz C, Zarazaga M, Torres C, Aarestrup FM. Expansion of a
plasmid classification system for Gram-positive bacteria and determination of the diversity of
380 G. A. Jacoby

plasmids in Staphylococcus aureus strains of human, animal, and food origins. Appl Environ
Microbiol. 2012;78(16):5948–55. https://doi.org/10.1128/AEM.00870-12.
176. Garcillan-Barcia MP, de la Cruz F. Ordering the bestiary of genetic elements transmissible by
conjugation. Mob Genet Elements. 2013;3(1):e24263. https://doi.org/10.4161/mge.24263.
177. Hancock SJ, Phan MD, Peters KM, Forde BM, Chong TM, Yin WF, Chan KG, Paterson
DL, Walsh TR, Beatson SA, Schembri MA. Identification of IncA/C plasmid replication
and maintenance genes and development of a plasmid multilocus sequence typing scheme.
Antimicrob Agents Chemother. 2017;61(2). https://doi.org/10.1128/AAC.01740-16.
178. Jacoby GA. Resistance plasmids of Pseudomonas. In: Sokatch JR, editor. The bacteria, vol.
X. The biology of Pseudomonas. Orlando: Academic Press, Inc.; 1986. p. 265–93.
179. Hughes VM, Datta N. Conjugative plasmids in bacteria of the ‘pre-antibiotic’ era. Nature.
1983;302(5910):725–6.
180. Datta N, Hughes VM. Plasmids of the same Inc groups in Enterobacteria before and after the
medical use of antibiotics. Nature. 1983;306(5943):616–7.
181. Partridge SR. Analysis of antibiotic resistance regions in Gram-negative bacteria. FEMS
Microbiol Rev. 2011;35(5):820–55. https://doi.org/10.1111/j.1574-6976.2011.00277.x.
182. Siguier P, Gourbeyre E, Varani A, Ton-Hoang B, Chandler M. Everyman’s guide to bac-
terial insertion sequences. Microbiol Spectr. 2015;3(2):MDNA3-0030-2014. https://doi.
org/10.1128/microbiolspec.MDNA3-0030-2014.
183. Lartigue MF, Poirel L, Aubert D, Nordmann P. In vitro analysis of ISEcp1B-mediated mobi-
lization of naturally occurring β-lactamase gene blaCTX-M of Kluyvera ascorbata. Antimicrob
Agents Chemother. 2006;50(4):1282–6. https://doi.org/10.1128/AAC.50.4.1282-1286.2006.
184. Toleman MA, Bennett PM, Walsh TR. ISCR elements: novel gene-capturing systems of the
21st century? Microbiol Mol Biol Rev. 2006;70(2):296–316.
185. Labbate M, Case RJ, Stokes HW. The integron/gene cassette system: an active
player in bacterial adaptation. Methods Mol Biol. 2009;532:103–25. https://doi.
org/10.1007/978-1-60327-853-9_6.
186. Partridge SR, Tsafnat G, Coiera E, Iredell JR. Gene cassettes and cassette arrays in
mobile resistance integrons. FEMS Microbiol Rev. 2009;33(4):757–84. https://doi.
org/10.1111/j.1574-6976.2009.00175.x.
187. Guillard T, Grillon A, de Champs C, Cartier C, Madoux J, Bercot B, Lebreil AL, Lozniewski
A, Riahi J, Vernet-Garnier V, Cambau E. Mobile insertion cassette elements found in
small non-transmissible plasmids in Proteeae may explain qnrD mobilization. PLoS One.
2014;9(2):e87801. https://doi.org/10.1371/journal.pone.0087801.
188. Guglielmini J, Quintais L, Garcillan-Barcia MP, de la Cruz F, Rocha EP. The repertoire of
ICE in prokaryotes underscores the unity, diversity, and ubiquity of conjugation. PLoS Genet.
2011;7(8):e1002222. https://doi.org/10.1371/journal.pgen.1002222.
189. Fournier PE, Vallenet D, Barbe V, Audic S, Ogata H, Poirel L, Richet H, Robert C, Mangenot
S, Abergel C, Nordmann P, Weissenbach J, Raoult D, Claverie JM. Comparative genomics
of multidrug resistance in Acinetobacter baumannii. PLoS Genet. 2006;2(1):e7. https://doi.
org/10.1371/journal.pgen.0020007.
190. Roy Chowdhury P, Scott M, Worden P, Huntington P, Hudson B, Karagiannis T, Charles
IG, Djordjevic SP. Genomic islands 1 and 2 play key roles in the evolution of extensively
drug-resistant ST235 isolates of Pseudomonas aeruginosa. Open Biol. 2016;6(3). https://doi.
org/10.1098/rsob.150175.
191. Colomer-Lluch M, Jofre J, Muniesa M. Quinolone resistance genes (qnrA and qnrS) in bac-
teriophage particles from wastewater samples and the effect of inducing agents on pack-
aged antibiotic resistance genes. J Antimicrob Chemother. 2014;69(5):1265–74. https://doi.
org/10.1093/jac/dkt528.
192. Quiros P, Colomer-Lluch M, Martinez-Castillo A, Miro E, Argente M, Jofre J, Navarro
F, Muniesa M. Antibiotic resistance genes in the bacteriophage DNA fraction of human
fecal samples. Antimicrob Agents Chemother. 2014;58(1):606–9. https://doi.org/10.1128/
AAC.01684-13.
11 Transmissible Antibiotic Resistance 381

193. Rice LB, Carias LL, Hutton RA, Rudin SD, Endimiani A, Bonomo RA. The KQ element,
a complex genetic region conferring transferable resistance to carbapenems, aminoglyco-
sides, and fluoroquinolones in Klebsiella pneumoniae. Antimicrob Agents Chemother.
2008;52(9):3427–9. https://doi.org/10.1128/AAC.00493-08. AAC.00493-08 [pii]
194. Gillings MR, Paulsen IT, Tetu SG. Genomics and the evolution of antibiotic resistance. Ann
N Y Acad Sci. 2017;1388(1):92–107. https://doi.org/10.1111/nyas.13268.
195. Liebert CA, Hall RM, Summers AO. Transposon Tn21, flagship of the floating genome.
Microbiol Mol Biol Rev. 1999;63(3):507–22.
196. Jacoby GA, Munoz-Price LS. The new ß-lactamases. N Engl J Med. 2005;352(4):380–91.
197. Shaw KJ, Poppe S, Schaadt R, Brown-Driver V, Finn J, Pillar CM, Shinabarger D, Zurenko
G. In vitro activity of TR-700, the antibacterial moiety of the prodrug TR-701, against
linezolid-resistant strains. Antimicrob Agents Chemother. 2008;52(12):4442–7. https://doi.
org/10.1128/AAC.00859-08.
198. Arias CA, Mendes RE, Stilwell MG, Jones RN, Murray BE. Unmet needs and prospects for
oritavancin in the management of vancomycin-resistant enterococcal infections. Clin Infect
Dis. 2012;54(Suppl 3):S233–8. https://doi.org/10.1093/cid/cir924.
199. Shen Z, Ding B, Bi Y, Wu S, Xu S, Xu X, Guo Q, Wang M. CTX-M-190, a novel β-lactamase
resistant to tazobactam and sulbactam, identified in an Escherichia coli clinical isolate.
Antimicrob Agents Chemother. 2017;61(1). https://doi.org/10.1128/AAC.01848-16.
200. Livermore DM, Mushtaq S, Warner M, Vickers A, Woodford N. In vitro activity of cefepime/
zidebactam (WCK 5222) against gram-negative bacteria. J Antimicrob Chemother.
2017;72(5):1373–85. https://doi.org/10.1093/jac/dkw593.
201. Shields RK, Chen L, Cheng S, Chavda KD, Press EG, Snyder A, Pandey R, Doi Y,
Kreiswirth BN, Nguyen MH, Clancy CJ. Emergence of ceftazidime-avibactam resistance
due to plasmid-­borne blaKPC-3 mutations during treatment of carbapenem-resistant Klebsiella
pneumoniae infections. Antimicrob Agents Chemother. 2017;61(3). https://doi.org/10.1128/
AAC.02097-16.
202. Davies J, Davies D. Origins and evolution of antibiotic resistance. Microbiol Mol Biol Rev.
2010;74(3):417–33. https://doi.org/10.1128/MMBR.00016-10. 74/3/417 [pii]
203. Bates J, Jordens JZ, Griffiths DT. Farm animals as a putative reservoir for vancomycin-­
resistant enterococcal infection in man. J Antimicrob Chemother. 1994;34(4):507–14.
204. Kunin CM, Tupasi T, Craig WA. Use of antibiotics. A brief exposition of the problem and
some tentative solutions. Ann Intern Med. 1973;79(4):555–60.
205. Barlam TF, Cosgrove SE, Abbo LM, MacDougall C, Schuetz AN, Septimus EJ, Srinivasan A,
Dellit TH, Falck-Ytter YT, Fishman NO, Hamilton CW, Jenkins TC, Lipsett PA, Malani PN,
May LS, Moran GJ, Neuhauser MM, Newland JG, Ohl CA, Samore MH, Seo SK, Trivedi
KK. Implementing an antibiotic stewardship program: guidelines by the Infectious Diseases
Society of America and the Society for Healthcare Epidemiology of America. Clin Infect Dis.
2016;62(10):e51–77. https://doi.org/10.1093/cid/ciw118.
Chapter 12
Antibiotics and Resistance
in the Environment

Marilyn C. Roberts

12.1 Introduction

The discovery and use of antibiotics was one of the greatest public health achieve-
ments of the twentieth century. Antibiotics have saved millions of human and ani-
mal lives, reduced agricultural losses, and contributed to increased food production.
These agents have extended the lives of people with genetic conditions and have
become indispensible in modern medicine. The majority of antibiotics currently in
use were originally produced by living microbes that were then modified by man.
Antibiotics either inhibit growth of other microbes or kill them by interacting with
specific microbial targets. Most of the targets are unique to microbes, which has led
to the agents being safe enough to use with eukaryotic organisms.
In the mid-twentieth century, antibiotics became the foundation for treating bac-
terial infections in both humans and animals. Antibiotic-resistant bacteria [ARB]
and antibiotic resistance genes [ARGs] were recognized within a year after penicillin
was first used in humans, and soon after it was seen with agricultural use [1, 2].
ARB infections now contribute to thousands of deaths each year plus increased
morbidity and medical cost. Currently, it is estimated that ~10 million deaths due to
antibiotic-resistant infections occur each year; this number is expected to rise in
coming years [3]. In essence, antibiotic resistance has changed treatable infections
into untreatable diseases, thereby moving us closer to the “post-antibiotic era.”
Multidrug-resistant pathogens were first identified in the 1950s [4]. ARB were
initially limited to hospital settings and few outbreaks occurred; ARB were not seen
as a major concern for general community medicine. Today it is known that
antibiotic use in humans and agriculture results in increased antibiotic resistance in

M. C. Roberts (*)
Department of Environmental and Occupational Health Sciences, School of Public Health,
University of Washington, Seattle 98195, WA, USA
e-mail: [email protected]

© Springer International Publishing AG, part of Springer Nature 2018 383

I. W. Fong et al. (eds.), Antimicrobial Resistance in the 21st Century, Emerging Infectious
Diseases of the 21st Century, https://doi.org/10.1007/978-3-319-78538-7_12
384 M. C. Roberts

all types of bacteria, ranging from pathogenic to environmental species. A major

paradigm shift occurred in the 1970s, with the identification of ampicillin-resistant
Haemophilus influenzae and penicillin-resistant Neisseria gonorrhoeae, both
community-acquired pathogens. Resistance to the preferred therapy led to changes
in the recommended therapies for disease arising from these pathogens. The need
for monitoring ARB and ARGs and periodic changes of first-line therapies has
become an ongoing issue for many different pathogens. Resistance has also lead to
a new industry of diagnostics in which new methods and techniques continue to be
developed for rapid identification of resistance in clinically important bacteria.
In the past, surveillance of the environment locally, nationally, and internation-
ally has not been a priority, but that has changed [5], as we are beginning to examine
the issue and assess the impacts of ARB/ARGs on human and animal health, agri-
cultural and food production, aquaculture, human and animal waste management,
and the impact and contamination of the environment globally [6–11].
Antibiotic uses and abuses are directly responsible for the increases in the level
of ARB and ARGs isolated in agricultural as well as aquacultural settings, the food
chain, man, and built and natural environments [12–14]. Much has been said about
the uses of antibiotics as growth promoters in Europe and the USA as being a major
source of antibiotic resistance. In June 2015, the US Food and Drug Administration
published a final rule known as the veterinary feed directive (https://www.gpo.gov/
fdsys/pkg/FR-2015-06-03/pdf/2015-13393.pdf), which limits the use of antibiotics
as feed additives for growth promotion. The rule became effective on October 1,
2015, and may have widespread impact on use and prescribing of medically
important antibiotics in food animals, both in the years leading up to implementation
and after implementation (Jan 2017).
In the early years of antibiotic usage, there were new antibiotics available to
replace the older antibiotics as bacteria became resistant. Thus, when one antibiotic
failed to work, another was available to take its place. Today there are very few new
antibiotics in development to replace the less effective, older antibiotics [3]. The
current lack of new and novel antibiotics coming into the market, along with the
high cost of newer antibiotics, has led researchers to anticipate a time when there
will be no useable antibiotic for many common bacterial diseases. Thus, animals,
plants, and people will die of infections that were once easily treated with antibiotics
but are now resistant to all available therapies [11]. The factors that contribute to
emergence and dissemination of bacterial resistance are complex and require
attention in both industrialized and developing countries [12, 13].
Concerns over the spread of antibiotic resistance have stimulated several groups
to assess the impact of ARB/ARGs on human and animal health, agricultural and
food production, and agricultural and human waste management [15, 16]. One of the
primary outcomes of emerging reports is a call for increased surveillance of ARB
and ARGs in agricultural and environmental settings, with a particular interest in
identifying transmission routes of ARB and ARGs throughout the world [11, 15].
Keys to the success of current and future surveillance efforts are strategies to deter-
mine which types of resistant genes to monitor and how to support the surveillance
effort, especially for environmental settings and in developing and resource-poor
12 Antibiotics and Resistance in the Environment 385

countries. This is a major task given that in the USA there is no national surveillance
program for the most common pathogens across most states. Instead, the Centers for
Disease Control and Prevention (CDC) has used surveillance systems that focus on
nine representative states [17]. The European Union does a more comprehensive job
of covering their member states (http://ecdc.europa.eu/en/healthtopics/antimicro-
bial-resistance-and-consumption/antimicrobial_resistance/EARS-Net/Pages/
EARS-Net.aspx); other parts of the world have varying success with human surveil-
lance systems [17]. The problem is difficult, because ARGs are not randomly distrib-
uted among bacterial species. Data suggests that a clear link exists between bacterial
taxonomy and specific types of ARG [18, 19]. This phenomenon has been particu-
larly well documented with tetracycline resistance genes [20–22].
The environmental dissemination of ARGs and the development of ARGs are
thought to be primarily due to horizontal gene transfer. The most common way
bacteria exchange ARGs is by conjugation, which allows rapid transfer of ARGs
between species and genera within and between ecosystems [21]. However, our
knowledge is limited in regard to how the environment contributes to transmission
between the environment, wildlife, domesticated animals, plants, and humans. It is
critical when examining specific antibiotic resistance genes to know whether a
given gene is normally associated with a mobile element and whether that element
has a narrow or broad host range. Clearly a mobile element with a broad host range
will allow for wider transmission across multiple genera than a narrow host range
element [23, 24]. It is important to identify the specific ARGs associated with
specific bacterial species and/or genera within the environment. Durso et al. [18]
suggested that the same antibiotic resistance gene might have different risks for
environmental transmission that depends on the specific bacterial taxa within which
it is found. For example, if the bacteria are widely distributed among a variety of
environments, the ARGs associated with them are more likely to spread widely. If,
on the other hand, the bacteria have a limited environmental range, the ARGs will
tend to remain associated with them specifically. If they have a limited host range,
they may also not be widely distributed. It is equally important to know how ARGs
and ARB are distributed among human and animal populations and how these
ecosystems interact with various environments. Moreover, we need to know how
microbial distribution differs by region, nation, and worldwide [25]. Other issues
include the fact that most environmental studies look at a selected group of ARGs
by qPCR, which determines the presence or absence of particular genes [26], or
they use microbiome studies that usually do not look specifically for selected ARGs
[25]. Thus, environmental studies should include bacterial culturing, in addition to
molecular studies, to fully understand the distribution within the bacterial ecosystem
of various environments. The more comprehensive analysis is especially important
because many of the new ARGs are coming from the environment rather than from
either human or animal sources, which makes it difficult to know the bacterial
source of a given ARG (http://faculty.washington.edu/marilynr/).
Organized environmental surveillance of ARGs/ARB will hopefully allow iden-
tification of major gaps in our understanding of the forces that act on selection and
transmission of bacterial resistance. This effort in turn may lead scientists in direc-
386 M. C. Roberts

tions that could either slow or stop the march to a time when common infections and
minor injuries kill, as they did prior to the introduction and widespread use of anti-
biotics (this phenomenon is well illustrated by the recent spread of NDM-1
β-lactamase carrying bacteria [27]). It is clear that a global “One Health” approach
is needed in which animal and human usage and environmental contamination are
considered together, along with an understanding of how ARGs and ARB move
between the ecosystems.

12.2  ntibiotics Used for Conditions Due to Non-bacterial

A
[Noninfectious] Conditions

Some antibiotics have non-bacterial effects on humans and animals and have been
used to treat non-bacterial conditions, especially skin diseases. A review of the non-­
antibiotic properties of minocycline by Garrido-Mesa et al. [28] is a useful guide to
other properties that this antibiotic has and the non-bacterial conditions for which
minocycline is used as treatment. A 2013 paper [29] reported that minocycline
improves symptoms of fragile X syndrome when given to children and adolescents.
Another study explored the use of tetracyclines with cancer targets through a
randomized phase II trial [30]. In a third case, the macrolide azithromycin stimulated
immune and epithelial cell modulation of transcription factors AP-1 and NFκB with
subsequent delayed inhibitory effects on cell function and may cause lysosomal
accumulation of the macrolide with disruption of protein and lipid transport through
the Golgi apparatus and effects the surface receptor expression, including
macrophage phenotype changes and autophagy [31]. In addition, azithromycin
inhibits quorum sensing and biofilm formation by Pseudomonas aeruginosa, even
though the drug does not inhibit growth. Moreover, azithromycin, given
prophylactically, can reduce the incidence of ventilation-associated pneumonia
[31]. It is important to note that the use of antibiotics for non-bacterial conditions
increases the exposure of individuals’ microbiomes to the selective pressures
underlying the emergence of bacterial resistance (M. Roberts unpublished results).
It also increases the potential for environmental contamination by the antibiotic, its
residues, and ultimately selection of ARGs and ARB resistant to these antibiotics.

12.3 The One Health Approach

The One Health approach contrasts with the traditional practice of human and
animal medicine, which have been studied and practiced in isolation rather than as
part of an ecosystem. The environmental contribution to global health has not been
generally considered, or if studied, rarely, until recently in connection with the
health of man and/or animals. The world microbial ecosystem includes the micro-
biomes associated with each domain of life and the direct and indirect mixing of
12 Antibiotics and Resistance in the Environment 387

the different microbiomes that, in some cases, may lead to disease. The human-­
animal interface is ancient, but it has expanded with the development of farming
animals and fish. It is a continuum of contacts and interactions that allow for bar-
rier breaches of pathogens to occur and an increased driver of infections. This is
illustrated by the estimation that ~75% of emerging infectious diseases in humans
over the last 20 years have been zoonotic, i.e., the pathogen spread from animals
or insects to people. In some cases, the pathogen becomes established and then
spreads within human populations. However, more commonly, there are recurrent
events of transmission from an animal/insect reservoir to humans, with limited
human-to-­human transmission. An example of this situation is observed with the
Zika virus [32, 33]. Other examples include many foodborne bacterial infections,
such as those caused by E. coli O157:H7 and enterotoxigenic E. coli O114:H4. E.
coli O114:H4 caused a huge outbreak in 2011, which, besides causing death and
infections, created tension among EU members involving boycotts of vegetables
within the EU [34]. Dealing with emerging and reemerging infections that cross
species barriers not only impacts humans but also impacts livestock, pets, wildlife,
crops, and aquaculture. These pathogens can contaminate the environment, and in
worse cases, they may impact food resources and food security. The importance of
global ecological changes due to human impact on the environment and techno-
logical changes in society, along with important changes in how food is produced,
processed, and transported, combines to increase the potential risk of disease
transmission [32]. With environmental contamination as a major by-product of
these endeavors, changing the downward spiral of increasing global contamination
can only be addressed by improved communication, cooperation, and collabora-
tion across disciplines and the realization that there are multiple ways contamina-
tion can enter the food chain.
How a particular antibiotic can influence where and how antibiotic resistance
and ARB develop and spread from one domain to all three has been illustrated in the
literature. One good example is the development of vancomycin-resistant
Enterococcus faecium (VRE) in North America, the EU, and the rest of the world.
In Europe and other parts of the world, a vancomycin-related drug avoparcin was
used as a growth promoter in livestock. Over time, VRE developed in chickens and
swine to where it could be readily detected in processed meat [35]. Transmission of
VRE genes, or the intact bacterium, from animals to humans occurs in the EU
setting. Once VRE was established in livestock populations, farmers and those
slaughtering the VRE+ animals acquired VRE in their intestinal tracts. VRE
ultimately was isolated from hospitalized persons [36]. In contrast, avoparcin was
never used as a food additive in the USA. Early studies suggested that VRE was not
found in chickens in the USA, and there was little evidence to suggest transmission
of VRE in healthy adults prior to 2000 [35]. In contrast to the EU, which did not use
vancomycin heavily in the hospital setting, vancomycin was used extensively in US
hospital. The result was the emergence of VRE as a major nosocomial pathogen
within US hospitals [38]. This was due, in part, to the persistence of viable VRE on
contaminated surfaces within the hospital for weeks and even months. Rooms
housing patients colonized or infected with VRE were difficult to clean.
388 M. C. Roberts

Consequently, these rooms served as reservoirs for transmission of VRE to new

patients [37]. More recently, VRE has been cultured from the general community
environment in the USA, as illustrated by VRE recovered from wild crows and
recreational beach sand and water in North America [39–41]. Currently, if a patient
enters a hospital and exhibits a VRE infection within 48 h of entrance, the infection
is considered community acquired rather than nosocomial. The occurrence of
outpatient VRE depends on geographic location, occupation of the people,
differences between urban or rural settings, and/or recent attendance at a medical/
dental outpatient clinic or office. VRE in the USA spread from hospitalized humans
to the community and environment, while in the EU and other parts of the world,
VRE developed in livestock receiving avoparcin and then spread to the farm
workers, local communities, and ultimately hospitalized patients.

12.4 The Environment

Most studies on ARGs, over the last 70 years, have focused on clinically important
bacteria found in humans and animals. It is estimated that there are ~5 × 1030 bacteria
on earth, with only a small subset adapted to live either in or on humans and animals.
More striking is the estimate that <1% of the total number of bacteria in the world
have been cultured [42]. The natural world is rich in chemicals made by living
organisms and human activity – antibiotics are not the only compounds that have
influenced the evolution of microbiomes [43].
As stated above, knowing which type of bacterium carries a particular ARG can
be critical in designing studies of the environment. For example, many in the field
use E. coli as a model system for ARG carriage. Yet many ARGs found in E. coli are
unique to Gram-negative facultative aerobes and not found in anaerobes, other
Gram-negative bacteria, or Gram-positive species [22]. Thus, when E. coli is the
model, most acquired ARGs are associated with plasmids that independently
replicate and tend to have a host range limited to Gram-negative bacteria. In contrast,
many ARGs in Gram-positive and anaerobic species are on mobile elements that are
normally found in the chromosome; thus, they have a different host range dynamics
that can be much broader than classical Gram-negative plasmids. Therefore, by
looking for both classical Gram-negative and Gram-positive ARGs, researchers can
select ARGs that are likely to be most important for a particular ecosystem from a
large set that confers resistance to the same class of antibiotic. This is especially
important when molecular methods of detection are used, because only a limited
number of genes can be assayed, and if the rare genes are chosen, it will bias the
results leading to an unrealistic picture in that ecosystem.
Environmental studies are moving away from culturing bacteria, instead of
determining which ARGs are present by using either PCR or qPCR. These molecular
assays have now been used for direct detection of ARGs in food [44], animal feeding
facilities, and agricultural soils amended with manure [45] and as indicators for
water quality changes [46]. In these studies, known ARGs were used without
12 Antibiotics and Resistance in the Environment 389

determining their likely distribution in the particular sample source, which can lead
to biased results. For example, if only ARGs present in Gram-negative bacteria are
used for screening, then no information will be obtained regarding the Gram-­
positive and anaerobic component of the sample source. Such studies have a very
limited ability to identify novel resistance genes.
To overcome the shortcomings of nucleotide sequence-dependent methods, soil
bacteria were screened for the ability to degrade or inactivate antibiotics. In one
study, strains were randomly isolated from 11 diverse rural and urban soils, and
they were then tested for the ability to utilize 18 different antibiotics as sole sources
of carbon and nitrogen [47]. Many of the bacteria were Burkholderia spp. and
Pseudomonas spp., which are naturally part of the soil microbiome and only rarely
cause disease. These bacteria could grow on antibiotic-supplemented media and
were resistant to multiple antibiotics at clinically relevant concentrations, suggest-
ing the presence of an unappreciated reservoir of antibiotic resistance genes in
these soils [47]. This work led to development of the functional metagenomic
approach (described below) in which the antibiotic resistome of an environment is
examined. This has led to identification of novel resistance genes in addition to
know ARGs [22].
Functional genomic studies have also been used to study a variety of microbial
environments [48]. This assay determines whether the cloned DNA can be expressed
and confer resistance when transferred into a host E. coli. When the cloned DNA
allows the host bacteria to grow in the presence of antibiotic-supplemented media,
the resistance-conferring DNA fragments can be sequenced and compared to known
ARGs. The Antibiotic Resistance Genes Database now lists ~20,000 potential
resistance genes( [49]; http://ardb.cbcb.umd.edu/), while the Comprehensive
Antibiotic Resistance Database [CARD] also has a large number of resistance genes
that can be used to screen sequences [50]. A variety of new potential ARGs have
been identified using this method ([51]; http://faculty.washington.edu/marilynr/).
One issue with these databases is they rely on GenBank information, which can be
confusing and inaccurate because the system allows authors to name their own
genes rather than going through a system. They also make it difficult to change
names. Thus many of the ARGs in GenBank do not have the correct nomenclature
for specific ARGs. Consulting other sources such as http://faculty.washington.edu/
marilynr/ and http://www.lahey.org/studies/ should be used to get the correct names
for tetracycline, for macrolide-lincosamide-streptogramin genes, and for β-lactamase
genes, respectively. Recent reviews are also good sources for the current
nomenclature [22].
The term antibiotic “resistome” is defined as the collection of all genes that can
either directly or indirectly contribute antibiotic resistance to its bacterial host [52].
Research groups have been examining the microbial resistome of natural and
clinical environments [51, 53–55]. Studies have looked for ARGs in samples linked
to human activity, such as food production [56, 57], polluted waterways, and
wastewater treatment plants [26]. Resistome studies suggest that environmental
bacteria may be antibiotic resistant by virtue of both previously characterized,
known genes and unknown resistance genes, mutations, and resistance genes on
390 M. C. Roberts

mobile elements [55, 56]. A number of new ARGs have been identified from these
studies; in most cases, the bacterial host is unknown (http://faculty.washington.edu/
marilynr/). One example of a ARG identified in a molecular study is the tet(43),
which encodes an efflux protein. It was isolated from metagenomic analysis of soil
taken from an apple orchard that had been repeatedly treated with streptomycin
[56]. It is unknown what type of bacterium actually carries tet(43), and little is
known about the distribution of this gene. Similarly, the nine genes [tet(47)-tet(55)],
identified more recently, code for enzymes that inactivate tetracycline. These were
identified, cloned, and sequenced from soil samples where functional metagenomic
analysis was done [58]. Again the bacterial sources of each of the genes are not
known. This later work increased the number of characterized enzymes that
inactivate tetracycline from 3 to 13 (http://faculty.washington.edu/marilynr/),
clearly showing that a variety of new ARGS may be present in environments.
Many bacteria, including environmental bacteria, encode β-lactamases, which
hydrolyze and inactivate β-lactam antibiotics (http://www.lahey.org/studies/). They
are the most widely distributed of all ARGs [3]. One example is ampC, which was
originally an inducible chromosomal cephalosporinase found in a variety of
Enterobacteriaceae. This gene has been found in opportunistic pathogens belong-
ing to normal intestinal floral of humans and animals, in bacterial species that nor-
mally live in either soil or water, and in both pathogenic and nonpathogenic bacteria
[59, 60]. It has been proposed that ampC originated in environmental bacteria. The
first AmpC-positive clinical strains were E. coli isolated in the 1940s, just as the
first antibiotics were being developed and used. In a host background that has porin
deficiencies, ampC, when expressed, confers carbapenem resistance due to
increased production of the AmpC β-lactamase. This increased production of the
AmpC β-lactamase is usually due to mutations that up-regulate expression of the
enzyme. Today the chromosomal AmpC β-lactamases are associated with plas-
mids, which was first noticed in the 1980s. These mobile plasmids often tend to be
large, and they carry multiple ARGs. Plasmid-mediated AmpC β-lactamases have
greatly expanded the host range of this group of enzymes that are now found in
epidemic human pathogens such as E. coli ST131. This E. coli strain has been iso-
lated from fresh vegetables, food-producing animals, fish farms, pets, and water
environments [61, 62].
Many ARGs are associated with soil antibiotic producers such as Streptomyces.
Some of these natural ARGs have the same mode of action as those found in
clinically resistant bacteria [3]. In the past, it was assumed that most environmental
bacteria were poorly adapted for life in humans and animals. However this idea is
changing, as progress in medical science allows severely immunocompromised
patients to live in the community where they can be infected with environmental
organisms. Other susceptible persons include those who have foreign objects
permanently implanted in their bodies and persons with various types of occupational
exposure [63]. Moreover, the distinction between environmental and non-­
environmental bacteria has become difficult, because the mixing of the two sources
of bacteria has become increasingly common as human and agricultural
contamination of the environment has become widespread. Indeed, very few
12 Antibiotics and Resistance in the Environment 391

ecosystems around the world have not been touched by the activities of human
civilization – whether it is in polar regions or the Amazon jungle [64, 65]. The
continual mixing of environmental and non-environmental bacteria provides
opportunities for horizontal genetic exchange of ARGs between man, animal, and
environmental bacteria.
Antibiotics, antibiotic residues, ARBs, and ARGs move by water and wind [66],
wastewater treatment discharges [26, 67], biosolids, and manure applications [68],
isolated from recreational beaches [40]. They are also moved along with the
transportation of goods and people around the world [69, 70]. One result of this
movement has been the global spread of specific strains, such as Clostridium difficile
NAP1/027/BI [71]. C. difficile spores are robust, and they can survive in hospital
dust for extended time periods. C. difficile was originally classified as a nosocomial
pathogen. Today C. difficile is known to be a foodborne and community pathogen.
Similarly, 25 years ago, Acinetobacter baumannii was a rarely identified human
pathogen. At that time, Acinetobacter spp. were primarily found in the environment.
They are well adapted to grow at a wide range of temperature and pH values, can
use a variety of carbon and energy sources, and persist in both moist and dry
locations for extended time periods. Today multidrug-resistant A. baumannii is
considered an opportunistic pathogen that has become a major concern for military
trauma patients and causes infections that are very difficult to treat due to limited
therapeutic options [72].

12.5 Agriculture

Antibiotics are used for both human and agricultural activities for prevention and
treatment of infections. They are also used as food additives and growth promot-
ers in food production in the USA. However this widespread use is changing. In
June 2015, the US Food and Drug Administration published a final rule, known as
the veterinary feed directive (VFD), that extends the use of veterinary feed direc-
tives to an increased number of medically important antimicrobials used in food
animal production [73]. The rule became effective on October 1, 2015, and may
have impacted use and prescribing of medically important antibiotics in food ani-
mals in years prior to implementation because evidence supporting this idea is
derived from the experience of the EU. On July 1, 1989, an EU-wide ban on the
use of four growth-promoting antibiotics, spiramycin, tylosin, bacitracin zinc,
and virginiamycin, came into effect. The result of this ban was a dramatic drop in
the sales of antimicrobial growth-promoting agents. In 2006, the remaining anti-
biotics used as growth promoters (monensin, avilamycin, salinomycin, and flavo-
mycin) came under an EU-wide ban. It is projected that a further dramatic
decrease in sales of growth promoters will occur [74]. Therefore, it is hoped that
the US FDA ruling will reduce overall uses of antibiotics used annually in live-
stock raised in the USA.
392 M. C. Roberts

Antibiotics can be found in domestic animal manure, which may be transferred

when manure is applied to fields or stored in lagoons. In the USA, manure, regardless
of whether the host animals are treated with antibiotics or not, is considered an
organic product. Domestic animal manure can be placed on crops that will be grown
organically. In addition, “organic farms” are usually on land that was originally
farmed conventionally. Therefore, antibiotics, antibiotic residues, ARGs, and ARB
are normally present in the farm environment. The ARB can colonize the “organic”
livestock, while the ARGs may be incorporated into the livestock microbiome. As a
result, both organic and conventionally grown meats may have ARB [75].
Antibiotics are sprayed onto crops which contaminates the surrounding soil,
sediment, and groundwater. This practice exerts selective pressure on the associated
microbiomes and increases the prevalence of resistance in bacterial pathogens of
fruit orchards. Antibiotics may also be incorporated into food given to farm animals
and fish, which will, in turn, contaminate the surrounding area and ultimately enter
the water system.
Antibiotics from human therapeutic use, especially from hospital effluents, are a
continual source of pollution and are considered part of the “emerging contaminants”
in municipal waste (concentrations of tetracycline vary from ng/L to μg/L). At these
levels, antibiotics may select for tetracycline-resistant environmental bacteria
which, once present, may persist in the environment for long times. The environment
may become a reservoir for tetracycline resistance genes and for other antibiotic
resistance genes, since co-selection with other ARGs is common. Antibiotic-­
resistant bacteria and residues have been identified in tap water, urban water
supplies, milk, meat, vegetables, and processed and unprocessed foods [76]. All of
these sources contaminate both built and natural environments, either directly or
indirectly, and provide selective pressure on the resident environmental bacteria to
become antibiotic resistant [3]. In some cases, transfer of specific antibiotic
resistance genes is increased with exposure to low levels of antibiotics [77].

12.6 Human Influences on Environments

Human activity may directly influence the development of ARBs in built-up envi-
ronments. For example, several studies have recovered antibiotic-resistant E. coli
and S. aureus from air in homes that are enriched relative to samples from outside
of the home, even though the latter have higher bacterial levels. However, there was
variability both in study design and results [66]. Potentially, ARB may contaminate
the environment either directly, as occurs when manure is applied to enrich agricul-
tural fields, or indirectly due to sewage contamination of receiving waters where
the final effluent is deposited such as a river, lake, or ocean. The first description of
the tet(M) gene in Bacillus spp. and of TcrBacillus cereus strains carrying the
tet(M) gene, on a Tn916 mobile element, was found in animal manure and in fields
where the manure was spread. These results suggested that the presence of tet(M)-
carrying B. cereus in fields was a direct result of manure application to the soil.
12 Antibiotics and Resistance in the Environment 393

Whether tet(M)-carrying B. cereus will act as a donor and transfer the tet(M) gene
to either related B. anthracis or B. thuringiensis is unknown. However, some toxin-
encoding plasmids are shared among these three species [68].
An example of human wastes increasing ARB was illustrated by a 1980’s study
that observed three groups of wild baboons in Kenya. Two of the groups lived in
their natural habitat with either limited or no human contact; these groups had low
levels of antibiotic-resistant Gram-negative enteric bacteria. The third group lived
close to a tourist lodge that provided opportunities for daily contact with unprocessed
human refuse. From these animals, high levels of antibiotic-resistant Gram-negative
enteric bacteria were recovered with >90% tetracycline resistant [Tcr]. These results
suggested that contact with human refuse greatly increased the carriage of Tcr
bacteria in these wild primates [78]. Unfortunately, the surrounding environmental
bacteria were not sampled. One could speculate that the level of environmental
ARB was likely higher around the human refuse site than in areas where the two
other baboon groups lived in a more natural setting. Other studies have recovered
antibiotic-resistant E. coli from arctic and subarctic seals [79], wild boars [80], and
wild rabbits [81]. More recently, bacteria carrying extended-spectrum beta-­
lactamases (ESBLs) have been isolated from water birds in remote locations [82].
Birds and wild animals can also be found feeding either in or around wastewater
treatment ponds, waste landfill sites, and septic tank discharges. Birds have the
potential for long-distance dissemination of ARB and ARGs to remote environments.
Such transmission sources may explain why ARB and ARGs can be found in
environments having little anthropogenic activity, such as the remote arctic [66].
In many studies it has been assumed that ARG flowed from humans and animals
to the environment. But in other cases, the use of antibiotics for food production has
created antibiotic-resistant bacteria in animals and farm environment that has spread
to man. One classic example of animal-to-human spread is the use of avoparcin in
farm animals in the EU [83]. Vancomycin-resistant enterococci [VRE] develop on
these farms, contaminating the farm ecosystem, including animal, environmental,
and human microbiomes. The VRE strains were passed to farm workers and families
living on the farm. In other cases, the plasmids carrying the vanA/vanB genes were
transmitted from animal to human enterococci [40]. In contrast, VRE development
in hospital settings in North America has occurred because vancomycin was
commonly used in hospitalized individuals but not in the general community
population. More recently, VRE strains have spread to the environment in the USA
where they are now isolated in a variety of settings, from recreational beaches to
birds to farms [39, 40, 84].

12.7 Aquaculture

As the taste for seafood and shellfish increases, the use of aquaculture around the
world, especially in Asia, has increased. Integrated aquaculture is a traditional
practice used by small-scale farmers in Asia. The fish are raised in ponds along
394 M. C. Roberts

with livestock. The livestock manure is used to feed the fish. This system allows
for mixing of ARGs and ARB, as well as for creating recombinant influenza
viruses [85]. Other parts of the world are less likely to practice integrated aquacul-
ture. Varying sizes of fish farms, both of the fresh water and marine type, grow
many types of fish for global export. Tilapia (Oreochromis niloticus) is among the
most cultured and internationally traded food fish, with an estimated 1.45 million
tons produced in China in 2013 [85].
ARGs are enriched in sediments below fish farms in Finland, even though selec-
tive pressure from antibiotics was low. A new study, which looked at 364 PCR
primer sets for detecting ARGs, mobile genetic elements, and 16S rRNA genes,
detected 28 genes in fish feces and fish farm sediments. The ARGs included
aminoglycoside (aadA1, aadA2), chloramphenicol (catA1), macrolide (mef(A),
msr(A)), sulfonamide (sul1), trimethoprim (dfrA1), and tetracycline ribosomal
protection genes [tet(32), tet(M), tet(O), tet(W)]. The same ARGs were found in fish
feces, suggesting that fish contribute to the ARG enrichment of the farm sediments
even though no antibiotic treatment of the fish in the farms was performed. Individual
farms had their own unique resistome compositions [86]. The Baltic Sea has no tide,
and water circulation is slow; thus, ARGs in the sediment underneath the fish pens
and up to 200 m from the fish farms were expected to reflect activity in the farm.
Muziasari et al. [86] concluded that their findings provide indirect evidence for the
hypothesis that selected ARGs are introduced into the sediment underneath fish
farms in the Northern Baltic Sea by farmed fish. The antibiotic concentrations in the
sediments were ~1–100 ng/g of sediment.
Tetracyclines have been used extensively in aquaculture, and Tcr bacteria have
been characterized from numerous sources, including fish pathogens and envi-
ronmental bacteria associated with finfish aquaculture from around the world
[87–91]. Tcr bacteria can be found in fish feed, in the sediment under the fish
pens, as well as in the water entering and leaving fresh water ponds [92]. Some
of the greatest diversity in Tcr genes has been identified in the aquaculture envi-
ronment. In one of our studies, ~40% of the Tcr bacteria isolated from Chilean
salmon fish farms carried previously unidentified Tcr genes, suggesting that the
diversity in the types of tet resistance genes is higher than routinely found in col-
lections from either man or other food animals [57]. Some of these bacteria were
later found to carry tet(39), while other genes are still unknown [93]. It is com-
mon to find previously characterized tet genes in new bacterial genera. Many of
these tet genes were not readily transferred under laboratory conditions, thereby
raising the question of how some of the genes are transferred to bacteria across
the world and from very different environments [57]. The diversity of type and
number of Tcr bacteria found in the aquaculture setting suggests that this may be
one environment where there is rapid evolution of Tcr bacteria and a hotspot for
ARG transmission.
12 Antibiotics and Resistance in the Environment 395

12.8 Wastewater Treatment Plants (WWTP)

Municipal wastewater is a mixture of everything that is flushed down a toilet or

washed down a drain. This can include commercial, industrial, hospital, and resi-
dential waste, in addition to stormwater. The latter is especially important when
excessive rain leads to floods. Flooding is expected to become more common, as
the climate continues to change. Contamination of the sewer system by stormwa-
ter may also occur when storm and sanitary sewers are combined. Previously,
municipal wastewater and biosolids were considered waste products that required
disposal. However, as drought conditions continue, there has been a paradigm
shift. Municipalities are increasingly considering the final wastewater and bio-
sludge as resources to be utilized, rather than as waste products to be disposed of
[94, 95]. This change is occurring throughout the world, although it is not a new
idea ([96]; https://woods.stanford.edu/news-events/event/wastewater-resource-
focus-bay). WWTP do not specifically have a goal of reducing the level of ARGs
and ARBs in their final waste products.
Relatively little is known about the risk to farmers, exposed community mem-
bers, and WWTP workers to the pathogens, ARGs, and ARB present in WWTP
products. In most cases, a link between the presence of WWTP products and human
health has not been established. However, one study looking at the reuse of waste-
water found higher levels of intestinal parasitic infections among Uganda farmers
than in other persons [97]. Fenollare et al. [98] found that sewage workers were
more likely to be colonized with Tropheryma whipplei, the causal agent of Whipple’s
disease, than nonexposed people. Few other studies have looked at occupational risk
of WWTP products.
Human pathogens, including shiga toxin-producing E. coli and enteric viruses,
typically die off within a 3-month period in WWTP products, while Clostridium
spp. can persist for years as dormant endospores [99]. Spores include those from C.
perfringens and C. difficile, with the majority of work focused on C. perfringens
[100]. Several examples of the human opportunistic/pathogens associated with
WWTP effluents and biosolids are discussed below.
Wastewater treatment plants and their by-products [biosludge and effluents] have
been considered potential reservoirs, amplifiers, and transmitters of ARGs and ARB
in a variety of settings [95, 101, 102]. This is of concern because biosludge is an
important by-product of the WWTP process and is now considered an economically
important resource. Biosludge has been used for a variety of agricultural purposes,
including growing food for public consumption; effluent has been used to recharge
aquifers, for water landscaping and agriculture, and as a contributor to drinking
water [94]. These uses suggest that ARGs/ARB found in biosludge and effluent may
be transferred to food products, including shellfish. They can also contaminate
waterways, lakes, rivers, recreational waters, and oceans worldwide. Some studies
have speculated that the wastewater treatment process may increase the proportion
396 M. C. Roberts

of ARB in outlets [102]. Hotspots of ARGs and ARB may be at WWTP outflows
where wastewater effluents are discharged into bodies of water. Thus, WWTP efflu-
ent may contribute to the dissemination of specific ARGs in the natural environment
[102, 103]. Similarly, other studies have shown that use of reclaimed water is a
reservoir for ARGs which increase in the soils after repeated irrigation with
reclaimed water. This has potential implications for human health [104].
Residual ARB/ARGs in the final effluent are normally deposited into bodies of
water where they can then be taken up by fresh water and marine wildlife and ulti-
mately cycle back to humans, land animals, and/or marine life [105]. Preliminary
data supports this hypothesis. High levels of ARGs were detected where WWTP
and CSO outflows discharged into Puget Sound WA USA (Dr. L. Rhodes personnel
communications). This release may be one reason why the southern resident killer
whales carry Gram-negative and Gram-positive resistant and multiresistant bacteria
in their respiratory tracts, as determined by cultures from exhaled breath samples
[106]. Similarly, antibiotic-resistant enterococci have been isolated from feces of
sea turtles, seabirds, and marine mammals from the southern coast of Brazil [105].
We conclude that the major waterways are sources and reservoirs of ARGs and ARB
worldwide.
Conventional wastewater treatment does reduce the total number of fecal bacte-
ria, but it does not necessarily reduce the fraction of ARGs/ARB present. Over
30 years ago, Walter and Vennes [107] showed that between 0.35% and 5% of the
coliforms from a domestic sewage system were resistant to ≥1 antibiotics, with
~75% of the multiple resistant strains capable of resistance gene transfer. Other
studies have isolated and characterized multidrug-resistant fecal coliforms and/or
enterococci from municipal water from multiple geographical areas [108, 109]. To
complicate the issue, wastewater effluent is now being used for urban landscaping
and to replenish urban aquifers. Thus what is in the effluent can make its way into
the drinking water ([110]; http://www.ocwd.com/what-we-do/water-reuse/).
The wastewater treatment process, besides increasing the abundance of ARGs
and the diversity of ARBs, may also provide selective pressure to increase the
diversity of antibiotic-resistant phenotypes and transmission of ARGs to new
bacterial species. These final WWTP products can ultimately contaminate a variety
of ecosystems, with particular impact on health through aquaculture, agriculture,
the human workers in these industries, and persons who consume these products
[104]. Occupational exposure risk to human and animal health is just now being
recognized [110]. ARGs and ARB have been found throughout the wastewater treat-
ment process, from raw influent, primary and secondary effluent, aeration tanks,
activated sludge, and residual biosolids [111, 112]. The biosolids represent the
majority of the biomass and thus the highest concentration of the ARGs and ARB
from the treatment process. This material is now widely used to enrich both urban
and agricultural environments. This can lead to environmental contamination of soil
and water and, most importantly, the potential to contaminate food consumed by the
general public [101]. This potential contamination needs to be considered when
12 Antibiotics and Resistance in the Environment 397

t­ rying to determine where the bacteria causing an outbreak were introduced into the
food product of interest. Moreover, knowing which specific ARG(s) are found in
which bacterial species and/or genera in WWTP products is critical when selecting
specific ARGs for regional, national, and international surveillance studies. It is
likely that there are common microbes in most WWTP systems (E. coli and entero-
cocci), but they may differ in the carriage of ARGs. Thus, unlike isolating bacteria,
which can also lead to biases, determining which ARGs are carried by specific bac-
teria is key to the success of future surveillance efforts using molecular methods.
The use of whole genome sequencing of WWTP products with emphasis on a large
number of different ARGs would be extremely useful in determining which suite of
ARGs should be examined when screening various components of the WWTP. This
needs to be done in different types of WWTP systems in both rural and urban setting
and both economically advantaged and disadvantaged nations.
Few studies have been conducted concerning metagenome analysis of plasmids
[113] or the microbiome of human sewage [114]. More research needs to be done to
determine whether there are variations by geographical location, seasons, and other
factors. Thus most studies in the literature that screen for specific ARGs and/or
resistant plasmids are inherently biased, because of the very large number of differ-
ent ARGs that are known. This bias should be taken into account when reviewing
the literature, including studies cited below.
A variety of studies have looked for specific ARGs in influent wastewater, after
primary settling, treated effluent, activated sludge, and treated biosolids. Most of
these studies select a small subset of the known antibiotic resistance genes
characterized by conferring resistance to a particular antibiotic class. For example,
one study looked at 10 different tet genes out of 59 that are known ([22]; http://
faculty.washington.edu/marilynr/). The genes included Gram-negative-specific
efflux genes tet(A), tet(E), and tet(G) and ribosomal protection genes tet(M), tet(O),
tet(Q), and tet(S) that are found in both Gram-negative and Gram-positive bacteria
[115] from the 18 samples over a 12-month period. The Gram-negative efflux tet(A)
and tet(C) genes were identified from all samples (n = 18). The other Gram-negative
efflux genes were isolated from 9–16 of the samples. The least common Gram-
negative efflux gene, tet(D), was identified in 9 of the 18 samples. The results are
not surprising, given the distribution of the different tet genes (http://faculty.wash-
ington.edu/marilynr/). It is interesting that most common efflux gene, tet(L), which
is isolated in similar numbers of Gram-negative (n = 19) and Gram-positive (n = 22)
bacteria, was not examined ([22]; http://faculty.washington.edu/marilynr/). This is a
common issue with many of the environmental sample studies published. The
authors selected tetracycline resistance genes to survey based on what previous
studies have used rather than base the work on abundance or on those most widely
distributed ARGs among different genera in the system they are studying. This
approach provides a significant bias to many of the environmental studies, including
those on WWTP products [101, 116].
398 M. C. Roberts

12.9  elective Examples of ARGs Found in Environmental

S
Bacteria

Bacteria carrying Tcr are widely distributed throughout the world. They have been
isolated from deep, subsurface trenches; in wastewater, surface water, and
groundwater, sediments, and soils; and in pristine environments untouched by
human civilization, such as penguins in Antarctica and seals in the Arctic [42, 56,
65, 79]. Seventeen (39%) of the 43 known tet genes including 12 (44%) of the
efflux, 3 (25%) of the ribosomal protection, and 2 (66%) of the enzymatic tet genes
are uniquely ascribed to environmental bacteria. Whether this is an accurate
representation, with some tet genes being truly “unique” to environmental bacteria,
or whether these genes have not been used in surveillance studies of either animal
or human bacteria is unclear. As of 2017, there are 59 tet genes with many of the
new genes not having been identified in specific bacteria (http://faculty.washington.
edu/marilynr/).
Five different resistance genes from Streptomyces, designated otr(A), otr(B),
otr(C), tcr3, and tet, have been identified in the chromosome of antibiotic-pro-
ducing strains. Today the otr(A) and otr(B) are now found in classical Bacillus
and Mycobacterium species that were primarily environmental bacteria but
recently have caused animal and human disease. It is possible that over time
other environmental “tet genes” will move into bacteria of clinical importance
and become associated with animals and man. For example, Clostridium spp. are
found in the environment, but they are also associated with the intestinal tract of
humans and animals. The tetA(P) and tetB(P) genes appear to be unique to
Clostridium spp. Other environmental genes included are the tet(V) gene that
has been found in Mycobacterium smegmatis, which is thought to be an environ-
mental bacterium; the tet(30) gene in Agrobacterium; the tet(33) that has been
found in environmental Arthrobacter and Corynebacterium spp.; the tet(35)
gene in Vibrio and Stenotrophomonas spp., which can cause human disease; and
the tet(41) gene in Serratia spp. which rarely causes human disease. The tet(42)
gene found in Bacillus, Microbacterium, Micrococcus, Paenibacillus, and
Pseudomonas spp. was isolated from a deep-sea trench. The tet(34) gene was
first described in Vibrio spp. and more recently identified in Pseudomonas spp.
and Serratia spp.(http://faculty.washington.edu/marilynr/). To determine if these
genes are truly environmental will require new surveillance studies in human
and animal bacteria to determine if some of genes currently assigned as “uniquely
environmental” are really only associated with bacteria isolated in the
environment.
Among the 97 genes that confer resistance to one or more macrolide, lincos-
amide, and streptogramin (MLS) antibiotics, there are a number of resistance genes
that are exclusively identified in the Streptomyces spp. including rRNA methylase
genes [erm(H), erm(I), erm(N), erm(O), erm(S), erm(U), erm(Z), erm(30), erm(31),
and erm(32)], ATP-binding transporters [car(A), ole(C), srm(B), tlr(C)], and a
major facilitator [lmr(A)] gene. Other rRNA methylases are found innately in vari-
12 Antibiotics and Resistance in the Environment 399

ous environmental Mycobacterium spp., [erm(37) to erm(41)], while environmental

bacteria carry a variety of the known MLS resistance genes (http://faculty.washing-
ton.edu/marilynr/). Other than genes associated with Streptomyces spp. and
Mycobacterium spp., there are relatively few genes exclusively associated with
environmental bacteria. Why a difference occurs in the distribution between tet and
MLS genes in environmental bacteria is not clear.
β-Lactamases are enzymes that provide resistance to β-lactam antibiotics such as
penicillins, cephamycins, and carbapenems (ertapenem). These β-lactamase
enzymes have random mutations that modify the spectrum of resistance to varying
classes of this antibiotic group. There are hundreds of these modified β-lactamase
genes. β-Lactamase genes are ancient and have been identified in remote and
isolated environments, suggesting that β-lactamases occur in nature [66]. Another
class of β-lactamases, the CTX-M genes, which hydrolyze expanded-spectrum
cephalosporins, originated in environmental Kluyvera spp. Bacteria with CTX-M
genes were first identified in 1989. Today these genes can be found across the world
[3]. The qnr genes originated in waterborne Aeromonas, Shewanella, and Vibrio
spp. [52]. Data from a 30,000-year-old permafrost sample showed that the sample
carried genes conferring resistance to a variety of different classes of antibiotics
[β-lactams, tetracycline, and glycopeptides]; thus, resistance existed in the
environment before antibiotics were used by man.

12.10 Conclusions

The environmental microbiome, which is difficult to define, remains largely unex-

plored. However, a few studies suggest the wide distribution of ARB and ARGs in
the environment. For example, antibiotic-resistant marine bacteria have been iso-
lated 522 km offshore and at depths of 8200 m [117]. The degree of pollution in the
environment correlates with the prevalence of resistance, suggesting that over time
even the more “pristine” environments will become increasingly contaminated with
ARGs and ARB. This phenomenon will ultimately increase resistance among
opportunistic and pathogenic bacterial species having human and animal impor-
tance. Increased selection pressure for antibiotic resistance in environmental micro-
organisms is likely to continue, since human activities will likely continue to pollute
the environment. Natural forces, such as wind and movement of water, will continue
to contaminate areas of relatively uninhabited environments.
The One Health concept is a worldwide strategy for expanding interdisciplinary
collaborations and communications in all aspects of health care for humans, animals,
and the environment. The aim is for inclusive collaborations dedicated to improving
the lives of all species through the integration of human medicine, veterinary
medicine, and environmental science. This concept recognizes that using
compartmentalized (silo) mentality to approach the three disciplines is not adequate,
since the distinction of environment from non-environment, especially at the
bacterial level, has become increasingly difficult. It is clear that the introduction of
400 M. C. Roberts

a new ARG into a human, animal, agricultural, or environmental microbial

ecosystem often leads to cross-transmission and dissemination of ARGs and ARB
within and between the ecosystems [3].
The data summarized in this chapter indicate that the environment is an impor-
tant reservoir for ARGs and ARB; it needs to be considered in future studies. There
is a large diversity of resistance genes in the environment, and many of these genes
have yet to be identified or characterized. Horizontal gene transfer within the micro-
bial world knows few boundaries, and our ability to experimentally mimic what
occurs in nature has significant limitations. Indeed, the role that the natural environ-
ment plays in the evolution, maintenance, and transmission of ARB and ARGs is
just now being examined. However, it is generally agreed that human anthropo-
genic changes are impacting natural ecosystems that will ultimately impact human
and animal health.
It is clear that ARB and ARGs are spread among animals, the environment, and
humans and from one geographic location to another throughout the world. The
environment is an important reservoir for these resistance genes, with WWTP
products being an important component as reservoirs, potential amplifiers, and/or
transmitters of ARB and ARGs in the environment. These contaminants not only
degrade the local environment but ultimately influence the health of humans and
animals associated with that environmental landscape. The environment has
provided an increasing number of novel ARGs that have not been found in bacteria
traditionally associated with animals or humans (http://faculty.washington.edu/
marilynr/). It is unclear whether these “new genes” will impact the treatment of
animal and human infections in the future, but NDM-1 and CTX-M genes have
been associated with bacterial pathogens. Evidence also exists that WWTP plays a
role in the evolution of multidrug-resistant opportunistic and pathogenic bacteria.
WWTP is thought to be a hotspot for the contamination of environments including
receiving waters of effluent and of soil and agricultural lands where biosolids are
utilized. This is very important, as WWTP biosolids and final effluents are considered
to be resources that should be used for agricultural purposes and, in some
communities, as water resources. Thus it is plausible that there is a human health
risk associated with WWTP products; however, data backing this hypothesis is
currently very limited. Reducing the levels of ARGs/ARB in WWTP by-products
before they are recycled is an important component in the multipronged approach to
reduce the global spread and distribution of ARGs. Advanced wastewater treatments
using ozone, UV, ultrafiltration, chlorination, dry-air beds, and membrane bioreactor
processes are effective in reducing the number of bacteria. These processes may be
useful in reducing the level of ARB/ARGs in effluents and biosolids before they are
utilized by communities, thereby reducing the risk to humans (113). Unfortunately,
recent studies report that UV/H2O2 disinfection processes do not eliminate the pos-
sible spread of antimicrobial resistance in the receiving environment [118].
Moreover, cost-effectiveness is an important consideration with advanced wastewa-
ter treatment options. To comprehensively assess AMR-related impacts on risks to
human health, we need to gain a better understanding of the role that biosolids and
effluents play as amplifiers, reservoirs, and transmitters of these bacteria and genes.
12 Antibiotics and Resistance in the Environment 401

It is important that members of human communities understand that they con-

tribute to the contamination of their environment – practices such as discarding food
and food waste products inappropriately may have downstream consequences. Thus
education of the general community, from young children through adults, is an
important mission that many scientist in the field neglect – it is potentially the most
cost-effective use of resources.
Major Points
Limited work on various environmental ecosystems limits our understanding the
relationships between environmental bacteria and the stressors that lead to selec-
tions and retention of ARGS/ARB in only one system. Preliminary data indicates
that certain places such as WWTP and the receiving waters of this material along
with the biosolids produced in the WWTP are hotspots for the exchange of ARGs
among the bacterial microbiome. How to deal with these products to reduce the
number and diversity of ARGs and ARB is not clear. Using a One Health approach,
it is clear that ARGs and ARB can flow from humans and/or animals into the envi-
ronment and environmental bacteria, and genes can flow back into human and/or
animal bacteria. Looking at the complete picture will provide better information for
specific ARGs and ARB and with this knowledge perhaps ways of reducing overall
transmission from one sector to the other. This requires resources and science at all
levels to stabilize and hopefully reduce the human-generated impact on the environ-
ment including contamination as well as changes in climates which can disturb the
natural web of life as well as increase food insecurity to millions.

References

1. Feighner SD, Dashkevicz MP. Subtherapeutic levels of antibiotics in poultry feeds and their
effects on weight gain, feed efficiency, and bacterial cholyltaurine hydrolase activity. Appl
Environ Microbiol. 1987;53:331–6.
2. Marshall BM, Levy SB. Food animals and antimicrobials: impacts on human health. Clin
Microbiol Rev. 2011;24:718–33.
3. Davies J, Davies D. Origins and evolution of antibiotic resistance. Microbiol Mol Biol Rev.
2010;74:417–33.
4. Watanabe T. Fukasawa T. Episome-mediated transfer of drug resistance in Enterobacteriaceae.
1. Transfer of resistance factors by conjugation. J Bacteriol. 1961;81:669–78.
5. WHO Global Action Plan on Antimicrobial Resistance. 2015. [cited 2017 May 20.] Available
from: http://www.who.int/drugresistance/global_action_plan/en/
6. Collignon PC, Conly JM, Andremont A, McEwen SA, Aidara-Kane A. World Health
Organization ranking of antimicrobials according to their importance in human medicine:
a critical step for developing risk management strategies to control antimicrobial resistance
from food animal production. Clin Infect Dis. 2016;63(8):1087–93.
7. Boucher HW, Bakken JS, Murray BE. The United Nations and the urgent need for coordinated
global action in the fight against antimicrobial resistance. Ann Intern Med. [serial online].
2016. [cited 2017 May 4.] Available from: www.annals.org
8. Center for Diseases Dynamics, Economics & Policy. The State of the World’s Antibiotics
2015. [cited 2017 May 4.] Available from: http://www.cddep.org/publications/state_worlds_
antibiotics_2015#sthash.3wasnwH1.dpbs
402 M. C. Roberts

9. Wall, BA, Mateus A, Marshall L, Pheiffer DU. Drivers, dynamics and epidemiology of
antimicrobial resistance in animal production. Food and Agriculture Organization of the
United Nations. Rome, 2016. [cited 2017 May 4.] Available from: http://www.fao.org/3/a-
i6209e.pdf
10. Collignon P. The importance of a one health approach to preventing the development and
spread of antibiotic resistance. Cur Top Microbiol Immunol. 2013;366:19–36.
11. Center for Disease Control. Antibiotic Resistant Threats in the United States. 2013. [cited
2017 May 4.] Available from: http://www.cdc.gov/drugresistance/threat-report-2013/pdf/ar-
threats-2013-508.pdf
12. WHO Antibiotic Resistance Global Report of Surveillance 2014. [cited 2017 June 2].
Available from: http://apps.who.int/iris/bitstream/10665/112642/1/9789241564748_eng.
pdf?ua=1
13. Van Boeckel TP, Brower C, Gilbert M, et al. Global trends in antimicrobial use in food ani-
mals. PNAS. 2015;112:5649–54.
14. Berendonk TU, Manaia CM, Merlin C, et al. Tackling antibiotic resistance: the environmen-
tal framework. Nat Rev Microbiol. 2015;13:310–7.
15. National Strategy for Combating Antibiotic-Resistant Bacteria. 2014. [cited 2017 June 2.]
Available from: https://www.whitehouse.gov/sites/default/files/docs/carb_national_strategy.
pdf
16. Taylor, J, Hafner M, Yerushalmi E, et al. Estimating the economic costs of antimicrobial
resistance: Model and results. RAND. 2014 [cited 2017 May15] Available from: http://www.
rand.org/content/dam/rand/pubs/research_reports/RR900/RR911/RAND_ RR911.pdf
17. Tillotson GS. Where in the world? The role of geography in antibiotic resistance and the
potential impact in pulmonary infections. Postgrad Med. 2016;128(5):449–50.
18. Durso LM, Miller DN, Wienhold BJ. Distribution and quantification of antibiotic resis-
tant genes and bacteria across agricultural and non-agricultural metagenomes. PLoS ONE.
[serial online] 2012. [cited 2017 June 2] Available from: https://doi.org/10.1371/journal.
pone.0048325
19. Roberts M C, Schwarz S, Aarts H. Acquired antibiotic resistance genes: an overview. Front
Microbiol. [serial online] 2012 [cited. 2017 May 30] 3:384; 1–17. Available from: http://www.
frontiersin.org/Antimicrobials,_Resistance_and_Chemotherapy/10.3389/fmicb.2012.00384/full
20. Chopra I, Roberts MC. Tetracycline antibiotics: mode of action, applications, molecular biol-
ogy and epidemiology of bacterial resistance. Microbiol Mol Biol Rev. 2001;65:232–60.
21. Roberts MC. Mechanisms of bacterial antibiotic resistance and lessons learned from environ-
mental tetracycline resistant bacteria. In: Keen P, Montforts M, editors. Antibiotic resistance
in the environment: John Wiley & Sons, Inc.; 2012. p. 93–121.
22. Roberts MC, Schwarz S. Tetracycline and phenicol resistance genes and mechanisms: impor-
tance for agriculture, the environment and humans. J Environ Quality. 2016;45:576–92.
23. Roberts MC, Knapp JS. Host range of the conjugative 25.2 Mdal tetracycline resistance plas-
mid from Neisseria gonorrhoeae. Antimicrob Agents Chemother. 1988;32:488–91.
24. Roberts MC, Knapp JS. Transfer of β-lactamase plasmids from Neisseria gonorrhoeae to
Neisseria meningitidis and commensal Neisseria species by the 25.2-Megadalton conjugative
plasmid. Antimicrob Agents Chemother. 1988;32:1430–2.
25. Hu Y, Yang X, Lv N, et al. The bacterial mobile resistome transfer network connecting the
animal and human microbiomes. Appl Environ Microbiol. 2016;82(22):6672–81.
26. Rizzo L, Manaia C, Merlin C, et al. Sci Total Environ. 2013;447:345–60.
27. Nordmann P, Naas T, Poirel L. Global spread of carbapenemase-producing Enterobacteri­
aceae. Emerg Infect Dis. 2011;17:1791–8.
28. Garrido-Messa N, Zarzuelo A, Galvez J. Minocycline: far beyond an antibiotic. Br
J Pharmacol. 2013;169:337–52.
29. Leigh MJS, Nguyen DV, My Y, et al. A randomized double-blind, placebo-controlled trial of
minocycline in children and adolescents with Fragile X syndrome. J Devel Behav Pediatr.
2014;34:147–55.
12 Antibiotics and Resistance in the Environment 403

30. Dezube BJ, Krown SE, Lee JY, et al. Randomized phase II trial of matrix metalloproteinase
inhibitor COL-3 in AIDS-related Kaposi’s sarcoma: an AIDS Malignancy Consortium Study.
J Clin Oncol. 2006;24:1389–94.
31. Parnham MJ, Haber VE, Giamarellos-Bourboulis EJ, Perletti G, Verleden GM, Vos
R. Azithromycin: mechanism of action and their relevance for clinical applications.
Pharmacol Ther. 2014;143:225–45.
32. Lindahl JF, Grace D. The consequences of human actions on risks for infectious disease:
a review. Infect Ecol Epidem. [serial online] 2015 [cited 2017 May 25]; Available from:
5:30048-https://doi.org/10.3402//lee.v5.30048
33. USDA Animal and Plant Health Inspection service Veterinary services 2015 One Health.
What is One Health? [cited 2017 May 20.] Available from: https://www.aphis.usda.gov/ani-
mal_health/one_health/downloads/one_health_info_sheet.pdf.
34. Heymann DL, Dixon M. The value of the one health approach: shifting from emergency
response to prevention of zoonotic disease threats at their source. In: Atlas RM, Maloy S, edi-
tors. One health people, animals, and the environment. Washington, DC: American Society
for Microbiology; 2014. p. 17–31.
35. Cetinkaya Y, Falk P, MayHall CG. Vancomycin-resistant enterococci. Clin Microbiol Rev.
2000;13:686–707.
36. van den Bogaard AE, Willems R, London N, Top J, Stobberingh EE. Antibioitc resistance
of faecal enterococci in poultry, poultry farmers and poultry slaughterers. J Antimicrob
Chemoth. 2002;49:497–505.
37. Eckstein BC, Adams DA, Eckstein EC, et al. Reduction of Clostridium difficile and vanco-
mycin-resistant Enterococcus contamination of environmental surfaces after an intervention
to improve cleaning methods. BMC Infect Dis. [serial online] 2007 [cited 2017 May 20];7:61
Available from:http://www.biomedcentral.com/1471-2334/7/61
38. Rice LB. The complex dynamics of antimicrobial activity in the human gastrointestinal tract.
Trans Am Clin Climatol Assoc. 2013;124:123–32.
39. Oravcova V, Zurek L, Townsend A, et al. American crows as carriers of vancomycin-resistant
enterococci with vanA gene. Environ Microbiol. 2014;16:939–49.
40. Roberts MC, Soge OO, Giardino MA, Mazengia E, Ma G, Meschke JS. Vancomycin- resis-
tant Enterococcus spp. in environments from the west coast of the USA. J Appl Microbiol.
2009;107:300–7.
41. Roberts MC, No DB, Marzluff JM, DeLap JH, Turner R. Vancomycin resistant Enterococcus
spp. from crows and their environment in Metropolitan Washington State, USA: is there a cor-
relation between VRE positive crows and the environment? Vet Microbiol. 2016;194:48–54.
42. Kümmerer K. Resistance in the environment. J Antimicrob Chemother. 2004;54:311–20.
43. Wright GD. The antibiotic resistome: the nexus of chemical and genetic diversity. Nat Rev
Microbiol. 2007;5:175–86.
44. Guarddon M, Miranda JM, Rodriquez JA, et al. Real-time polymerase chain reaction for
the quantitative detection of tetA and tetB bacterial tetracycline resistance genes in food. Int
J Food Microbiol. 2011;146:284–9.
45. Peng S, Wang Y, Zhou B, et al. Long-term application of fresh and composted manure
increase tetracycline resistance in the arable soil of eastern China. Sci Total Environ.
2015;506-507:279–86.
46. Harnisz M, Korzeniewska E, Ciesielski S, et al. tet genes as indicators of changes in the water
environment: relationships between culture-dependent and culture-independent approaches.
Sci Total Environ. 2015;505:704–11.
47. Dantas G, Sommer MOS, Oluwasegun RD, et al. Bacterial subsisting on antibiotics. Science.
2008;320:100–3.
48. Moore AM, Patel S, Forsberg KJ, et al. Pediatric fecal microbiota harbor diverse and novel
antibiotic resistance genes. PLoS One. [serial online] 2013 [cited 2017 June 1]; 8:e78822
Available from: http://www.plosone.org/article/fetchObject.action?uri=info:doi/10.1371/
journal.pone.0078822&representation=PDF
404 M. C. Roberts

49. Lin B, Pop M. ARDB-Antibiotic Resistance Genes Database. Nucleic Acids Res. 2009 [cited
2017 May 11]; Jan:37(Database issue):D443-7 Available from:http://ardb.cbcb.umd.edu/
50. Jia B, Raphenya AR, Alcock B, et al. CARD 2017: expansion and model-centric curation of
the comprehensive antibiotic resistance database. Nucl Acids Res. [serial online] 2017; [cited
June 1]; 45:D566–D573 Available from: doi:https://doi.org/10.1093/nar/gkw1004.
51. D’Costa V, McGrann KM, Hughes DW, et al. Sampling the antibiotic resistome. Science.
2006;311:374–7.
52. Perry JA, Wright GE. The antibiotic resistance “mobiolime”: Searching for the link between
environment and clinic. Front Microbiol. [serial online] 2013 [cited 2017 June 2];4:138
Available from: doi: https://doi.org/10.3389/fmicb.2013.00138.
53. Forsberg KJ, Reyes A, Wang B, et al. The shared antibiotic resistome of soil and human
pathogens. Science. 2012;337:1107–11.
54. Perry JA, Westman EL, Wright GE. The antibiotic resistome: what’s new? Curr Opin
Microbiol. 2014;21:45–50.
55. Thaker M, Spanogiannopoulos P, Wright GD. The tetracycline resistome. Cell Mol Life Sci.
2010;67:419–31.
56. Donato JJ, Moe LA, Converse BJ, et al. Metagenomic analysis of apple orchard soil reveals
antibiotic resistance genes encoding predicted bifunctional proteins. Appl Environ Microbiol.
2010;76:4396–401.
57. Miranda CD, Kehrenberg C, Ulep C, et al. Diversity of tetracycline resistance genes in bac-
teria from Chilean salmon farms. Antimicrob Agents Chemother. 2003;47:883–8.
58. Forsberg KJ, Patel S, Wencewicz TA, Dantas G. The tetracycline destructases: a novel family
of tetracycline-inactivation enzymes. Chem Biol. 2015;22:888–97.
59. de Oliveria DV, Nunes LS, Barth AL, Van Der Sand. Genetic background of β-lactamases in
Enterobacteriaceae isolates from environmental samples. Microb Ecol. [serial online] 2017
[cited 2017 May 16.] Available from:doi:https://doi.org/10.1007/s00248-017-0970-6
60. Capkin E, Terzi E, Altinok I. Occurrence of antibiotic resistance genes in culturable bacte-
ria isolated from Turkish trout farms and their local aquatic environment. Dis Aquat Org.
2015;114:127–37.
61. Guyomard-Rabenirina S, Darton C, Falord M, et al. Resistance to antimicrobial drugs in
different surface waters and wastewaters of Guadelopupe. PLoS One. [serial online] 2017.
[cited 2017 May 16]; Mar 2;12(3):e0173155. Available from: doi:https://doi.org/10.1371/
journal.pone.0173155.
62. van Hoek AHAM, Schouls L, van Santen MG, Florihn A, de Greeff SC, van Duijkeren
E. Molecular characteristics of extended-spectrum cephalosporin-resistant Enterobacteriaceae
from humans in the community. PLoS One. [serial online] 2015 [cited 2017 May 16]:
10(6):e0129085. Available from; doi:https://doi.org/10.1371/journal.pne.0129085
63. Rowlinson M-C, Bruckner DA, Hinnebusch C, et al. Clearance of Cellulosimicrobium cel-
lulans bacteremia in a child without central venous catheter removal. J Clin Microbiol.
2006;44:2605–54.
64. Pallecchi L, Bartonloni A, Riccobono E, et al. Quinolone resitance in absence of selective
pressure: The experience of a very remote community in the Amazon forest. PLoS Neg
Trop Dis. [serial online] 2012 [cited 2016 July 10]; 6:e1790 Available from: doi: https://doi.
org/10.1371/journal.pntd.0001790
65. Rahman MH, Sakamoto KQ, Nonaka L, et al. (2008). Occurrence and diversity of tetra-
cycline tet(M) in enteric bacteria of Antarctic Adelie penguins. J Antimicrob Chemother.
2008;62:627–8.
66. Allen HK, Donato J, Wang HH, et al. Call of the wild: antibiotic resistance gene in natural
environments. Nat Rev Microbiol. 2010;8:251–9.
67. Berglund B, Fick J, Lindgren PE. Urban wastewater effluent increases antibiotic resis-
tance genes concentrations in a receiving northern European river. Environ Toxicol Chem.
2015;34:192–6.
12 Antibiotics and Resistance in the Environment 405

68. Agersø Y, Jensen LB, Givskov M, et al. The identification of a tetracycline resistance gene
tet(M), on a Tn916-like transposon, in the Bacillus cereus group. FEMS Microbiol Lett.
2002;214:251–6.
69. Dobbs FC, Goodrich AL, Tomson FS III, et al. Pandemic serotypes of Vibrio cholerae iso-
lated from ships’ ballast tanks and coastal waters: assessment of antibiotic resistance and
virulence genes (tcpA and ctxA). Microb Ecol. 2013;65:969–74.
70. MacFadden DR, Bogoch II, Brownstein JS, et al. A passage from India: association between
air traffic and reported cases of New Delhi metallo-beta-lactase 1 from 2007 to 2012. Travel
Med Infect Dis. 2015;13:295–9.
71. Gould LH, Limbago B. Clostridium difficile in food and domestic animals: a new foodborne
pathogen? Clin Infect Dis. 2010;51(5):577–82.
72. Abbo A, Navon-Venezia S, Hamemer-Muntz O, et al. Multidrug-resistant Acinetobacter bau-
mannii. Emerg Infect Dis. 2005;11:22–9.
73. U.S. Food and Drug Administration (FDA) FACT SHEET: Veterinary Feed Directive
Final Rule and Next Steps. [cited 2017 May 30.] Available from: https://www.fda.gov/
AnimalVeterinary/DevelopmentApprovalProcess/ucm449019.htm
74. Lord Soulsby of Swaffham Prior. The 2008 Garrod Lecture: antimicrobial resistance--ani-
mals and the environment. J Antimicrob Chemother. 2008;62(2):229–33.
75. Price LB, Johnson E, Vailes R, Silbergeld E. Fluoroquinolone-resistant Campylobacter
isolates from conventional and antibiotic-free chicken products. Environ Health Perspect.
2005;113:557–60.
76. Hao R, Zhao R, Qiu S, et al. Antibiotics crisis in China. Science. 2015;348:1100–1.
77. Facinelli B, Roberts MC, Giovanetti E, et al. Genetic basis of tetracycline resistance in food
borne isolates of Listeria innocua. Appl Environ Microbiol. 1993;59:614–6.
78. Rolland RM, Hausfater G, Marshall B, et al. Antibiotic-resistant bacteria in wild pri-
mates: increased prevalence in baboons feeding on human refuse. Appl Environ Microbiol.
1985;49:791–4.
79. Glad T, Kristiansen VF, Nielsen KM, Brusetti L, Wright A-DG, Sundset MA. Ecological
characterisation of the colonic microbiota in arctic and sub-arctic seals. Microb Ecol.
2010;60:320–30.
80. Poeta P, Radhouani H, Pinto L, et al. Wild boars as reservoirs of extended-spectrum beta-­
lactamase (ESBL) producing Escherichia coli of different phylogenetic groups. J Basic
Microbiol. 2009;49:584–8.
81. Marinho C, Igrehas G, Goncalves A, et al. Azorean wild rabbits as reservoirs of antimicrobial
resistant Escherichia coli. Anaerobe. 2014;30:116–9.
82. Ardiles-Villegas K, Gonzalez-Acuna D, Waldenstrom J, Olsen B, Hernandez J. Antibiotic
resistance patterns in fecal bacteria isolated from Christmas shearwater (Puffinus nativitatis)
and masked booby (Sula dactylatra) at remote Easter Island. Avian Dis. 2011;55:4896–489.
83. Nilsson O (2012) Vancomycin resistant enterococci in farm animals-occurrence and impor-
tance. Infect Ecol Epidemiol. [serial online] 2012 [cited 2017 June 2; 2:16959 Available
from: https://doi.org/10.3402/lee.v2i0.16959
84. Gordoncillo MJN, Donabedian S, Bartlett PC, et al. Isolation and molecular characterization
of vancomycin-resistance Enterococcus faecium from swine in Michigan, USA. Zoonoses
Pub Health. 2012. 2012;60:319–26.
85. Li K, Petersen G, Barco L, Hvidtfeldt K, Liu L, Dalsgaard A. Salmonella Weltevreden in inte-
grated and non-integrated tilapia aquaculture systems in Guangdong, China. Food Microbiol.
2017;65:19–24.
86. Muziasari WI, Pikanen LK, Sorum H, Stedtfeld RD, Tiedje JM, Virta M. The resistome of
farmed fish feces contributes to the enrichment of antibiotic resistance genes in sediments
below Baltic Sea fish farms. Front Microbiol. [serial online] 2017 [cited 2017 May 8]
Available from: doi:https://doi.org/10.3380/fmicb.2016.02137ss.
87. Akinbowale OL, Peng H, Barton MD. Diversity of tetracycline resistance genes in bacteria
from aquaculture sources in Australia. J Appl Microbiol. 2007;103:2016–25.
406 M. C. Roberts

88. DePaola A, Roberts MC. Class D and E tetracycline resistance determinants in gram-negative
catfish pond bacteria. Mol Cell Probes. 1995;9:311–3.
89. Furushita M, Shiba T, Maeda T, et al. Similarity of tetracycline resistance genes isolated from
fish farm bacteria to those from clinical isolates. Appl Environ Microbiol. 2003;69:5336–42.
90. Jacobs L, Chenia HY. Characterization of integrons and tetracycline resistance determinants
in Aeromonas spp. isolated from South African aquaculture systems. Int J Food Microbiol.
2007;114:295–306.
91. Nawaz M, Khan AA, Khan S, et al. Isolation and characterization of tetracycline-resistant
Citrobacter spp. from catfish. Food Microbiol. 2008;25:85–91.
92. Miranda CD, Zemelman R. Bacterial resistance to oxytetracycline in Chilean salmon farms.
Aquaculture. 2002;212:31–47.
93. Roberts MC, No D, Kuchmiy E, Miranda CD. The tetracycline resistant tet(39) gene identi-
fied in three new genera of bacteria isolated in 1999 from Chilean Salmon farms. J Antimicrob
Chemother. 2015;70:619–20.
94. Hayden EC. California faces arid future. Nature. 2015;526:14–5.
95. Reinthaler FF, Posch J, Feierl G, et al. Antibiotic resistance of E. coli in sewage and sludge.
Water Res. 2003;37:1685–90.
96. Hespanhol I. Wastewater a as resource, Chpt 4. In R Helmer, I Hespanhol, editors. Water
Pollution Control-A guide to the use of water quality management principles. 1997. Available
from: http://www.who.int/water_sanitation_health/resourcesquality/wpcchap4.pdf
97. Fuhrimann S, Winkler MS, Schneeberger PHH, et al. Health risk assessment along the waste-
water and faecal sludge management and reuse chain of Kampala, Uganda: a visualization.
Geospat Health. 2014;9:251–5.
98. Fenollare F, Marth T, Lagier JC, Angelakis E, Raoult D. Sewage workers with low antibody
response may be colonized successively by several Tropheryma whipplei strains. Int J Infect
Dis. 2015;35:51–5.
99. Rouch DA, Mondal T, Pai S, et al. Microbial safety of air-dried and rewetted biosolids.
J Water Health. 2011;9:403–14.
100. Xu C, Weese SJ, Flemming C, Odumeru J, Warriner K. Fate of Clostridium difficile dur-
ing wastewater treatment and incidence in Southern Ontario watersheds. J Appl Microbiol.
2014;117:891–904.
101. Burch TR, Sadowsky MJ, LaPara TM. Air-drying beds reduce the quantities of antibiotic
resistance genes and class 1 integrons in residual municipal wastewater solids. Environ Sci
Technol. 2013;47:9965–71.
102. Dropa M, Lincopan N, Balsalobre LC, et al. Genetic background of novel sequence types of
CTS-M-8- and CTX-M-15-producing Escherichia coli and Klebsiella pneumoniae from pub-
lic wastewater treatment plants in Sao Paul, Brazil. Environ Sci Pollut Res. 2016;23:4953–8.
103. Proia L, von Schiller D, Sanchez-Melsio A, et al. Occurrence and persistence of antibiotic
resistance genes in river biofilms after wastewater inputs in small rivers. Environ Pollut.
2016;210:121–8.
104. Fahrenfeld N, Ma Y, O’Brien M, Pruden A. Reclaimed water as a reservoir of antibiotic resis-
tance genes: distribution system and irrigation implications. Front Microbiol. [serial online]
2013 [cited 2017 May 30] Available from: doi:https://doi.org/10.3389/micb.2013.00130.
105. Richula J, Pereira RI, Wachholz GR, et al. Resistance to antimicrobial agents among entero-
cocci isolated from fecal samples of wild marine species in the southern coast of Brazil. Mar
Pollut Bull. 2016;105(1):51–7.
106. Raverty SA, Rhodes LD, Zabek E, et al. Respiratory microbiome of endangered Southern
resident killer whales and microbiota of surrounding sea surface microlayer in the Eastern
North Pacific. Sci Reports. [serial online] 2017 [cited 2017 June 1];7:394 Available from:
doi:https://doi.org/10.1038/s41598-017-00457-5.
107. Walter MV, Vennes JW. Occurrence of multiple-antibiotic-resistant enteric bacteria in domes-
tic sewage and oxidation lagoons. Appl Environ Microbiol. 1985;50:930–3.
12 Antibiotics and Resistance in the Environment 407

108. da Costa PM, Vaz-Pires P, Bernardo F. Antimicrobial resistance in Enterococcus spp. isolated
in inflow, effluent and sludge from municipal sewage waters treatment plants. Water Res.
2006;40:1735–40.
109. Luczkiewica A, Jankowska K, Fudala-Ksiazek S, Olanczuk-Neyman K. Antimicrobial
resistance of fecal indicators in municipal wastewater treatment plant. Water Res.
2010;44:5089–97.
110. Rosenberg Goldstein ER, Micallef SA, Gibbs SG, et al. Occupational exposure to
Staphylococcus aureus and Enterococcus spp. among spray irrigation workers using
reclaimed water. Int J Environ Res Public Health. 2014;11:4340–55.
111. Diallo AA, Brugere H, Kerouredan M, et al. Persistence and prevalence of pathogenic and
extended-spectrum beta-lactase-producing Escherichia coli in municipal wastewater treat-
ment plant receiving slaughterhouse wastewater. Water Res. 2013;47:4719–29.
112. Huijbers PMC, Blaak H, de Jong MCM, et al. Role of the environment in the transmission of
antimicrobial resistance to humans: a review. Environ Sci Technol. 2015;49:11993–2004.
113. Szczepanowski R, Linke B, Krahn I, et al. Detection of 140 clinically relevant antibiotic-­
resistance genes in the plasmid metagenome of wastewater treatment plant bacteria showing
reduced susceptibility to selected antibiotics. Microbiology. 2009;155:2306–19.
114. Cai L, Ju F, Zhang T. Tracking human sewage microbiome in a municipal wastewater treat-
ment plant. Appl Microbiol Biotechnol. 2014;98:3317–26.
115. Auerbach EA, Seyfried EE, McMahon KD. Tetracycline resistance genes in activated sludge
wastewater treatment plants. Water Res. 2007;41:1143–51.
116. Naquin A, Shrestha A, Sherpa M, Nathaniel R, Boopathy R. Presence of antibiotic resis-
tance genes in a sewage treatment plant in Thibodaux, Louisiana, USA. Bioresour Technol.
2015;188:79–83.
117. Aminov RI. Horizontal gene exchange in environmental microbiota. Front Microbiol.
[serial online] 2011[cited 2017 May 22] Available from: doi:https://doi.org/10.3389/
fmmicb.2011.00158.
118. Ferro G, Guarino F, Castiglione S, Rizzo L. Antibioitc resistance spread potential in urban
wastewater effluents disinfected by UV/H2O2 process. Sci Total Environ. 2016;560-561:29–35.
Chapter 13
Phenotypic Tolerance and Bacterial
Persistence

Carl Nathan

13.1 Introduction

Antibiotics are among the most important achievements of biomedical science.

However, they are also among the most endangered. Not only are antibiotics suscep-
tible to rapid emergence of heritable resistance, but their action is resisted by a
much less well understood set of processes collectively termed “phenotypic toler-
ance” that gives rise to “persisters.” Persisters are the members of a population of an
antibiotic-susceptible strain of bacteria that survive exposure to the antibiotic at
concentrations that kill the vast majority of the population when tested under the
conditions used to define the antibiotic’s minimum inhibitory concentration (MIC)
and that when expanded in number and retested give rise to a population whose
MIC is unchanged. The major theme of this chapter is that mechanistically distinct
forms of phenotypic tolerance present different challenges for the development of
effective therapeutic approaches.
The chapter begins by describing what is at stake with the rise of antimicrobial
resistance (AMR) and then contrasts heritable AMR with its nonheritable form,
phenotypic tolerance. With tuberculosis (TB) as a case in point, I review the contri-
bution of host immunity to phenotypic tolerance. This sets the stage for contrasting
two major classes of phenotypic tolerance. Turning to the history of how phenotypic
tolerance was recognized, we will see that evidence for two major classes was evi-
dent from the outset, although the distinction was not perceived at the time. Finally,

This chapter reproduces, adapts, and updates portions of a wider-ranging essay, “Fundamental
Immunodeficiency and Its Correction” by C. Nathan, published in the Journal of Experimental
Medicine 214: 2175-2191, 2017. Passages from that publication are reproduced here with permis-
sion of The Rockefeller University Press.
C. Nathan (*)
Department of Microbiology and Immunology, Weill Cornell Medicine, New York, NY, USA
e-mail: [email protected]

© Springer International Publishing AG, part of Springer Nature 2018 409

I. W. Fong et al. (eds.), Antimicrobial Resistance in the 21st Century, Emerging Infectious
Diseases of the 21st Century, https://doi.org/10.1007/978-3-319-78538-7_13
410 C. Nathan

I discuss what is known about the mechanisms for each class and approaches to
overcome them.

13.2  nique Features of Antimicrobial Agents

U
among Medicines

Over the past six generations, humans have found or invented several thousand
medicines. Among them, the antimicrobial agents, discovered over the past four
generations, are unique in two aspects. First, until recently, antimicrobial agents
were the only medicines that cured large numbers of the sick, and they remain the
only medicines that do so routinely. Within the last two generations, some antineo-
plastic regimens have been curative, including some that are immunity-based, and
corticosteroids sometimes cure temporal arteritis. Second, antimicrobial agents are
the only medicines whose use hastens their loss of usefulness for people who have
not yet taken them.
The first claim hinges on using “cure” in the true sense. Administration of an
appropriately chosen antimicrobial agent has the routine capacity to restore an indi-
vidual to the state of wellness that prevailed before the onset of an illness that would
not otherwise have resolved, that would not otherwise have resolved as quickly, or
whose unaided resolution would not restore the individual to their prior state of
wellness. In contrast, when the administration of most other medicines stops, the
individual returns to the state of illness that invited intervention, unless the illness
had resolved spontaneously or from a change in contributory factors, such as diet.
Some other medicines help prevent the onset of illness rather than treating it.
The definition of “cure” given above is admittedly idealized. Clinical cure can be
ambiguous. “Cure” does not return the patient to the previous state of health if tissue
damage already caused by the pathogen or the host’s reaction to it is irreparable, as
is often the case in successfully treated TB. Finally, cure achieved with broad spec-
trum antimicrobial agents often comes at the cost of a long-lasting perturbation of
the microbiota, and in that sense an important component of the host’s overall
makeup has not returned to its preexistent state. Nonetheless, within the bounds of
these ambiguities and qualifications, antimicrobial agents stand out among medi-
cines for their ability to cure large numbers of people routinely.
However, the ability of antimicrobial agents to cure the majority of patients for
whom such drugs are appropriately prescribed is handicapped by the second unique
feature of this class of medicines: their use eventually selects for resistance. The
resistant pathogens are eventually shared among hosts, or the determinants of resis-
tance are eventually shared among pathogens. Thus we are all likely to need antimi-
crobial agents, yet the more a given agent is used, the nearer it comes to being
useless.
In sum, antimicrobial agents are at once among the most important and least
permanent achievements of medicine.
13 Phenotypic Tolerance and Bacterial Persistence 411

13.3  ising Stakes: The Growing Reach and Recognition

R
of Antimicrobial Resistance (AMR)

Beginning with the use of penicillin in civilian populations in the mid-1940s, physi-
cians, scientists, and much of the public quickly came to regard antimicrobial agents
as both indispensable and invincible [1]. Beginning just 20 years later, taking anti-
microbial agents for granted put us on a path to losing them.
Over the past few decades, a declining rate of success in discovering new antimi-
crobial agents discouraged much of the pharmaceutical industry from continuing
the search [2]. Meanwhile, levels of AMR continue to rise. These respectively fall-
ing and rising curves have crossed in recent years for one pathogen after another;
antimicrobial agents are now lacking to treat a significant proportion of formerly
curable infections caused by nearly a dozen different bacterial species. As the
remaining agents become less often useful, elective surgery and cancer chemother-
apy may become prohibitively risky, trauma care ineffective, premature babies non-
viable, and incidental wounds potentially lethal.
To imagine what it might be like to return to a pre-antibiotic era, consider the
reaction to the introduction of penicillin to public use after World War II. Alexander
Fleming “was showered with gifts of carnations… people whose lives had been
saved by penicillin … now knelt before him to kiss his hands” [3]. In 1964, the city
of Madrid installed statues of Fleming and of a bullfighter saluting him outside the
municipal bullring, because antibiotics had so greatly reduced the lethality of mata-
dors’ wounds.
One of the first postwar impacts of penicillin was the cure of gonorrhea with a
single injection. Yet Neisseria gonorrhoeae is one of the bacterial pathogens some
of whose clinical isolates are now resistant to most antibiotics. Others include
Enterococcus faecium; Staphylococcus aureus; Klebsiella pneumonia; Acinetobacter
baumannii; Pseudomonas aeruginosa; Enterobacter species; some Salmonella,
including invasive, non-typhoidal strains; some Shigella; and Mycobacterium tuber-
culosis. Leaving out the single most prevalent instance of AMR—drug-resistant
tuberculosis— it is estimated that drug-resistant bacterial pathogens now kill some
700,000 people a year, and if present trends continue, the toll will rise to 10 million
deaths per year by 2050 [4]. Authorities seem reluctant to factor drug-resistant
tuberculosis into this tally, perhaps fearing that its unfamiliarity to the citizenry of
economically advanced countries might blunt their concern. Nearly 500,000 people
a year develop drug-resistant tuberculosis; as matters now stand, over 50% of them
will die from it.
After decades of advocacy by scientists and physicians, beginning with Fleming
himself in his Nobel Prize acceptance speech in 1945, acknowledgment of the grav-
ity of AMR has finally come from leaders in business and government, as voiced by
the World Health Organization, the World Economic Forum, the G20, and the G7.
In 2015, President Obama issued a National Action Plan for Combating Antibiotic-­
Resistant Bacteria [5]. In May 2016 a panel commissioned by the British govern-
ment issued cogent recommendations for coordinated global action [6]. In July
412 C. Nathan

2016, NIH, the Department of Defense’s Biomedical Advanced Research and

Development Agency, the Wellcome Trust, the California Life Sciences Institute,
the Massachusetts Biotechnology Council, and the AMR Centre in the United
Kingdom announced that Kevin Outterson, a Boston University law professor inter-
ested in incentives to overcome AMR, will oversee the award of $350 million in
grants via a consortium called the Combating Antibiotic-Resistant Bacteria
Biopharmaceutical Accelerator (CARB-X) [7, 8]. In September 2016, NIH
announced a $20 million Antimicrobial Resistance Diagnostic Challenge, and the
government of China announced a national initiative to counter antimicrobial mis-
use and to find new antimicrobials (http://scim.ag/Chinaresistance). Also in
September 2016, the United Nations General Assembly declared AMR to be a risk
to global health security, placing it alongside HIV/AIDS, noncommunicable dis-
eases, and Ebola virus as only the fourth global health issue prioritized for discus-
sion and action in the history of the General Assembly. The UN’s 193 member
nations agreed to develop an action plan [9].

13.4  MR as a Scientific Challenge; Tuberculosis as a Case

A
in Point

There is now a cross-sector consensus that preserving antibiotics as a mainstay of

human medicine will require overcoming obstacles of four kinds—scientific, regu-
latory, economic, and political [1, 10–12]. Among the several scientific challenges
confronting the development of new antimicrobial agents [13], one stands out as
most needful of fresh thinking: the nature of AMR itself.
The discussion that follows deals only with bacterial infections and antibacterial
agents, now generally called “antibiotics” without regard to whether they are of
microbial origin, as the term was originally used. This focus is for purposes of illus-
tration; it is not meant to discount the urgency of developing antimicrobial agents
for viral, fungal, protist, and helminthic infections.
M. tuberculosis (Mtb) serves, for further focus, for the following reasons [14].
That these four points are all true reveals serious shortcomings in existing approaches
to antibiotic development and use: (i) Mtb is now the single leading cause of death
from infectious disease, (ii) despite causing a curable infection, (iii) one that is now
becoming progressively incurable because of AMR. (Among potentially lethal bac-
terial pathogens displaying AMR, Mtb is estimated to account for the highest num-
ber of cases, even though the vast majority of cases of drug-resistant tuberculosis go
undiagnosed, given that drug sensitivity testing is lacking in many endemic areas.
The fate of people whose tuberculosis displayed extensive AMR was recently moni-
tored: 5% were cured, 73% died, and 10% failed all efforts to treat them and were
discharged into the community in a contagious state [15, 16].) (iv) Even in its drug-­
sensitive form, tuberculosis takes longer to cure than almost any other bacterial
infection.
13 Phenotypic Tolerance and Bacterial Persistence 413

That an immunologic perspective might help derives from four additional points:
(i) Mtb has no known naturally transmitting host but humans. (ii) As noted earlier,
for its transmission, Mtb needs a live human whose immune response is vigorous
enough to liquefy infected lung and erode into an airway. (This dependency proba-
bly accounts for the striking finding that the nucleotide sequences most highly con-
served among 1226 clinical isolates of Mtb were those encoding human T cell
epitopes, that is, the specific oligopeptides within a given protein that bind to anti-
gen receptors on T lymphocytes [16].) (iii) Untreated, the active disease has a fatal-
ity rate of 50% or more. (iv) Nonetheless, after an estimated 70,000 years of
parasitism, neither species—Mtb nor humans—has eliminated the other.
From these considerations we can reach four conclusions: Mtb has evolved the
ability to incite, titrate [17], survive, and exploit the human immune response.
To the degree that we understand the host-pathogen relationship in tuberculosis,
we should be able to apply strategies for drug development that accommodate or
even capitalize on those relationships rather than ignoring them and paying the price
for unappreciated antagonism.

13.5 Heritable AMR

The best understood form of AMR is heritable. There are bacterial genes that encode
resistance to antibiotics that were not invented or deployed at the time that the bac-
teria acquired the genes [18], and it is usually possible to isolate bacteria that have
become heritably resistant to any new antibiotic as soon as there is enough of the
antibiotic on hand to conduct a selection [19]. Apparent exceptions [20–22] are
likely to involve compounds with multiple targets or no specific target. Only a few
such agents are sufficiently selective to be clinically useful. In general, the issue
with heritable AMR is not whether but when the deployment of a given antibiotic
will select for the emergence of heritable resistance in clinical settings.
While correct use of antibiotics will usually lead in time to heritable AMR, other
forms of use hasten its emergence: misuse, overuse, and underuse.
Misuse is exemplified by feeding over half of the United States’ antibiotic ton-
nage to healthy food animals and plants to accelerate their growth; the proportion is
thought to be higher in China [23]. Another form of misuse is the routine failure to
account for individual variation in drug levels attained with standard dosing,
although it is possible to conduct therapeutic drug monitoring on finger-prick blood
spots [24]. Without dose adjustment, peak rifampin levels in the blood vary by
nearly two orders of magnitude in people treated for tuberculosis [25], with some
40–70% being undertreated [26]. Undertreatment fosters the emergence of
resistance.
Overuse results from lack of rapid, point-of-care diagnostics. An estimated 30%
of antibiotic prescriptions in the United States are written for the wrong indication,
typically a viral infection [27]. Overuse is also fostered in settings where the
414 C. Nathan

p­ rescribers are the purveyors or the consumers, that is, where doctors sell the drugs
or patients purchase them without recourse to doctors.
Underuse is a problem when the drugs are diluted by inexpert manufacture or
fraudulent intent or when patients discontinue them prematurely because they feel
better, feel worse, or cannot afford to buy more of them.
Mechanisms of heritable AMR are still being discovered. They include mutation
or posttranslational modification of the target so that it continues to support the
viability of the organism but no longer binds the antibiotic, increased expression of
the target so that it titrates the antibiotic, expression of a pathway that compensates
for the impairment caused by the antibiotic, inactivation of the antibiotic inside the
bacterium [28] or by a secreted bacterial product [29], decreased activation of a
prodrug form of the antibiotic, and decreased uptake or increased export of the
antibiotic.
Discovery of mechanisms of AMR has profoundly impacted both basic science
and clinical care. In basic science, studies of heritable AMR played a prominent role
in introducing the concept that small chemical compounds can have specific macro-
molecular targets in biological systems and can serve as tools to identify the targets’
functions [30]. Clinically, mechanistic understanding of heritable AMR allowed the
design of combination chemotherapy with agents that thwart resistance. For exam-
ple, the World Health Organization’s list of essential medicines includes the combi-
nation of amoxicillin, which is a β-lactam, with clavulanate, an inhibitor of some
bacterial β-lactamases. Moreover, mechanistic understanding of heritable AMR
allows combination chemotherapy with agents to which bacteria manifest resistance
by different mechanisms. Combination chemotherapy was introduced to the prac-
tice of medicine in the 1950s with the discovery that there was no other way to avoid
routine emergence of resistance in the treatment of TB [31]. The practice was later
adopted for the treatment of cancer and HIV/AIDS.

13.6  ntagonism Between Immunity and Antimicrobial

A
Agents

To set the stage for a discussion of phenotypic tolerance as a major form of AMR, it
helps to acknowledge the seemingly paradoxical negative impact of host immunity
on the action of anti-infectives that were developed without taking immunity into
account.
Because a primary function of the immune system is to protect the host from
infection and the purpose of administering antimicrobial agents is the same, then
immunity and antimicrobial chemotherapy can be expected to exert additive or syn-
ergistic effects, and no special effort should be necessary to take advantage of their
common actions.
Indeed, it is sometimes difficult to cure an infection with antibiotics in someone
whose encoded immune system is dysfunctional. For example, most patients with
13 Phenotypic Tolerance and Bacterial Persistence 415

nontuberculous mycobacterial infections who are discovered to have autoantibodies

that neutralize interferon gamma (IFNγ) fail to clear the pathogen in response to
treatment with antimicrobial agents [32]. One of the genes induced by IFNγ is
inducible nitric oxide synthase (iNOS) [33]. Tuberculosis can be cured in most mice
with isoniazid and pyrazinamide [34], but apparent cure is quickly followed by
relapse if the mice are deficient in iNOS [14]. Such observations indicate that anti-
microbial agents not only synergize with host immunity but can depend on host
immunity to effect clinical cure.
At the same time, immune mechanisms often act at cross-purposes with antimi-
crobial agents. When antibiotics are selected for their activity against replicating
bacteria, as is almost always the case, they usually work best, or only, against repli-
cating bacteria. When immunity serves to halt the replication of some infecting
bacteria but fails to kill all of them, as is often the case at the time that an infection
manifests as clinically apparent disease, then immunity can antagonize antibiotic
action. Such antagonism has been demonstrated in axenic culture [35], in cultured
macrophages [36], in rabbits [37], and in mice [38].
In fact, some of the foregoing examples underscore that the same antibiotic and
the same element of host immunity can work both for and against each other in the
same disease. As noted, apparent clinical cure of tuberculosis in mice with isoniazid
and pyrazinamide was sustained in the majority of wild type mice [34] but was
rapidly followed by relapse in all mice that lacked iNOS [14]. Yet the action of iso-
niazid in Mtb-infected mice was partially impaired by iNOS [38] because products
of iNOS block replication of Mtb and, in vitro at least, isoniazid only kills Mtb
when the bacteria are replicating. There may be diverse mechanisms for such antag-
onisms. For example, reactive nitrogen species (RNS) target cytochromes involved
in electron transport; the reduction in energy generation can block uptake of amino-
glycoside antibiotics [39]. Bacteria themselves can generate RNS that induce their
own antioxidant defenses, covalently modify antibiotics, and confer resistance [40].
Host-derived RNS may do the same.
Like generation of RNS, generation of reactive oxygen species (ROS) is a major
element of host immunity against infection. Genetic deficiency in the primary ROS-­
generating enzyme of phagocytes, NADPH oxidase 2 (NOX2), predisposes to life-­
threatening bacterial and fungal infections [41], including by Staphylococcus
aureus. Yet the autotoxicity of NOX2-derived ROS for host myeloid cells can impair
the ability of antibiotics to cure S. aureus pneumonia [42].
When immunity adversely impacts the action of antimicrobial agents, it creates
a form of AMR. The more we understand about the mutual antagonism between
antimicrobial chemotherapy and partially effective host immunity, the more oppor-
tunity we have to identify drug targets in the bacterial pathogen whose inhibition
may convert a non-curative response to chemotherapy into a cure [11, 43].
416 C. Nathan

13.7  onheritable AMR: Phenotypic Tolerance and Its

N
Subtypes

In contrast to the situation with heritable AMR, we have very limited understanding
of nonheritable AMR, also called “phenotypic tolerance,” a term introduced by
Tuomanen [37]. Phenotypic tolerance can be defined as conditional drug resistance
that is not attributable to changes in the nucleic acid sequence of the pathogen’s
genome. Phenotypic tolerance gives rise to bacterial persistence: survival of bacte-
ria during treatment of a host with a drug to which the same strain of pathogen is
susceptible under standard laboratory conditions at concentrations achieved in the
host. Phenotypic tolerance predisposes to emergence of mutants with heritable
resistance [44].
The first two studies of phenotypic tolerance hold such important lessons for
today that they deserve detailed discussion. The purification of penicillin was
reported in 1942 [45]. That same year, Gladys Hobby and her colleagues reported
that at 37 °C, about 1 streptococcus remained viable after 48 hours of exposure to
penicillin for every 106 present in the control culture at the end of that period. The
authors did not comment on that but drew attention to the survival of nearly all the
penicillin-treated streptococci if the exposure took place at 4 °C, conditions in
which there was no increase in bacterial number in the untreated control culture.
The authors concluded, “It is apparent that penicillin is capable of destroying bacte-
ria only if multiplication takes place” [46].
In 1944, Joseph Bigger repeated and extended the experiments using staphylo-
cocci [47]. He introduced the term “persisters” to stress the observation that about
1 in 106 staphylococci survived the treatment of logarithmically replicating cultures
at body temperature. He inferred that persisters to penicillin must be “cocci …
which happen to be, when exposed to it, in a phase in which they are insusceptible
to its action,” because “If persisters had an abnormally high resistance, either natu-
ral [that is, heritable and existing prior to the experiment] or acquired [that is, heri-
table but acquired during the experiment], it is probable that their descendants
would also possess abnormally high resistance. The descendants of a number of
persisters which had survived contact with 1 unit per c.cm. penicillin for 3–5 days
were found to be killed by 1/8 unit per c.cm. within 46 hours and to have no greater
tendency than normal forms to produce persisters” [47].
Bigger went on to confirm the observation of Hobby et al. [46] that cooling the
bacteria elevated the frequency of persisters to nearly 100%, that is, by 6 orders of
magnitude. He demonstrated the same effect by acidifying the medium or lowering
its tonicity. He concluded that “persisters are cocci which survive contact with peni-
cillin because they are in dormant, non-dividing phase” [47].
In fact, within 2 years of the report of penicillin’s publication, the two groups
mentioned above, working on two continents with two different pathogens, had each
observed two different classes of phenotypic tolerance, but without distinguishing
them. It took another 70 years before the distinction was made, driven by the recog-
nition that the two classes have different implications for drug discovery [11].
13 Phenotypic Tolerance and Bacterial Persistence 417

Class I phenotypic tolerance can be viewed as a form of bacterial bet-hedging

manifest by a minority of a population in conditions permissive for growth. The
upper limit of the size of the minority population that can display class I phenotypic
tolerance is set by the precision of the assay used to determine the minimum inhibi-
tory concentration (MIC) of the antibiotic. If the MIC is defined as the concentra-
tion that inhibits growth by 90%, then 10% of the population could be phenotypically
tolerant without changing the population’s MIC. Typically, in a wild type popula-
tion, the frequency of class I phenotypic tolerance is about 1 in 106. Certain muta-
tions can increase the frequency of class I phenotypic tolerance by orders of
magnitude without changing the MIC and without conferring heritable AMR. The
phenotypically tolerant minority may be non-replicating at the time, as Hobby et al.
[46] and Bigger [47] inferred and others then assumed and asserted, or it may be
replicating, as documented in later studies. The key feature is that a population of
class I persisters, once expanded in the absence of the antibiotic, succumbs in the
same proportion to the same concentration of antibiotic as did the population from
which the persisters were recovered.
Heritable AMR can emerge more readily after antibiotics select for a mutation
that increases the frequency of class I phenotypically tolerant bacteria in the popula-
tion. Such mutations can arise in diverse genes, including those encoding antitoxins
or enzymes that catalyze metabolic processes [44]. Mutations that augment class I
phenotypic tolerance increase the proportion of bacteria that survive one exposure
to antibiotic, providing a larger population in which mutants may arise that confer
heritable resistance to a subsequent exposure [44].
In contrast, class II phenotypic tolerance is a bacterial response to exogenous
stress, including non-sterilizing immunity. It is imposed by conditions that impair
growth and pertains to all of the bacteria whose growth is impaired, which may
be most or all of the bacterial population in a given site at the time that chemo-
therapy is administered. Conditions that impair growth can be imposed by the
host environment, host immune chemistries, or exposure to sublethal levels of
other antibiotics (Table 13.1). The stresses that lead to class II phenotypic toler-
ance can foster the emergence of heritable AMR by increasing the frequency of
mutation [49, 50].
A particularly challenging form of class II phenotypic tolerance is displayed by
bacteria whose non-replicative state is not reversed by plating them on a rich
medium rendered semisolid with agar. That is, they are not colony-forming units,
yet their viability is demonstrable by some other means, such as growth after limit-
ing dilution in liquid culture or injection into an experimental host. Over 80 bacte-
rial species have been shown to have the property of becoming what Rita Colwell
and colleagues originally called “viable but non-culturable” [51]. Strikingly, in
two studies to date, most of the Mtb in the sputum of most treatment-naïve patients
with tuberculosis were unable to replicate as CFU and were detected instead by
limiting dilution [52–54]. Similarly, limiting dilution rather than plating on agar
was necessary to detect 90–99% of the Mtb remaining in vitro after sequential
starvation and exposure to a rifamycin in an in vitro model of “differentially detect-
able” Mtb [55].
418 C. Nathan

Table 13.1 Classes of phenotypic tolerance and their therapeutic implications

Class I Class II
Growth state Most cells replicating Most cells not replicating
of bacterial
population
Persistence Small minority; different cells Large majority; same cells tolerate many
phenotype tolerate different antibiotics antibiotics
Inducers of Unknown; stochastic Acidification, ROS, RNS, hypoxia,
persistence deprivation of C, N, P or Fe; sublethal
exposure to antibiotics
Speculative Epigenetic, transcriptional, Decreased uptake, increased export, or
mechanisms translational, or posttranslational increased catabolism of drug; metabolic
expression or suppression of any stress leading to oxidative stress and
process for which genetic change adaptation; increase in proteostasis
can produce heritable resistance pathways; preferential transcription and
translation; alternate respiratory pathways
and electron acceptors
Therapeutic Combine different drugs that each Include new kinds of drugs active on
implications reach the sites of infection non-replicating cells that reach the sites of
infection
Based on Nathan [11] and modified from Nathan and Barry [48]

Not all the anti-infectives that kill Mtb in some non-replicating states kill Mtb in
other non-replicating states. For example, rifampin generated rather than killed the
differentially detectable Mtb described above, while thioridazine did not generate
such cells but did kill them [55]. These antimicrobial agents serve as chemical
probes to teach us that class II phenotypic tolerance encompasses a spectrum of
states—at our present state of knowledge, at least two. Class IIa phenotypic toler-
ance is characteristic of bacteria that stop replicating in response to a given set of
stresses but form CFU when those stresses are relieved. Class IIb phenotypic toler-
ance is a feature of bacteria that stop replicating in response to different stresses and
remain viable when those stresses are removed, but do not form CFU [55]. This
complicates the task of finding anti-infectives that can kill bacteria displaying phe-
notypic tolerance.
To the extent that individual bacteria in an otherwise antibiotic-susceptible popu-
lation manifest class I phenotypic tolerance to two different antibiotics by different
mechanisms, then the cells that are phenotypically tolerant to the first antibiotic are
likely to be susceptible to the second. In such a case, to kill the whole population, it
should suffice to combine antibiotics in such a way that no one bacterium is pheno-
typically tolerant to all of them, provided that each of the drugs in the combination
reaches the bacteria in adequate concentrations at the same time. (In vitro, class I
phenotypically tolerant Mtb could be killed by forcing them to produce extra ROS
in the presence of rifampin or isoniazid by supplying them with small thiols [56].)
In contrast, if all the bacteria in a population are phenotypically tolerant to several
different antibiotics, then each individual bacterium must be tolerant to each of
13 Phenotypic Tolerance and Bacterial Persistence 419

them, and combinations of those antibiotics are unlikely to be effective. Instead, it

will be necessary to discover antibiotics that can kill non-replicating bacteria.
The foregoing theses constitute a practical imperative for distinguishing classes
of phenotypic tolerance (Table 13.1). Other classifications of nonheritable AMR
are also useful, for example, to frame mechanistic questions [57]. A caveat of all
classifications based on in vitro observations is that the relationship is complex
and variable between the MIC measured in low-protein, host cell-free media over
short periods of time and the dosing regimens of antibiotics required for clinical
cure [58, 59].

13.8 Mechanisms of Class I Phenotypic Tolerance

Class I phenotypic tolerance can theoretically arise by any mechanism that confers
heritable AMR, from epigenetic regulation to posttranslational modification, as
long as the mechanism does not depend on a change in the pathogen’s coding
sequence. As noted earlier, the size of the tolerant subpopulation may be affected by
a change in coding sequence, as long as the tolerant subpopulation remains such a
minority that the overall population does not manifest an increase in the antibiotic’s
MIC.
Much of the research in this field has wrestled with a descriptive question,
whether class I phenotypic tolerance is as tightly linked with non-replication as
Hobby et al. [46] and Bigger [47] inferred. In short, the answer is “no.”
The first study to use time-lapse photomicroscopy of bacteria in microfluidic
chambers to study phenotypic tolerance at the single cell level [60] revealed that in
an otherwise replicating population of E. coli, most of the few cells that survived
ampicillin were non-replicating at the time of exposure to the drug. However, some
of the other surviving E. coli had been replicating. This study was rendered feasible
by using E. coli with compound mutations in hipA that raised the frequency of class
I phenotypically tolerant E. coli by several orders of magnitude without changing
the MIC of the overall population.
Nine years later, a study of similar design reached a different conclusion while
studying the action of isoniazid on Mtb [61]. Isoniazid is a prodrug whose activation
depends on the Mtb catalase-peroxidase KatG. The investigators showed that sto-
chastic extinction of KatG expression conferred resistance to isoniazid. Growth rate
had nothing to do with it [61].
The same year, Orman and Brynildsen showed that E. coli persisters to ampicil-
lin and fluoroquinolones are enriched among the non-replicating subpopulation, but
not confined to it nor highly prevalent in it [62]. Natural clinical and veterinary
isolates of E. coli each showed the same MICs to a given antibiotic, yet each showed
different levels of persistence to different sets of antibiotics [63]. This suggested
that different individual cells were phenotypically tolerant to different antibiotics,
meaning that non-replication of a given cell could not be a universal explanation for
phenotypic tolerance.
420 C. Nathan

The same conclusion was reached in studies of persisters among antibiotic-­

stressed E. coli during diauxic transition. The frequency with which E. coli per-
sisted in the face of exposure to ampicillin increased from about 1 in 104 to about
1 in 2.5 × 103 during the transition from replication in glucose to utilization of fuma-
rate [64]. Results were similar with ofloxacin. However, co-treatment with ampicil-
lin and ofloxacin reduced the frequency of persisters in diauxie by about tenfold,
suggesting that about 90% of them were phenotypically tolerant to one or the other
of the antibiotics, but not both [64]. In this instance as well, non-replication could
not serve as a universal explanation for phenotypic tolerance to all antibiotics tested.
Working with Mtb, Javid and co-workers discovered a growth-rate independent
form of class I phenotypic tolerance to rifampin and defined its molecular mecha-
nism [65]. Individual Mtb cells mistranslate different proportions of individual cop-
ies of rifampin’s target, RNA polymerase subunit B (RpoB). The basis of
mistranslation is the propensity of Mtb’s glutaminyl-tRNA synthetase to charge
tRNA not only with glutamine but also with glutamate and of Mtb’s asparaginyl-­
tRNA synthetase to charge tRNA not only with asparagine but also with aspartate.
The errors are corrected by a glutamine amidotransferase, but not perfectly. If a
given cell’s collection of RpoB molecules includes enough copies in which Asn170
has been replaced with Asp, the cell can survive a dose of rifampin that kills geneti-
cally identical siblings. Heritable mutations in the gene encoding a subunit of the
amidotransferase increased the frequency of class I phenotypically tolerant Mtb in
a population but, as with hipA mutations in E. coli discussed above, did not allow
the persisters, when grown up without antibiotic, to display a higher MIC than the
population from which they were recovered [65].
Some view class I phenotypic tolerance as an outcome of noise: random varia-
tion arising from imperfect execution or synchronization of various processes. In
contrast, others argue that the high value of class I phenotypic tolerance for survival
of a replicating population in the face of emergent stress, together with its suscepti-
bility to genetic regulation, make a case for the existence of specific, evolved mech-
anisms. Both views are likely to be correct, depending on the setting.
Our understanding of class I phenotypic tolerance in diverse bacterial species
would be greatly enriched if we could study the phenomena not only in mono-­
species planktonic cultures in optimal growth media during exposure to clinically
relevant antibiotic concentrations but in natural, multi-species environments with
their complex chemical language of cooperation and competition.

13.9 Mechanisms of Class II Phenotypic Tolerance

One of the most important challenges for antibiotic research is to understand mech-
anisms of class II phenotypic tolerance, a state for which incompletely effective
immunity and sublethal antibiotic therapy bear much of the responsibility.
We have a long way to go. We do not know if a given bacterial species that enters
a non-replicating state in response to different host conditions manifests class II
13 Phenotypic Tolerance and Bacterial Persistence 421

phenotypic tolerance to the same antibiotic by different mechanisms nor whether a

given bacterial species that enters a non-replicating state in response to the same
host condition manifests class II phenotypic tolerance to different antibiotics by dif-
ferent mechanisms.
Following the reasoning that Bigger advanced three quarters of a century ago
[47], some scientists today argue that non-replicating bacteria are phenotypically
tolerant to inhibitors of biosynthetic processes because they are “dormant,” where
dormancy is inferred from the cells’ survival of exposure to inhibitors of biosyn-
thetic processes. For example, it was recently stated that “Tolerance is a property of
dormant, nongrowing bacterial cells in which antibiotic targets are inactive, allow-
ing bacteria to survive.” [66].
Such reasoning is circular. Although class II phenotypic tolerance is associated
with non-replication by definition, non-replication does not constitute a mechanistic
explanation of class II phenotypic tolerance. In fact, non-replication offers bacteria
no blanket reprieve from the need for biosynthetic processes, such as generation of
energy to maintain membrane potential. Generation of energy requires the action of
enzymes. Stresses associated with imposition of non-replication cause damage to
macromolecules. Some such damage is reparable; most repair requires energy.
Some damage is irreparable. Replacement of irreparably damaged molecules
requires synthesis, which again requires energy, and usually requires transcription
as well. Indeed, non-replicating Mtb maintains its membrane potential [67–69] and
a large, altered transcriptome [70, 71].
In short, non-replication is a state associated with class II phenotypic tolerance
but not a mechanism accounting for it. Only recently have underlying mechanisms
begun to come into focus. Non-replicating states can lead to reduced antibiotic
uptake [72] or reduced retention [73] and perhaps to altered drug catabolism. Stress
can lead to upregulation of antioxidant pathways, as seen, for example, in a pro-
teomic analysis of M. smegmatis exposed to sublethal concentrations of rifampin
[74]. To the extent that antibiotic action is augmented by generation of reactive
oxygen species secondary to disordered metabolism [75], the increase in antioxi-
dant defenses may contribute to phenotypic tolerance [76], as may the increased
expression of proteostasis pathways for macromolecular preservation and repair.
Non-replicating bacteria may switch to alternate respiratory pathways and use alter-
nate electron acceptors. During non-replication, an essential process may occur so
slowly that its corruption by the antibiotic only leads to death after the period of
observation. Condition-dependent changes in gene essentiality may lead to prioriti-
zation of the transcription and translation of newly essential genes in the face of
partial inhibition of overall transcription or translation.
It is a separate question how stresses suppress replication. Some stresses limit the
supply of exogenous precursors for an increase in biomass. Many stresses activate
the stringent response, leading to inactivation of antitoxins in toxin-antitoxin mod-
ules, of which Mtb has over 80 [77]. The activated toxins can cleave specific tRNAs,
mRNAs, or ribosomal RNAs; phosphorylate and inhibit specific tRNA synthetases;
interfere with DNA gyrase; ADP-ribosylate DNA [78]; and reduce the proton
motive force [77]. The stringent response in some organisms includes induction of
422 C. Nathan

hibernation factor and ribosome modulation factor, proteins that bind ribosomes
and inhibit translation [79]. It is clear how these actions could suppress replication,
but as noted above, suppression of replication does not suffice as a general explana-
tion of phenotypic tolerance.

13.10 I s It Possible to Find New Antibiotics that Can Kill

Bacteria Displaying Class II Phenotypic Tolerance
to Existing Antibiotics?

Tuberculosis illustrates the importance of answering this question. A central hypoth-

esis is that class II phenotypic tolerance to existing TB drugs is a major contributor
to the failure of these drugs to reduce the time it takes to cure TB to less than
6 months for over 86% of individuals with drug-sensitive disease. If most of the Mtb
at a given site in the host are non-replicating because of conditions they encounter
at that site, such as hypoxia, nutritional restriction, acidity, or reactive species of
oxygen or nitrogen, and, in association with those conditions, are phenotypically
tolerant to every antibiotic that reaches the site, then chemotherapy that combines
those drugs is not likely to be effective.
The following considerations illustrate one way that immunologic thinking can
suggest new targets for unconventional antibiotics against Mtb to complement the
action of conventional antibiotics.
Mechanisms of Mtb’s resistance to host immunity can be understood in terms of
successive lines of resistance. First, Mtb can suppress host immunity (e.g., [80]).
Failing that or in addition, Mtb can detoxify host effector molecules (e.g., [68, 81–
84]). Next, the pathogen can adapt to effector molecules whose production it failed
to block and whose level it failed to reduce (e.g., [85]). If macromolecules are none-
theless damaged, the bacteria can repair them (e.g., [86]). If repair is inadequate, the
bacteria can degrade damaged macromolecules to avoid their toxic gain of function
(e.g., [87, 88]). Some macromolecules are too damaged to be repaired, such as irre-
versibly oxidized proteins that cannot be unfolded for degradation by chambered
proteases. These can be sequestered [89]. If all else fails, some bacteria can survive
long periods without replicating, awaiting the return of conditions in which replica-
tion can be sustained. In many cases, enzymes have been identified that mediate
these microbial defenses and compounds have been identified that inhibit these
enzymes [81, 88, 90–92]. Where human homologs exist, it has been possible to
identify Mtb-selective inhibitors that spare the corresponding human enyzmes [81,
88, 90–92].
Almost all antibiotics that were selected on the basis of their ability to kill repli-
cating bacteria are much less effective, or ineffective, against the same organisms
when they are non-replicating. While rifampin, fluoroquinolones, and bedaquiline
are active against non-replicating Mtb, much of that effect in short-term in vitro
assays appears to be an artifact of carry-over of antibiotic from the non-replicating
13 Phenotypic Tolerance and Bacterial Persistence 423

stage of the assay to the stage of the assay where recovery is assessed under condi-
tions that support replication [93]. Rifampin has genuine bactericidal action on non-­
replicating Mtb in vitro but at far higher concentrations than needed to kill replicating
Mtb, and even then, the maximum extent of killing in vitro is far less [93]. This is
not meant to disparage the proven clinical utility of these drugs but rather to suggest
that they do not represent an ideal solution to the problem of class II phenotypic
tolerance.
Fortunately, compounds can be found that extensively kill bacteria in a state that
confers class II phenotypic tolerance to conventional antibiotics. An early example
was a thioxothiazolidine that killed Mtb only when the Mtb was non-replicating,
without regard to diverse conditions tested that imposed non-replication [81].
Another target-based screen led to two chemically distinct classes of Mtb-selective
proteasome inhibitors [88, 92] that killed Mtb that was rendered non-replicating by
nitrosative stress [88, 92] or starvation [94]. A whole-cell screen designed to iden-
tify compounds that kill non-replicating Mtb identified oxyphenbutazone [35] and
other compounds [95]. Subsequently, over 100 compounds have been reported to
kill non-replicating Mtb selectively, including novel cephalosporins [96]. However,
in only a few cases did the investigators exclude the possibility that carry-over of
compound into the replicative phase of the assay may have led to a false impression
of activity in the preceding, non-replicative phase of the assay [30].
Why are some compounds only able to kill non-replicating bacteria, sparing the
same cells when they replicate? Barring compound modification under one of the
two sets of assay conditions, and assuming equivalent uptake under both, the ques-
tion becomes why some targets are nonessential under conditions that support rep-
lication but essential under conditions that do not. For example, at least four sets of
Mtb enzymes involved in central carbon metabolism—hydroxyoxoadipate syn-
thase, dihydrolipoamide acyltransferase, lipoamide dehydrogenase, and the isoci-
trate lyases—are dispensable for survival under nonstressed conditions but become
essential for Mtb to withstand oxidative or nitrosative stresses that impose non-­
replication [68, 76, 81, 84]. This invites the speculation that some pathways that
would afford redundancy in a critical function targeted by the antibiotic are inacti-
vated under non-replicative conditions, or a singular essential pathway incompletely
inhibited by the antibiotic is further inhibited by the non-replicative conditions.
Even more encouraging are antibiotics that can kill bacteria extensively not only
when they are replicating but also when they are not replicating and are phenotypi-
cally tolerant to other antibiotics. With respect to tuberculosis, this has been reported
with 8-hydroxyquinolines [97, 98] and nitazoxanide, an antibiotic approved for
other indications [20]. In vitro, the nitroimidazole PA-824 (Pretomanid) kills both
replicating and non-replicating Mtb to comparable extents and at comparable con-
centrations [30, 99]. Under non-replicating conditions, the mechanism involves
generation of reactive nitrogen species [99], a striking example of a synthetic anti-
biotic mimicking host immunity [100].
424 C. Nathan

Major Points
• Phenotypic tolerance prevents an antimicrobial agent from eradicating a patho-
gen population; it likely accounts for relapse and contributes to the emergence of
heritable resistance.
• Type I phenotypic tolerance occurs when a minority (subpopulation) survives
antibiotic treatment in conditions permissive for growth of the majority popula-
tion and individual tolerant bacterial cells are each tolerant to a different
antibiotic.
• Type II phenotypic tolerance is a bacterial response to exogenous stress that
impairs growth and pertains to all of the bacteria whose growth is impaired and
individual bacterial cells are each tolerant to multiple antibiotics.
• Host cell immunity can foster phenotypic tolerance and thereby work at cross-­
purposes with antimicrobials.
• Better mechanistic understanding of the different classes of phenotypic tolerance
will help improve antimicrobial chemotherapy and help reduce the emergence of
heritable antimicrobial resistance.

Acknowledgments I am grateful to K. Burns-Huang, B. Gold, K. Rhee, K. Saito, and T. Warrier

(Weill Cornell Medicine) for critical comments. I am thankful for the support of the Tri-Institutional
TB Research Unit (NIH U19 AI111143) and the Milstein Program in Chemical Biology and
Translational Medicine. The Department of Microbiology and Immunology is supported by the
William Randolph Hearst Charitable Trust.

References

1. Nathan C. Cooperative development of antimicrobials: looking back to look ahead. Nat Rev
Microbiol. 2015;13(10):651–7.
2. Payne DJ, Gwynn MN, Holmes DJ, Pompliano DL. Drugs for bad bugs: confronting the chal-
lenges of antibacterial discovery. Nat Rev Drug Discov. 2007;6(1):29–40.
3. Brown K. Penicillin man: Alexander Fleming and the antibiotic revolution. Stroud: History
Press; UK. 2013.
4. O’Neill J. Tackling drug-resistant infections globally. Review on Antimicrobial Resistance.
London, England, UK. 2014.
5. White House. National action plan for combating antibiotic-resistant bacteria. Washington,
DC. 2015. 62 pp.
6. O’Neill J. Tackling drug resistant infections globally: final report and recommendations.
Review on Antimicrobial Resistance. London, England, UK. 2016.
7. Outterson K, Rex JH, Jinks T, Jackson P, Hallinan J, Karp S, et al. Accelerating global
innovation to address antibacterial resistance: introducing CARB-X. Nat Rev Drug Discov.
2016;15(9):589–90.
8. Bagley N, Outterson K. How to avoid a post-antibiotic world. New York Times. 2017:Op-Ed.
9. United Nations. Draft political declaration of the high-level meeting of the General Assembly
on antimicrobial resistance. New York City, New York. 2016.
10. Nathan C. Antibiotics at the crossroads. Nature. 2004;431(7011):899–902.
11. Nathan C. Fresh approaches to anti-infective therapies. Sci Transl Med. 2012;4(140):140sr2.
12. Nathan C, Cars O. Antibiotic resistance–problems, progress, and prospects. N Engl J Med.
2014;371(19):1761–3.
13 Phenotypic Tolerance and Bacterial Persistence 425

13. Nathan C. Making space for anti-infective drug discovery. Cell Host Microbe.
2011;9(5):343–8.
14. Nathan C. Taming tuberculosis: a challenge for science and society. Cell Host Microbe.
2009;5(3):220–4.
15. Pietersen E, Ignatius E, Streicher EM, Mastrapa B, Padanilam X, Pooran A, et al. Long-term
outcomes of patients with extensively drug-resistant tuberculosis in South Africa: a cohort
study. Lancet. 2014;383(9924):1230–9.
16. Coscolla M, Copin R, Sutherland J, Gehre F, de Jong B, Owolabi O, et al. M. tuberculosis
T cell epitope analysis reveals paucity of antigenic variation and identifies rare variable TB
antigens. Cell Host Microbe. 2015;18(5):538–48.
17. Marakalala MJ, Raju RM, Sharma K, Zhang YJ, Eugenin EA, Prideaux B, et al.
Inflammatory signaling in human tuberculosis granulomas is spatially organized. Nat Med.
2016;22(5):531–8.
18. Bhullar K, Waglechner N, Pawlowski A, Koteva K, Banks ED, Johnston MD, et al. Antibiotic
resistance is prevalent in an isolated cave microbiome. PLoS One. 2012;7(4):e34953.
19. Kling A, Lukat P, Almeida DV, Bauer A, Fontaine E, Sordello S, et al. Antibiotics. Targeting
DnaN for tuberculosis therapy using novel griselimycins. Science. 2015;348(6239):1106–12.
20. de Carvalho LP, Lin G, Jiang X, Nathan C. Nitazoxanide kills replicating and nonreplicating
Mycobacterium tuberculosis and evades resistance. J Med Chem. 2009;52(19):5789–92.
21. Ling LL, Schneider T, Peoples AJ, Spoering AL, Engels I, Conlon BP, et al. A new antibiotic
kills pathogens without detectable resistance. Nature. 2015;517(7535):455–9.
22. Moreira W, Aziz DB, Dick T. Boromycin kills mycobacterial persisters without detectable
resistance. Front Microbiol. 2016;7:199.
23. Van Boeckel TP, Brower C, Gilbert M, Grenfell BT, Levin SA, Robinson TP, et al. Global
trends in antimicrobial use in food animals. Proc Natl Acad Sci U S A. 2015;112(18):5649–54.
24. Alsultan A, Peloquin CA. Therapeutic drug monitoring in the treatment of tuberculosis: an
update. Drugs. 2014;74(8):839–54.
25. Wilkins JJ, Savic RM, Karlsson MO, Langdon G, McIlleron H, Pillai G, et al. Population phar-
macokinetics of rifampin in pulmonary tuberculosis patients, including a semimechanistic
model to describe variable absorption. Antimicrob Agents Chemother. 2008;52(6):2138–48.
26. Um SW, Lee SW, Kwon SY, Yoon HI, Park KU, Song J, et al. Low serum concentrations
of anti-tuberculosis drugs and determinants of their serum levels. Int J Tuberc Lung Dis.
2007;11(9):972–8.
27. Fleming-Dutra KE, Hersh AL, Shapiro DJ, Bartoces M, Enns EA, File TM Jr, et al. Prevalence
of inappropriate antibiotic prescriptions among US Ambulatory Care Visits, 2010–2011.
JAMA. 2016;315(17):1864–73.
28. Warrier T, Kapilashrami K, Argyrou A, Ioerger TR, Little D, Murphy KC, et al. N-methylation
of a bactericidal compound as a resistance mechanism in Mycobacterium tuberculosis. Proc
Natl Acad Sci U S A. 2016;113(31):E4523–30.
29. El-Halfawy OM, Klett J, Ingram RJ, Loutet SA, Murphy ME, Martin-Santamaria S, et al.
Antibiotic capture by bacterial Lipocalins uncovers an extracellular mechanism of intrinsic
antibiotic resistance. MBio. 2017;8(2)
30. Gold B, Nathan C. Targeting phenotypically tolerant Mycobacterium tuberculosis. In: Jacobs
Jr WR, McShane H, Mizrahi V, Orme I, editors. Tuberculosis and the tubercle Bacillus. 2nd
ed: American Society of Microbiolgy Press; 2017.
31. Fox W, Sutherland IA. Five-year assessment of patients in a controlled trial of streptomy-
cin, Para-aminosalicylic acid, and streptomycin plus Para-aminosalicylic acid, in pulmonary
tuberculosis. Q J Med. 1956;25(98):221–43.
32. Lin CH, Chi CY, Shih HP, Ding JY, Lo CC, Wang SY, et al. Identification of a major epitope
by anti-interferon-gamma autoantibodies in patients with mycobacterial disease. Nat Med.
2016;22(9):994–1001.
33. Xie QW, Cho HJ, Calaycay J, Mumford RA, Swiderek KM, Lee TD, et al. Cloning and
characterization of inducible nitric oxide synthase from mouse macrophages. Science.
1992;256(5054):225–8.
426 C. Nathan

34. McCune RM, Feldmann FM, Lambert HP, McDermott W. Microbial persistence.
I. The capacity of tubercle bacilli to survive sterilization in mouse tissues. J Exp Med.
1966;123(3):445–68.
35. Gold B, Pingle M, Brickner SJ, Shah N, Roberts J, Rundell M, et al. Nonsteroidal anti-­
inflammatory drug sensitizes Mycobacterium tuberculosis to endogenous and exogenous
antimicrobials. Proc Natl Acad Sci U S A. 2012;109(40):16004–11.
36. Helaine S, Cheverton AM, Watson KG, Faure LM, Matthews SA, Holden DW. Internalization
of Salmonella by macrophages induces formation of nonreplicating persisters. Science.
2014;343(6167):204–8.
37. Tuomanen E. Phenotypic tolerance: the search for beta-lactam antibiotics that kill nongrow-
ing bacteria. Rev Infect Dis. 1986;8(Suppl 3):S279–91.
38. Liu Y, Tan S, Huang L, Abramovitch RB, Rohde KH, Zimmerman MD, et al. Immune activa-
tion of the host cell induces drug tolerance in Mycobacterium tuberculosis both in vitro and
in vivo. J Exp Med. 2016;213(5):809–25.
39. Zemke AC, Gladwin MT, Bomberger JM. Sodium nitrite blocks the activity of amino-
glycosides against Pseudomonas aeruginosa biofilms. Antimicrob Agents Chemother.
2015;59(6):3329–34.
40. Gusarov I, Shatalin K, Starodubtseva M, Nudler E. Endogenous nitric oxide protects bacteria
against a wide spectrum of antibiotics. Science. 2009;325(5946):1380–4.
41. Group ICGDCS. A controlled trial of interferon gamma to prevent infection in chronic gran-
ulomatous disease. The International Chronic Granulomatous Disease Cooperative Study
Group. N Engl J Med. 1991;324(8):509–16.
42. Sun K, Yajjala VK, Bauer C, Talmon GA, Fischer KJ, Kielian T, et al. Nox2-derived oxidative
stress results in inefficacy of antibiotics against post-influenza S. aureus pneumonia. J Exp
Med. 2016;213(9):1851–64.
43. Nathan C, Gold B, Lin G, Stegman M, de Carvalho LP, Vandal O, et al. A philosophy of
anti-infectives as a guide in the search for new drugs for tuberculosis. Tuberculosis (Edinb).
2008;88(Suppl 1):S25–33.
44. Levin-Reisman I, Ronin I, Gefen O, Braniss I, Shoresh N, Balaban NQ. Antibiotic tolerance
facilitates the evolution of resistance. Science. 2017;355(6327):826–30.
45. Abraham EP, Chain E. Purification of penicillin. Nature. 1942;149:328.
46. Hobby GL, Meyer K, Chaffee E. Observations on the mechanism of action of penicillin. Proc
Soc Exp Biol Med. 1942;50:281–5.
47. Bigger J. Treatment of staphylococcal infections with penicillin by intermittent sterilisation.
Lancet. 1944;244:497–500.
48. Nathan C, Barry CE 3rd. TB drug development: immunology at the table. Immunol Rev.
2015;264(1):308–18.
49. Kohanski MA, DePristo MA, Collins JJ. Sublethal antibiotic treatment leads to multidrug
resistance via radical-induced mutagenesis. Mol Cell. 2010;37(3):311–20.
50. Boshoff HI, Reed MB, Barry CE 3rd, Mizrahi V. DnaE2 polymerase contributes to in vivo
survival and the emergence of drug resistance in Mycobacterium tuberculosis. Cell.
2003;113(2):183–93.
51. Xu HS, Roberts N, Singleton FL, Attwell RW, Grimes DJ, Colwell RR. Survival and viability
of nonculturable Escherichia coli and Vibrio cholerae in the estuarine and marine environ-
ment. Microb Ecol. 1982;8(4):313–23.
52. Chengalroyen MD, Beukes GM, Gordhan BG, Streicher EM, Churchyard G, Hafner R, et al.
Detection and quantification of differentially culturable Tubercle Bacteria in sputum from
tuberculosis patients. Am J Respir Crit Care Med. 2016;194(12):1532–40.
53. Dartois V, Saito K, Warrier T, Nathan C. Editorial: new evidence for the complexity of the
population structure of Mycobacterium tuberculosis increases the diagnostic and biologic
challenges. Am J Resp Crit Care Med. 2016;194:1448–50.
54. Mukamolova GV, Turapov O, Malkin J, Woltmann G, Barer MR. Resuscitation-promoting
factors reveal an occult population of tubercle Bacilli in Sputum. Am J Respir Crit Care Med.
2010;181(2):174–80.
13 Phenotypic Tolerance and Bacterial Persistence 427

55. Saito K, Warrier T, Somersan-Karakaya S, Kaminski L, Mi J, Jiang X, et al. Rifamycin action

on RNA polymerase in antibiotic-tolerant Mycobacterium tuberculosis results in differen-
tially detectable populations. Proc Natl Acad Sci U S A. 2017;114:E4832–40.
56. Vilcheze C, Hartman T, Weinrick B, Jain P, Weisbrod TR, Leung LW, et al. Enhanced respira-
tion prevents drug tolerance and drug resistance in Mycobacterium tuberculosis. Proc Natl
Acad Sci U S A. 2017;114(17):4495–500.
57. Brauner A, Fridman O, Gefen O, Balaban NQ. Distinguishing between resistance, tolerance
and persistence to antibiotic treatment. Nat Rev Microbiol. 2016;14(5):320–30.
58. Frimodt-Moller N. Correlation between pharmacokinetic/pharmacodynamic parameters and
efficacy for antibiotics in the treatment of urinary tract infection. Int J Antimicrob Agents.
2002;19(6):546–53.
59. Mueller M, de la Pena A, Derendorf H. Issues in pharmacokinetics and pharmacody-
namics of anti-infective agents: kill curves versus MIC. Antimicrob Agents Chemother.
2004;48(2):369–77.
60. Balaban NQ, Merrin J, Chait R, Kowalik L, Leibler S. Bacterial persistence as a phenotypic
switch. Science. 2004;305(5690):1622–5.
61. Wakamoto Y, Dhar N, Chait R, Schneider K, Signorino-Gelo F, Leibler S, et al. Dynamic
persistence of antibiotic-stressed mycobacteria. Science. 2013;339(6115):91–5.
62. Orman MA, Brynildsen MP. Dormancy is not necessary or sufficient for bacterial persis-
tence. Antimicrob Agents Chemother. 2013;57(7):3230–9.
63. Stewart B, Rozen DE. Genetic variation for antibiotic persistence in Escherichia coli.
Evolution. 2012;66(3):933–9.
64. Amato SM, Brynildsen MP. Persister heterogeneity arising from a single metabolic stress.
Curr Biol. 2015;25(16):2090–8.
65. Su HW, Zhu JH, Li H, Cai RJ, Ealand C, Wang X, et al. The essential mycobacterial ami-
dotransferase GatCAB is a modulator of specific translational fidelity. Nat Microbiol.
2016;1(11):16147.
66. Lewis K, Shan Y. Why tolerance invites reistance. Science. 2017;355:796.
67. de Carvalho LP, Darby CM, Rhee KY, Nathan C. Nitazoxanide disrupts membrane potential
and intrabacterial pH homeostasis of Mycobacterium tuberculosis. ACS Med Chem Lett.
2011;2(11):849–54.
68. Venugopal A, Bryk R, Shi S, Rhee K, Rath P, Schnappinger D, et al. Virulence of
Mycobacterium tuberculosis depends on lipoamide dehydrogenase, a member of three multi-
enzyme complexes. Cell Host Microbe. 2011;9(1):21–31.
69. Darby CM, Ingolfsson HI, Jiang X, Shen C, Sun M, Zhao N, et al. Whole cell screen for
inhibitors of pH homeostasis in Mycobacterium tuberculosis. PLoS One. 2013;8(7):e68942.
70. Schnappinger D, Ehrt S, Voskuil MI, Liu Y, Mangan JA, Monahan IM, et al. Transcriptional
adaptation of Mycobacterium tuberculosis within macrophages: insights into the phagosomal
environment. J Exp Med. 2003;198(5):693–704.
71. Voskuil MI, Schnappinger D, Visconti KC, Harrell MI, Dolganov GM, Sherman DR, et al.
Inhibition of respiration by nitric oxide induces a Mycobacterium tuberculosis dormancy
program. J Exp Med. 2003;198(5):705–13.
72. Sarathy J, Dartois V, Dick T, Gengenbacher M. Reduced drug uptake in phenotypically
resistant nutrient-starved nonreplicating Mycobacterium tuberculosis. Antimicrob Agents
Chemother. 2013;57(4):1648–53.
73. Adams KN, Takaki K, Connolly LE, Wiedenhoft H, Winglee K, Humbert O, et al. Drug toler-
ance in replicating mycobacteria mediated by a macrophage-induced efflux mechanism. Cell.
2011;145(1):39–53.
74. Giddey AD, de Kock E, Nakedi KC, Garnett S, Nel AJ, Soares NC, et al. A temporal pro-
teome dynamics study reveals the molecular basis of induced phenotypic resistance in
Mycobacterium smegmatis at sub-lethal rifampicin concentrations. Sci Rep. 2017;7:43858.
75. Kohanski MA, Dwyer DJ, Hayete B, Lawrence CA, Collins JJ. A common mechanism of
cellular death induced by bactericidal antibiotics. Cell. 2007;130(5):797–810.
428 C. Nathan

76. Nandakumar M, Nathan C, Rhee KY. Isocitrate lyase mediates broad antibiotic tolerance in
Mycobacterium tuberculosis. Nat Commun. 2014;5:4306.
77. Harms A, Maisonneuve E, Gerdes K. Mechanisms of bacterial persistence during stress and
antibiotic exposure. Science. 2016;354(6318)
78. Jankevicius G, Ariza A, Ahel M, Ahel I. The toxin-antitoxin system DarTG catalyzes revers-
ible ADP-ribosylation of DNA. Mol Cell. 2016;64(6):1109–16.
79. Beckert B, Abdelshahid M, Schafer H, Steinchen W, Arenz S, Berninghausen O, et al.
Structure of the Bacillus subtilis hibernating 100S ribosome reveals the basis for 70S dimer-
ization. EMBO J. 2017;
80. Rath P, Huang C, Wang T, Wang T, Li H, Prados-Rosales R, et al. Genetic regulation of
vesiculogenesis and immunomodulation in Mycobacterium tuberculosis. Proc Natl Acad Sci
U S A. 2013;110(49):E4790–7.
81. Bryk R, Gold B, Venugopal A, Singh J, Samy R, Pupek K, et al. Selective killing of nonrep-
licating mycobacteria. Cell Host Microbe. 2008;3(3):137–45.
82. Bryk R, Griffin P, Nathan C. Peroxynitrite reductase activity of bacterial peroxiredoxins.
Nature. 2000;407(6801):211–5.
83. Bryk R, Lima CD, Erdjument-Bromage H, Tempst P, Nathan C. Metabolic enzymes
of mycobacteria linked to antioxidant defense by a thioredoxin-like protein. Science.
2002;295(5557):1073–7.
84. Maksymiuk C, Balakrishnan A, Bryk R, Rhee KY, Nathan CF. E1 of alpha-ketoglutarate
dehydrogenase defends Mycobacterium tuberculosis against glutamate anaplerosis and
nitroxidative stress. Proc Natl Acad Sci U S A. 2015;112(43):E5834–43.
85. Vandal OH, Pierini LM, Schnappinger D, Nathan CF, Ehrt S. A membrane protein pre-
serves intrabacterial pH in intraphagosomal Mycobacterium tuberculosis. Nat Med.
2008;14(8):849–54.
86. Darwin KH, Nathan CF. Role for nucleotide excision repair in virulence of Mycobacterium
tuberculosis. Infect Immun. 2005;73(8):4581–7.
87. Darwin KH, Ehrt S, Gutierrez-Ramos JC, Weich N, Nathan CF. The proteasome
of Mycobacterium tuberculosis is required for resistance to nitric oxide. Science.
2003;302(5652):1963–6.
88. Lin G, Li D, de Carvalho LP, Deng H, Tao H, Vogt G, et al. Inhibitors selective for mycobac-
terial versus human proteasomes. Nature. 2009;461(7264):621–6.
89. Vaubourgeix J, Lin G, Dhar N, Chenouard N, Jiang X, Botella H, et al. Stressed mycobacteria
use the chaperone ClpB to sequester irreversibly oxidized proteins asymmetrically within and
between cells. Cell Host Microbe. 2015;17(2):178–90.
90. Bryk R, Arango N, Maksymiuk C, Balakrishnan A, Wu YT, Wong CH, et al. Lipoamide
channel-binding sulfonamides selectively inhibit mycobacterial lipoamide dehydrogenase.
Biochemistry. 2013;52(51):9375–84.
91. Bryk R, Arango N, Venugopal A, Warren JD, Park YH, Patel MS, et al.
Triazaspirodimethoxybenzoyls as selective inhibitors of mycobacterial lipoamide dehydro-
genase. Biochemistry. 2010;49(8):1616–27.
92. Lin G, Chidawanyika T, Tsu C, Warrier T, Vaubourgeix J, Blackburn C, et al. N,C-capped
dipeptides with selectivity for mycobacterial proteasome over human proteasomes: role of
S3 and S1 binding pockets. J Am Chem Soc. 2013;135(27):9968–71.
93. Gold B, Roberts J, Ling Y, Quezada LL, Glasheen J, Ballinger E, et al. Rapid, semiquantita-
tive assay to discriminate among compounds with activity against replicating or nonreplicat-
ing Mycobacterium tuberculosis. Antimicrob Agents Chemother. 2015;59(10):6521–38.
94. Russo F, Gising J, Akerbladh L, Roos AK, Naworyta A, Mowbray SL, et al. Optimization and
evaluation of 5-Styryl-Oxathiazol-2-one Mycobacterium tuberculosis proteasome inhibitors
as Potential Antitubercular Agents. ChemistryOpen. 2015;4(3):342–62.
95. Warrier T, Martinez-Hoyos M, Marin-Amieva M, Colmenarejo G, Porras-De Francisco E,
Alvarez-Pedraglio AI, et al. Identification of novel anti-mycobacterial compounds by screen-
ing a pharmaceutical small-molecule library against nonreplicating Mycobacterium tubercu-
losis. ACS Infect Dis. 2015;1(12):580–5.
13 Phenotypic Tolerance and Bacterial Persistence 429

96. Gold B, Smith R, Nguyen Q, Roberts J, Ling Y, Lopez Quezada L, et al. Novel cephalo-
sporins selectively active on nonreplicating Mycobacterium tuberculosis. J Med Chem.
2016;59(13):6027–44.
97. Darby CM, Nathan CF. Killing of non-replicating Mycobacterium tuberculosis by
8-­hydroxyquinoline. J Antimicrob Chemother. 2010;65(7):1424–7.
98. Shah S, Dalecki AG, Malalasekera AP, Crawford CL, Michalek SM, Kutsch O, et al.
8-Hydroxyquinolines are boosting agents of copper-related toxicity in Mycobacterium tuber-
culosis. Antimicrob Agents Chemother. 2016;60(10):5765–76.
99. Singh R, Manjunatha U, Boshoff HI, Ha YH, Niyomrattanakit P, Ledwidge R, et al. PA-824
kills nonreplicating Mycobacterium tuberculosis by intracellular NO release. Science.
2008;322(5906):1392–5.
100. Nathan C. An antibiotic mimics immunity. Science. 2008;322(5906):1337–8.
Chapter 14
Staphylococcus aureus Adaptation During
Infection

Bo Shopsin and Richard Copin

14.1 Introduction

Bacterial survival critically depends on the ability to swiftly respond to environmen-

tal change. To efficiently monitor the surrounding environment, microbial genomes
encode numerous, highly diverse proteins, such as two-component signaling sys-
tems, that sense particular extracellular stimuli. In response to diverse cues, includ-
ing nutrients, light, gases, and host and synthetic antimicrobial stress, systems
transmit signals to the intracellular environment and thereby elicit a response.
Molecular dissection of these signaling networks has increased our understand-
ing of communication processes and provides a platform for therapeutic interven-
tion. When organisms are forced into environments far beyond their normal situation
and when their mechanisms for responding to the new environment are over-
whelmed, an alternative path to adaptive evolution may occur through selection of
heritable genetic changes that “capture” the phenotype produced by a stimulus. The
emergence of antibiotic resistance in pathogenic microorganisms provides an excel-
lent example of such evolution, one that has profound consequences for human
health.
Antimicrobial resistance is based on selection for organisms that have an
enhanced ability to grow in the presence of a host or synthetic antimicrobial. The
evolution of drug resistance can be attributed to multiple factors that include: (1) an
increased frequency of intrinsically resistant variants, (2) the acquisition of mobile

B. Shopsin (*)
Departments of Medicine and Microbiology, New York University School of Medicine,
New York, NY, USA
e-mail: [email protected]
R. Copin
Department of Medicine, New York University School of Medicine, New York, NY, USA

© Springer International Publishing AG, part of Springer Nature 2018 431

I. W. Fong et al. (eds.), Antimicrobial Resistance in the 21st Century, Emerging Infectious
Diseases of the 21st Century, https://doi.org/10.1007/978-3-319-78538-7_14
432 B. Shopsin and R. Copin

resistance determinants, and (3) de novo accumulation of resistance mutations. The

evolutionary dynamics depend on the biology and population size of the microbe in
question, the drug, and the opportunity for genetic exchange of resistance determi-
nants. The observation that the emergence of drug resistance outpaces the develop-
ment of new antimicrobial agents underscores the crucial importance of
understanding the evolutionary mechanisms that lead to the development of
resistance.
Antimicrobial resistance has traditionally been approached from a mechanistic
perspective focused on identifying the cellular determinants that prevent a drug
from entering a cell, remove a drug from the cell, inactivate a drug, or prevent a drug
from inhibiting the normal activity of its target. Selection may occur at key regula-
tory loci, or it can be focused downstream at the effectors of the phenotype. None of
these mechanisms acts alone. Moreover, the phenotypic effects of mutations that
confer resistance depend on the genetic background of the strain and changes in the
genome that occur during clinical infection. This complexity is illustrated by the
observation that the development of resistance is often accompanied by a fitness
cost or deleterious effect on pathogen growth in the absence of the drug. Fitness
costs are often mitigated by the accumulation of compensatory mutations that
enhance the fitness of the resistant genotype in the absence of the drug. Fitness in
the presence of a drug is a complex trait affected by multiple loci, the bacterial spe-
cies involved, the ecological niche, and the host.
Using Staphylococcus aureus as an example pathogen, the present chapter
focuses on the mechanisms that potentiate the evolution of drug resistance, with an
emphasis on the central role of mutations in “off-target” genes having pleiotropic
effects. Substantial support exists for the role of metabolic changes that fuel the
accumulation of reactive oxygen species (ROS; superoxide, peroxide, and hydroxyl
radicals) in the live-or-die decision made by bacteria [1–4]. Thus, emphasis is
placed on how these principles apply to the lethal (bactericidal) cellular responses
to a variety of antimicrobials during bacterial growth. In addition, mechanisms by
which alterations in cellular states can influence the emergence of drug resistance,
including the effects of tolerant cells, will be highlighted. We also discuss how bac-
terial adaptations that are potentially beneficial within hosts and hospitals can be
used to track the evolution of hospital clones using whole-genome sequencing,
underscoring the need for rapid containment. Finally, the possibility of harnessing
evolution for therapeutic benefits through cellular perturbations will be explored.

14.2 The Accesory Gene Regulator (agr) Paradox

14.2.1 agr and Clinical S. aureus Infections

S. aureus is responsible for a large variety of diseases in both community and hospital
settings [5]. Despite advances in care, S. aureus infections remain associated with
considerable morbidity and mortality. In addition, treatment of methicillin-resistant
14 Staphylococcus aureus Adaptation During Infection 433

S. aureus strains (MRSA), an increasing cause of healthcare-­associated infections, is

complicated by the emergence of intermediate and fully vancomycin-resistant strains
[6–8]. MRSA surgical site infections are particularly devastating when hardware is
implanted in the patient (e.g., prosthetic joint, pacemaker, and vascular graft infec-
tions). As human populations age, more invasive procedures are being performed.
Consequently, the incidence of implant infection by MRSA is certain to increase.
While the outcome of an S. aureus encounter is usually asymptomatic coloniza-
tion, the propensity of S. aureus strains to produce invasive infection defines a
capacity to resist host innate immune clearance mechanisms. Infection is likely
transformative for the bacterium, since it must overcome host, and possibly syn-
thetic antimicrobials, to live within as well as upon the host. Thus, hope for develop-
ing new ways to control S. aureus rests in part on understanding how the bacterium
adapts to the new, in-host environment.
In a general sense, mutations and natural selection are expected to shape the
evolutionary dynamics of S. aureus within an individual host; however, the type,
frequency, and interaction of these events are largely unknown. Work on adaptation
to the host has focused on gene regulation; in contrast, population genetics has
addressed the genetic basis of evolution, with little overlap between the two disci-
plines. Recent work has combined molecular typing of field isolates with in vitro
experiments that examine how S. aureus evolves within hosts. These studies have
identified within-host variation in the agr locus, a quorum-sensing, global regulator
of virulence in S. aureus [9] (Fig. 14.1).
agr mutants are attenuated for virulence in animal models of infection [14], and
the majority of clinical isolates have a functional agr locus. However, agr-defective
strains are a common clinical occurrence, particularly in persistent infections in
which biofilms are thought to play a role, such as device-related infection and endo-
carditis [15–20]. In vivo selection for agr-defective strains was suggested by studies
of sequential isolates recovered from the blood of patients during antimicrobial
treatment [18, 21, 22], as well as in animal infection models [23]. The existing data
led to the paradoxical conclusion that survival of S. aureus in the bloodstream may
be enhanced by the inability of S. aureus to produce numerous virulence factors,
including cytotoxic leukocidins. Moreover, among patients with MRSA bacteremia,
the development of an agr-defective phenotype serves as a predictor of persistence
of the organism and of a higher incidence of infectious endocarditis [24] and mor-
tality [25]. Thus, the clinical consequences of agr activity are not obvious – depend-
ing on the patient – they could even make efforts to use agr and virulence as targets
for new antimicrobials ill advised [26].

14.2.2 Epidemiology of agr Dysfunction

The temporal and spatial variation in environmental conditions that opportunistic

microbes are exposed to within an individual host and during transmission between
hosts likely promote adaptation to a lifestyle that accomodates rapidly changing
434 B. Shopsin and R. Copin

AIP

B
C

D
A
P2 P3
A C D B

Target gene ? -hemolysin

RNAIII

Fig. 14.1 The agr quorum-sensing system. (A) The agr locus consists of two divergent transcrip-
tion units driven by promoters P2 and P3. The P2 operon encodes the signaling module, which
contains four genes – agrB, D, C, and A – each of which is required for transcriptional activation of
the agr regulon (reviewed in [9]). AgrC is the receptor-histidine kinase, and AgrA is the response
regulator. AgrD is the autoinducing, secreted peptide that is derived from a propeptide processed by
AgrB. The P3 transcript is a regulatory RNA (RNAIII) that also encodes the structural gene for
hemolysin. Regulation of target genes by agr occurs through two pathways: (1) an RNAIII-­
dependent regulation of virulence genes and (2) an RNAIII-independent, AgrA-mediated regulation
of metabolic genes and small cytolytic toxins known as phenol-soluble modulins (modulins). The
regulatory connection between these processes links virulence to metabolism. Agr has a dual, time-
dependent regulatory role (in vitro) that is characterized by (1) increased post-exponential produc-
tion of toxins and exoenzymes (e.g., α-hemolysin) that facilitate dissemination of bacteria via tissue
invasion; (2) decreased production of cell surface proteins that facilitate adherence and attachment
(e.g., fibronectin-binding proteins); and (3) decreased production of factors that promote the eva-
sion of host defense (e.g., protein A). Thus, the agr locus coordinates a switch from an adherent
state to an invasive state dependent on bacterial population density. This important duality has been
exploited by the use of agr quorum-sensing inhibitors for the prevention and treatment of experi-
mental S. aureus infections, including catheter and vascular prosthetic graft infection [10–13]

environments. Thus, a better understanding of how S. aureus redirects and fine-­

tunes its gene expression in response to the challenges of colonization, transmis-
sion, and infection is central to understanding the causal pathway between
commensalism and serious, complicated disease.
The observation that a significant fraction (~20% overall and ~70% in patients
with persistent bacteremia [18, 24]) of clinical isolates of S. aureus from infections
have genotypic agr defects provides a way to delineate the epidemiology-host-­
microbe relationship of this system, and therefore virulence, in disease. Nasal car-
riage is an important prerequisite for S. aureus infection, indicating the importance
14 Staphylococcus aureus Adaptation During Infection 435

of examining the role of agr in colonization. Screening assays to detect agr func-
tionality among isolates from healthy subjects indicate that although agr dysfunc-
tion is not an absolute barrier to colonization and transmission, carriage of
agr-defective strains is strongly associated with hospitalization rather than with
healthy patients [27]. Collectively, these observations suggested that: (1) agr-­
defective mutants are fit for transmission (they are not a “dead-end” state), and (2)
the hospital environment is a reservoir of attenuated, agr-defective variants.
Presumably, disruption of protective barrier functions by disease and clinical inter-
vention (e.g., intravenous catheter use) permits S. aureus lacking full virulence to
cause infection. Analysis of paired S. aureus clones from blood infection and nasal
carriage sites in individual hospitalized patients presenting with bacteremia indicate
that recovery of an agr-defective mutant from blood is usually predicted by the agr
status of carriage isolates [28]. Thus, fieldwork supports the idea that the transition
from commensalism to opportunism in S. aureus does not require full virulence in
hospitalized patients.
The strong association of agr dysfunction with the hospital environment and
infection suggests an unappreciated role for agr: colonization by S. aureus is
responsible for maintaining agr function. Indeed, although fully agr-defective
mutant isolates colonize and transmit, they do not persist indefinitely in natural
populations of hospital-associated MRSA [29]. This suggests that, in the case of agr
mutation, attenuation of virulence is the product of short-sighted evolution within
hosts – although attenuation of agr-mediated virulence may help S. aureus adapt to
host tissues in the short term, it appears to put S. aureus at a disadvantage in the long
term.
The combination of ubiquity and relatively short lifespan suggests that the
occurrence of agr-defective mutants results from frequent within-host selection in
situations such as persistent bacteremia [18, 21, 22, 28]. However, an experimen-
tal system demonstrating transmission following invasive bacteremia was lacking,
and thus implications of within-host adaptation for between-host transmission –
and therefore for hospital epidemiology – were unknown. While a disease-pro-
moting agr mutation that occurs during the course of bacteremia could confer a
transient advantage to the bacterium, such an adaption would be a dead end for the
bacterium in the absence of transmission. Recently, S. aureus was found to dis-
seminate to the gastrointestinal tract of mice via the gall bladder following intra-
venous injection, and the bacterium readily transmits to cohoused naive mice
[30]. These findings established an animal model to investigate gastrointestinal
dissemination of S. aureus and the role of adaptive mutations in genes such as agr.
The work suggests that selective processes taking place over the course of blood
infection can go beyond a single host. Both intestinal dissemination and transmis-
sion were linked to the production of virulence factors based on gene deletion
studies of two-­component virulence regulatory systems, including agr. Thus, the
animal data are consistent with data from hospital isolates that indicate that agr
inactivation can attenuate colonization-transmission but is selected during
bacteremia.
436 B. Shopsin and R. Copin

14.3  leiotropic Point Mutations in agr Illustrate a General

P
Mechanism of Adaptive Evolution During Infection

Adaptive evolution to a new niche, such as a novel host or host environment,

involves either fluctuating or completely novel conditions, and generally requires a
rapid shift in the expression of genes [31, 32]. Mutations in transcriptional regula-
tors produce a rapid pleiotropic, phenotypic effect on the expression of multiple
genes, and the mutations correlate with adaptive radiation [33]. Indeed, experimen-
tal [34–36] and observational [37–40] work suggests that global regulators consti-
tute a “one-step” mechanism of adaptation that drive adaptive leaps made by
microbes.
Global regulators and two-component signaling systems are highly abundant in
the S. aureus genome ([41], see also Chap. 15). They form a complex regulatory
network that modulates phenotypic plasticity and the expression of virulence genes,
cell division, and stress responses in response to environmental change [42].
Quorum sensing can be considered to lie at the top of the transcriptional regulatory
network hierarchy, not just in S. aureus but also in other pathogens. Consequently,
mutations that affect quorum-sensing loci during adaptation to novel environments
are likely to be a general phenomenon. Indeed, the lasR quorum-sensing system,
which is involved in the repression of biosynthesis of virulence factors and biofilm
in Pseudomonas aeruginosa, is a hot spot for mutations in isolates from chronically
infected cystic fibrosis patients [40, 43–46]. Accordingly, the present chapter
focuses on the role of quorum-sensing mutations as a prototype adaptive mutation.
Moreover, agr dysfunction and virulence attenuation are similar to the phenotype of
strains that have mutations or dysfunction in other regulators during infection. For
example, S. aureus mutationally adapts the global regulator rsp and virulence factor
expression in the course of infection [38]. Thus, multiple genetic mechanisms, as
well as the genetic background of the strain, control the induction of host-adatpted
states, indicating that the interplay between factors and the associated selective loss
of any one regulator are complex.

14.4 Evolution of agr-Defective Mutants

14.4.1 Host-Pathogen Interactions

Several explanations can potentially account for the selection of agr-defective

strains and their association with persistent infections. For example, endocarditis
vegetations may be regarded as biofilms in which the organisms are protected from
attack by phagocytes (and antibiotics). It has been demonstrated that agr-defective
mutants are enriched in biofilms [20, 47]. Organisms at the surface of a biofilm
express agr and those in the underlying layers have agr repressed [48]. Thus, agr-
defective strains could provide adhesins to stabilize the vegetations, and the
14 Staphylococcus aureus Adaptation During Infection 437

agr-positive strains could adhere to the agr-defective variants while producing their
toxic exoproteins. Additionally, inactivation of agr upregulates fibronectin-binding
proteins, which play an important role in the ability of S. aureus to colonize, persist
within, and damage cardiovascular tissue [49]. agr inactivation also increases resis-
tance to endogenous thrombin-induced microbicidal proteins [50], key mediators of
host defense that are secreted by platelets at sites of cardiovascular damage and
infection.
It is also possible that agr mutation promotes S. aureus survival inside host
cells, as opposed to within-host tissues. Although considered an extracellular
pathogen, S. aureus clearly thrives inside host phagocytic, epithelial, and endothe-
lial cells ([51–56]; reviewed in [57]). The importance of this intracellular lifestyle
is highlighted by the recent finding that ablation of intracellular S. aureus improves
outcome from experimental infection [54, 55]. While the large majority of work on
MRSA-­phagocyte interactions, including work from our laboratories, has been
performed with neutrophils, recent findings indicate that disseminated S. aureus
infection is tied to survival inside macrophages [52–55]. For some cell types, agr-
defective mutants exhibit prolonged intracellular residence due to attenuated cyto-
toxicity and delay in initiation of host cell death [58, 59]. But the agr-defective
phenotype is not noted in isolates from primary skin and soft tissue infection (e.g.,
[60]), suggesting that such attenuated toxicity and intracellular survival may be
particularly important in infections in which persistence is an issue. Persistence is
a particular problem with S. aureus endocarditis, while it is not such an issue with
acute infection of skin and soft tissue. Thus, a better understanding of the interac-
tion between-host cells, agr-mutant, and wild-type strains will generate the knowl-
edge needed to confront the growing problem of complicated disease and poor
outcomes.
Other investigators showed that the use of antibiotics, such as fluoroquinolones
or beta-lactams, is a risk factor for loss of agr functionality in vitro and during treat-
ment of infection in the hospital [61, 62]. Furthermore, agr-defective strains are
associated with the development of vancomycin tolerance in vitro and during treat-
ment of patients with bacteremia in vivo, perhaps owing to defects in autolysis and
consequent changes in cell wall structure that mitigate fitness costs associated with
the evolution of vancomycin resistance (reviewed in [17]). Indeed, all known van-
comycin intermediate-resistant S. aureus (VISA) strains are agr-defective.
Moreover, recent work demonstrates that vraR, a member of the two-component
vraRS regulatory system that is upregulated in vancomycin-resistant strains, sup-
presses transcription of agr [63]. Thus, vancomycin resistance appears to be an
example of how a fitness trait that is initially dependent on attenuation of agr can
evolve transcriptional independence.
Enhanced fitness when agr function is compromised can enrich underlying poly-
morphisms in the locus such that a threshold is passed and the phenotype is expressed
in all members of the population. The effect of antimicrobials on selection for agr-­
defective strains is discussed further in sections below in a different context – dimin-
ished antimicrobial lethality (tolerance) rather than inhibition of growth (resistance).
We conclude this section by noting that agr dysfunction offers potential advantages
438 B. Shopsin and R. Copin

to S. aureus not just during infection but also more generally by promoting protec-
tion against antimicrobials. Indeed, loss of agr expression has been described as a
potential “win-win” situation for a nosocomial pathogen [64].

14.4.2  itness and Protection from Antimicrobials in agr-­

F
Defective Mutants

The widespread increase in S. aureus antibiotic resistance has dramatically nar-

rowed treatment choices, especially with the appearance of resistance to key antimi-
crobials such as vancomycin and daptomycin. Resistance often emerges in vivo
during persistent infection, and a number of studies have investigated the genomic
basis for this phenomenon. Many mutations, such as those involved in target modi-
fication, are thought to lead directly to antimicrobial resistance. Other mutations
accumulate under combined antibiotic and host selective pressures, leading not only
to antimicrobial resistance but also to altered host-pathogen interactions that favor
persistent infection. For example, host thrombin-induced platelet microbicidal pro-
teins (tPMP) are one of the first-line innate defense mechanisms against S. aureus
infection, and a link has been demonstrated among agr mutations associated with
reduced vancomycin and daptomycin susceptibility, persistence, and reduced tPMP
susceptibility [24, 50, 65].
As discussed above, inactivation of agr also correlates with antibiotic use in
patients, suggesting that agr functionality is subject to a tradeoff – agr activation
promotes survival in host niches favoring acute virulence but represents a liability to
the bacterium during growth stress, especially antimicrobial treatment. The frequent
occurrence of agr mutants in serial passage of laboratory cultures in the absence of
antimicrobials supports the hypothesis that agr activity is metabolically costly [66],
as does the observation that the locus is itself highly expressed and that agr activates
the expression of many more genes than it inhibits [67]. Futhermore, previous
reports indicate a growth advantage for Δagr mutants in the presence of subinhibi-
tory concentrations of several antibiotics [42]. Fitness gains for Δagr mutants were
associated with inactivation of RNAIII, indicating that the growth advantage is agrA
independent. agr mutation may also mitigate fitness defects of resistance mutations.
For example, deformylase inhibitor-resistant S. aureus strains partly regain fitness
through mutation of agr while still retaining high-level resistance [68].
Superoxide, a metabolic product, may play a central role in the live-or-die deci-
sion made by bacteria when challenged with lethal antimicrobials, such as cipro-
floxacin, a gyrase-mediated DNA-damaging agent [1–4]. Activation of antioxidant/
oxidative stress-protective responses in bacteria would therefore be expected to pro-
mote antimicrobial tolerance (loss of lethal activity but retention of bacteriostatic
activity). From a clinical point of view, tolerance presents a major challenge: in con-
trast to the specificity of resistance, tolerance confers a survival advantage against a
broad spectrum of drugs and stresses. Additionally, it is likely that tolerance provides
14 Staphylococcus aureus Adaptation During Infection 439

a reservoir for relapse and the evolution to antibiotic resistance. Thus, understanding
tolerance is critical for addressing the decreasing efficacy of antibiotics. We note that,
although definitions have been debated, tolerance is related to but distinct from
another important contributor to pathogen survival – the phenomenon of persistence.
Persisters are considered to be slow growing, metabolically dormant cells that exhibit
tolerance and are less likely than growing cells to exhibit ROS-­mediated killing by
antimicrobials (for additional discussion of tolerance, see Chap. 13).
One oxidative stress-protective mechanism that might promote antimicrobial tol-
erance in S. aureus involves mutation of agr, which has a built-in oxidation-sensing
mechanism through an intramolecular disulfide switch possessed by the DNA-­
binding domain of the response regulator AgrA [69]. Oxidation of AgrA decreases
DNA-binding activity, which results in derepression of the bsaA gene, which
encodes the antioxidant glutathione peroxidase. As a result, agr-defective mutants
are less susceptible to oxidative stress.
The frequent occurrence of in vivo-selected agr-defective mutants during persis-
tent infection highlights a possible link between oxidative stress and antibiotic toler-
ance in this organism. The mechanism underlying agr dysfunction among strains
derived from serial passage in vitro and from clinical isolates is almost always traced
to inactivating mutations in agrC and agrA, the sensor component and response
regulator, respectively, of the agr system (Fig. 14.2). The intuition explaining this
observation is that selection for agr-defective strains occurs in mixtures containing
agr-positive parental strains. Accordingly, inactivation of agrD or agrB does not
silence agr owing to the production of autoinducing peptide in trans by the agr-
positive strain. However, this scenario does not explain why RNAIII, the effector of
the agr response, is not targeted by selection for loss of agr function. We hypothesize
that an agrA-dependent, bsaA-mediated antioxidant phenotype provides protection
against antibiotic-dependent oxidative damage, thereby resolving the dilemma.

9765 9771 9765 9764 P2 P3

A C D B RNAIII

9906 9772 9764

Fig. 14.2 Localization of inactivating mutations in agr. The mechanism underlying agr dysfunc-
tion can usually be traced to mutations in agr that inactivate the locus [18, 22, 27, 29, 66, 70, 71].
Representative mutations identified by DNA sequencing of the agr operon in different clinical
isolates that were negative for δ-hemolysin. (Adapted from [18]). The numbers on the figure refer
to the location of agr mutations for different isolates. The isolates were derived from patients with
various clinical infections (the strain number is an arbitrary designation). Some strains had more
than one mutation in agr, but complementation with the relevant gene on a plasmid showed that
only one mutation per strain resulted in agr inactivation. Notably, agr defects in S. aureus and
Staphylococcus epidermidis usually result from quorum-sensing deficiency (due to a mutation in
agrA or agrC) rather than from quorum-signaling deficiency (due to a mutation in agrB or agrD).
Presumably, selection for agr-defective strains occurs in mixtures with agr-positive parental
strains. Thus, inactivation of agrD or agrB cannot silence agr owing to the production of AIP in
trans by the agr-positive strain
440 B. Shopsin and R. Copin

To test our agrA-bsaA hypothesis, we used a range of both drug concentration

and treatment time to probe effects of agr status on the response of S. aureus to
lethal stress [72]. We note that, to study agr-stressor effects, it is important to distin-
guish phenotypes related to growth from those specific to survival. For example,
treatment with an antimicrobial leads to damage that is specific to the test agent.
This primary damage halts growth, which is measured as the minimal inhibitory
concentration (MIC). MIC reflects drug uptake, efflux, and target affinity; high MIC
values are associated with antimicrobial resistance. Some forms of primary damage
also kill cells, with much of the lethal process arising from a self-destructive bacte-
rial response to the primary damage (reviewed in [2, 3]; also Chap. 20). To focus
experimental measurements on the lethal response, lethal drug concentrations were
normalized to MIC. It is also important to recognize that lethal stress may be tran-
sient. For example, with S. aureus ROS can accelerate killing without increasing the
extent of killing [73]. Consequently, overnight killing assays, such as those com-
monly used to measure minimal bactericidal concentration (MBC), may be not pro-
vide much information [73].
Using highly lethal antimicrobials as probes for studying bacterial responses to
lethal stress, we found that wild-type agr stimulates the lethal action of several
stressors, including gentamicin and ciprofloxacin; thus, defective mutants will tend
to persist under stressful conditions rather than being killed by stressors that may
include synthetic antimicrobials and host defenses such as neutrophil-generated
ROS. Disruption of the RNAIII-dependent pathway had no effect on the stress
responses to lethal stress. In contrast, disruption of the agrA-dependent pathway
had effects, but they varied from one stressor to another. For example, with
ciprofloxacin-­mediated killing, agr facilitated the accumulation of toxic ROS,
which is known to be involved in quinolone-mediated killing of Escherichia coli
and S. aureus. agr action appears to be exerted by downregulating bsaA; thus, an
agr defect allows expression of a protective protein.
Within our sample of stressors, daptomycin was unusual in exhibiting greater
lethality with the agr-deficient mutant. Test conditions are important, as indicated
by consideration of previous work in which the opposite result was obtained with
nongrowing S. aureus in deep stationary phase, long after induction of agr and
expression of agr transcripts [74]. Daptomycin, a calcium-dependent molecule that
acts as a cationic antimicrobial peptide, releases membrane phospholipids that bind
to and inactivate the antibiotic. Although both wild-type and Δagr strains released
phospholipid in response to daptomycin, agrA-triggered secretion of phenol-­
soluble modulin (PSM) cytotoxins prevents antibiotic inactivation by wild-type
cells. In previous experiments, killing assays were performed using overnight cul-
tures, long after induction of agr and expression of its transcripts. Thus, the results
reflect agr-­mediated exoprotein secretion rather than a cellular response to stress-
mediated killing. Our experiments were performed in late exponential phase when
PSM levels may be lower and less protective [75]. The complex relationship
between daptomycin lethality, agr status, and bacterial physiological state illus-
trates the importance of understanding agr biology before applying novel therapies
that target agr [26].
14 Staphylococcus aureus Adaptation During Infection 441

Collectively, the data support the hypothesis that inactivation of agrA can result
in degradation of both antimicrobials themselves and the lethal response to
antimicrobial-­mediated stress. Given the data indicating that fitness gains are asso-
ciated with inactivation of RNAIII (mentioned above, [42]), we conclude that two
distinct subsets of agr antimicrobial fitness exist: an RNAIII-independent one that
impacts antimicrobial lethality and an RNAIII-dependent form that controls
antimicrobial-­associated fitness for growth (Fig. 14.3).

Fig. 14.3 Overview of agr mutation and its consequences in complex host environments. Effect
of agr deficiency on sublethal and lethal action. (A) Sublethal stress. By switching to the inactive
form and modulating expression of appropriate factors, S. aureus cells gain enhanced replicative
fitness (RNAIII pathway). In this model, RNAIII deficiency enhances energy resources. Decreased
protein synthesis and ATP could be involved in this coupling; however, as with many factors in the
agrA pathway, the physiological relevance, as well as the mechanism by which these factors act, is
poorly understood. (B) Lethal stress. agr deficiency modulates survival against lethal stress (agrA
pathway). agrA-mediated effects are stress dependent, for example, derepression of the antioxidant
bsaA results in protection against ciprofloxacin, whereas upregulation of unknown factors control
survival in trans against gentamicin (not shown). For both RNAIII and agrA pathways, enhanced
survivability is achieved at the cost of virulence, which may or may not be compensated for by the
presence of coinfecting agr-positive strains
442 B. Shopsin and R. Copin

14.4.3 Social Cheating

Given the metabolic burden associated with agr function, the question arises as
to whether it is advantageous for S. aureus populations to consist purely of
signaling-­proficient cells or whether there might be situations in which “cheat-
ers” would be favored. Social cheaters reap the benefit of public goods while
contributing less than average to the cost [76, 77]. In S. aureus, many agr-regu-
lated products are released into the extracellular environment and benefit not
only the producing cell but also its neighbors. Mutants that do not respond to
agr autoinducing peptide signals do not incur the cost of producing these “pub-
lic goods,” but they may gain the benefit of production of goods shared by
neighbors. Cheating theory therefore predicts that agr-defective cells should be
at a disadvantage to their wild-type counterparts when grown in monoculture. In
support of this hypothesis, it is well known that agr-­defective cells are rapidly
eliminated and are less virulent in animal models of acute infection [14, 23, 78,
79]. Indeed, it was unequivocally demonstrated that blocking of agr attenuates
staphylococcal virulence and that the administration of an agr-­positive superna-
tant along with agr-defective organisms protects the bacteria in an abscess
model of infection [14]. This suggested that the spread of agr-defective strains
within populations in vivo is due in part to the exploitation of shared products
produced by their wild-type neighbors. Thus, social cheating may be relevant
during infection owing to the differential impact of host defenses on bacterial
survival.

14.4.4 agr and Mutability

Evolution by natural selection involves two main steps: the generation of heritable
variations (e.g., mutations) and the differential proliferation of the variants in the
environment. Hypermutability may function to create a diverse bacterial population,
increasing the likelihood of environmentally adapted variants. Thus, enhanced
mutability may provide the substrate for selection of attenuated agr-mediated viru-
lence in S. aureus infection. However, recent work by Plata et al. suggests that heri-
table elevations of mutation frequency are not likely the cause of agr diversification:
agr-defective mutants, and their parent strains showed similar mutation frequencies
in the range of what is commonly found in the species [80]. Nonetheless, the authors
reported that generation of heterogeneous resistance to oxacillin is enhanced by agr
mutation and antimicrobial-related stress [80], giving rise to the related ideas that
(1) agr suppresses genome plasticity and (2) agr dysfunction can result in bursts of
mutations. In this scenario, agr dysfunction potentially serves as a “driver” muta-
tion that promotes accumulation of additional genetic alterations and rapid evolu-
tion in the complex environmental milieu of invasive infection. agr inactivation and
genetic instability in turn may result in subclones that display emergent properties,
14 Staphylococcus aureus Adaptation During Infection 443

including antimicrobial resistance, that pose challenges for therapeutic intervention.

agr-mediated effects may be superimposed on those of antimicrobials themseleves,
which may induce mutagenic “SOS” responses that contribute to the emergence of
resistance (e.g., see [81]).

14.4.5 Implications of agr-Mediated Antimicrobial Protection

The observations described above provide an entry point for additional screening to
identify other bacterial regulators having activities that can be self-protective,
depending on the type and level of lethal stress. Elucidating the basis of such effects
can be clinically significant when they inform efforts to personalize management of
antimicrobials through pathogen strain-specific characteristics. For example, the
use of anti-agr agents or therapeutic vaccines [26] may be counter-productive for
applications in which the absence of agr reduces lethal activity of an antimicrobial.
Likewise, identification of adaptations that erode the lethal activities of antimicrobi-
als might inform the development of novel strategies to selectively bolster antimi-
crobial effectiveness [82–84].

14.5 Role of Virulence in Acute Infection

14.5.1 Virulence and Outcome in Nosocomial Pneumonia

In persistent infections, such as complicated bacteremia, low virulence, and

enhanced adherence to prosthetic or host material might be expected to lie on
in the causal pathway leading to persistent infection and poor patient outcome.
In contrast, agr-mediated cytotoxic activity is integral to increased S. aureus
virulence in most models of acute infection, and agents that block agr and
quorum-sensing exhibit antiinfective properties [14]. Recently, virulence phe-
notypes were characterized among S. aureus isolates obtained at the time of
diagnosis from a large, prospective clinical trial that compared the efficacy of
linezolid and vancomycin for treatment of nosocomial pneumonia due to
MRSA [85]. The analyses took into account host-­related factors (virulence)
and organism-related factors (antimicrobial resistance). Virulence was mea-
sured by screening for functionality of agr. Since agr functionality alone does
not imply efficient expression of virulence, an additional, direct measure of
virulence factor production was sought. The leukocytotoxins, which are agr-
regulated pore-forming toxins (bicomponent leukocidins and alpha-­hemolysin),
and membrane-damaging cytolytic peptides, which are found in virtually all
staphylococcal isolates, are attractive candidates to be wide-ranging virulence
factors whose presence can effectively distinguish patient outcomes.
Accordingly, relative leukocytotoxic activity (in which individual leukotoxic
444 B. Shopsin and R. Copin

Distribution of CA- and HA-MRSA Cytotoxicity levels

(number of isolates) 100

Percentage of isolates (%)

80

60
118
40
263
20

0
High cytotoxicity Low cytotoxicity

HA-MRSA CA-MRSA HA-MRSA CA-MRSA

Fig. 14.4 Neutrophil cytotoxity values for CA-MRSA (orange) and HA-MRSA (blue) strains. In
our recent survey of more than 380 MRSA isolates from patients enrolled in a randomized con-
trolled trial of nosocomial pneumonia [85, 86], CA-MRSA lineages were found to be almost uni-
formly highly cytotoxic to human neutrophils (82% vs 11%; p < 0.0001). In contrast, HA-MRSA
lineages were weakly cytotoxic, often below levels expected to correlate with mortality in animals.
Furthermore, many strains (22/381 [6%] HA-MRSA) produced no detectable cytotoxic activity.
Left, overall, 69% (263/381) of isolates were assigned to CA- or HA-MRSA subsets based on
SCCmec type. More than 75% of CA-MRSA demonstrated ≥90% killing of neutrophils within
2 h. In contrast, ≥90% of HA-MRSA strains demonstrated cytotoxicity levels below 90%

activities are inseparable) were measured under normalized conditions [85], as

discussed below.
Cytotoxicity levels varied remarkably within each MRSA clonal complex and
even within groups of agr functional strains. Thus, agr functionality does not imply
full cytotoxic virulence. Community-acquired MRSA (CA-MRSA), defined by the
presence of the community-associated staphylococcal cassette chromosome mec
type IV genes that confer methicillin-resistance, was almost uniformly highly cyto-
toxic (Fig. 14.4). In contrast, most hospital-associated MRSA (HA-MRSA) strains
were weakly cytotoxic, often below levels expected to correlate with mortality in
animals. Indeed, several strains of HA-MRSA produced no detectable cytotoxic
activity. The low cytotoxicity phenotype was stable and heritable, indicating a
genetic perturbation.
HA-MRSA strains were the predominant isolate from patients with nosoco-
mial pneumonia. Moreover, the crude mortality rates for adequately treated
patients with MRSA due to strains of low cytotoxic activity were fourfold higher
than for patients infected with strains of high virulence, even after multivariate
logistic regression analysis with careful adjustment for bacterial, clinical, and
host factors. Thus, the low cytotoxic activity associated HA-MRSA strains is a
predictor of mortality in MRSA-mediated nosocomial pneumonia. Moreover,
isolates having low cytotoxicity, which were derived largely from healthcare-
associated clones, were recovered more often from patients who were older and
frailer. Collectively, these data suggested that the discrepancies in clinical out-
come and survival of low-virulence strains stem from confounding factors
14 Staphylococcus aureus Adaptation During Infection 445

related to differences in populations of patients infected with highly cytotoxic

and weakly cytotoxic MRSA. In this scenario, cytotoxic activity is a proxy for
a subtle factor that makes hospitalized patients more susceptible to infection
with hospital-associated bacteria of low virulence. That factor, referred to as
“host quality” [85], appears to be a multidimensional syndrome of decreased
reserve and resistance to stressors leading to increased vulnerability to adverse
outcomes.
The results provide a framework with which to further explore the relationship
between strain variation in MRSA and clinical outcomes. The issue is that, depend-
ing on the type of infection, patients who are infected with low-virulence MRSA are
inherently different from those infected with virulent MRSA and, in turn, that both
the natural history of disease and treatment effects between some drugs may be dif-
ferent in these patients. Future outcome studies should be designed so that the host
quality factor is assessed and evenly balanced between study groups.

14.5.2  ypothetical Model of Adaptation of Low-Virulence

H
MRSA in the Hospital Environment

Attempts to draw direct relationships between in vitro and animal model-based

virulence attributes, strain success, and patient outcome appear to have been too
simplistic. Substantial evidence supports the idea that adaptation to the host and
hospital environment, plus poor clinical outcome in invasive infection, are associ-
ated with low – not high – virulence [24, 25, 28, 38, 87–90]. Likewise, adaptation is
often accompanied by decreased susceptibility to killing by host factors and syn-
thetic antimicrobials. These observations enable formulation of a tentative model
for the events occurring during hospital adaptation as follows: after hospitalized
patients are infected with “wild-type” bacteria, host and antimicrobial factors pro-
mote diversification and the emergence of low-virulence variants. By evolving to a
host and potentially antimicrobial-tolerant form, the bacterium gains enhanced abil-
ity to persist in human tissues despite treatment. This persistence, however, requires
a downregulation of virulence (and possibly metabolic) functions. In this way, mod-
ulation of hospital adaptation is achieved at the cost of virulence, which may or may
not be compensated for by the presence of coinfecting wild-type strains. This dual-
ity of virulence and persistence may explain why such variants are preferentially
isolated from patients in the hospital compared with isolates from colonizing sites
in healthy community subjects. Supporting this hypothesis, agr mutations having
moderate or weak functional defects, such as the Gly55 amino acid change in agrC,
are associated with the successful and deadly hospital- and bacteremia-associated
clonal complex 30 lineage [88–90]. Apparently, the Goldilocks metaphor applies;
when high- and low-virulence states are balanced, infections are maintained, and
bacterial proliferation is not restricted, resulting in interhost spread and poor clini-
cal outcomes (Fig. 14.5).
446 B. Shopsin and R. Copin

Hospital adaptation
Hospital-associated Communiy-associated
S. aureus S. aureus

Antimicrobial
resistance/ Virulence
tolerance

Inflammation

low Moderate high

Low inflammation; Moderate inflammation; Excessive inflammation;

high tolerance moderate tolerance low tolerance

Poor response; Poor response; Good response;

low fitness high fitness low fitness

Response to treatment
Fig. 14.5 Variations in resistance and virulence, represented by the continuum symbol, influence
the balance between stimulation of proinflammatory host responses and antimicrobial-tolerant
states. High levels of virulence but low antimicrobial resistance results in immune and antimicro-
bial clearance among hospitalized patients with bacteremia. The same characteristics result in
exuberant bacterial growth and increased inflammation among hosts in the community [91–94]. In
the community, lack of prompt antimicrobial treatment leads to severe inflammation, tissue necro-
sis, and poor outcomes. Although lower levels of virulence result in decreased inflammation during
infection, high tolerance to antimicrobials and lack of effective immune responses may still result
in high bacterial burden and poor outcomes among hospitalized patients

14.6  olecular Basis of Adaptation of CA-MRSA to Low

M
Virulence States in the Hospital Environment

14.6.1  enetic Mechanisms Other than agr Mutation

G
in the Evolution of Attenuated Cytotoxicity

We now know that bacterial isolates from invasive infection often differ pheno-
typically from naturally occurring strains that colonize subjects in the commu-
nity, exhibiting high frequencies of decreased cytotoxicity and reduced agr
activity. Moreover, they are associated with worse outcomes [24, 25, 38, 70,
87]. Colonizing isolates represent the source for disease [28], and isolates in
hospital-associated settings are the reservoir for low-virulence phenotypes [27].
A better understanding of low-virulence strains is critical not only to evaluate
therapeutic efficacy of vaccine and monoclonal antibody strategies under
14 Staphylococcus aureus Adaptation During Infection 447

development for prophylaxis and therapy but also for the important first step of
target identification. To date, efforts to define the polymorphism and expression
heterogeneity of antigens associated with targets of protection have been mostly
limited to individual agr-regulated toxin antigens, such as hla, that are expressed
poorly, if at all, among HA-MRSA lineages that are implicated in complicated
infections, such as bacteremia [24, 89, 90]. However, it is possible that adaptive
variants result from changes that do not alter agr sequences or expression. To
test for non-agr specific selection processes, additional mutations can be sought
using genome sequencing to define the complete set of changes that accompany
the transition to host and hospital-adapted phenotypes.
Recent studies have used genome sequencing to provide evidence for rapid
pathogen genome diversification, some of which could potentially affect the course
of disease (reviewed in [95]). Much of the variation described has been measured
between bacterial isolates of a single patient during infection, either at a single time
point or longitudinally. Although longitudinal studies in single patients have been
highly informative (e.g., [21]), they often focus on resistance mutations during anti-
biotic exposure; effects of selection are superimposed upon mutational patterns
generated by antimicrobial damage and repair processes. In contrast, the overlap
between the within-host and the between-host levels that defines adaptation to the
hospital environment results in heterogeneity of selective pressures that derive from
from functional trade-offs, wherein mutations in genes such as agr may promote
survival in certain host niches in the short term but represent a liability to clones
over the longer term. A consequence of the differential populational stability of
mutations is that pathogens evolve toward fitness on a global scale. Thus, to under-
stand adaptation to a complex environment, such as the hospital, we must examine
evolution in populations as well as in single patients.
Divergence of HA-MRSA strains from community isolates occurred decades
ago, complicating the identification of specific genetic changes associated with
the transition from community to hospital. CA-MRSA strains, such as USA300,
the predominant CA-MRSA clone in the United States, are only distantly related
to hospital-associated MRSA; until recently, they were not linked to the health-
care environment [96, 97]. To limit confounding effects of genetic variability,
genetic background can be controlled by leveraging the recent introduction and
dissemination of CA-MRSA strain USA300 into hospitals. The limited number
of functional mutations associated with this lineage helps identify specific vari-
ants that are associated with the evolution to low-virulence phenotypes in hospi-
tals. Using such an approach, recent genomic studies demonstrate that a number
of MRSA genes, including but not limited to agr, are targets for mutation during
hospital adaptation [70, 98]. Among the non-agr genes are sucD, clpC, tsK, and
rpsA [70].
Identification of genetic changes that potentially drive or are associated with
emergence of host- and hosptial-adapted strains highlight the need to rapidly
reduce bacterial burden and restrict the emergence of adapted mutants, as well
as vigorous surveillance and early public health intervention to limit further
adaptation. Additionally, the mutations that accumulate during adaptation pro-
448 B. Shopsin and R. Copin

vide vital biological information crucial for the interpretation of outbreak

sequence evolution. Unique epidemiology and pathogens make the course of
nosocomial outbreaks unpredictable, complicating efforts to plan for the next
outbreak. Nonetheless, a better understanding of infection traits involved in the
epidemiology (e.g., virulence, metabolism, resistance) will likely facilitate inte-
gration of genomic and epidemiological analysis, allowing more effective tar-
geting of intervention strategies.

14.6.2  otential Role of Mobile Genetic Element Diversity

P
During Hospital Adaptation

Although genome sequencing has improved our understanding of the mutations and
genetic polymorphism associated with S. aureus strains isolated from patients dur-
ing invasive and hospital-associated infections [99–101], the contribution of mobile
genetic elements (MGE) remains under-investigated. MGE discovery and identifi-
cation are important goals for genome analysis because almost all S. aureus strains
harbor one or more MGE with potentially syndrome-specific and tissue-specific
functions. Staphylococcal MGEs include plasmids, transposons, integrons, genomic
islands, S. aureus pathogenicity islands (SaPIs), integrative conjugative elements,
staphylococcal chromosome cassettes, and phages [102–104]. MGEs are abundant,
representing 15–20% of the S. aureus genome. Staphylococcal MGEs are thus an
important source of variation in S. aureus strains and remain, for the most part,
functionally uncharacterized. Together, phages and plasmids are the main source of
MGE diversity among S. aureus strains [105].
Phages are bacterial viruses that can integrate into the S. aureus chromosome. As
the primary vehicles for horizontal transfer of MGEs between strains, phages are
major drivers of staphylococcal genome evolution [106]. All S. aureus genomes
sequenced to date contain at least one phage, and many strains carry up to four
[107]. In addition to numerous hypothetical proteins, phages carry antibiotic resis-
tance genes and virulence factors, such as the Panton-Valentine leukocidin (PVL)
toxin [108, 109]. Phages also mobilize other MGEs, such as plasmids and staphylo-
coccal pathogenicity islands (SaPIs) [103]. SaPIs, which integrate into specific
attachment sites in the chromosome, employ phage machinery for replication and
dissemination [102]. Thus, phages impact both the virulence and adaptability of S.
aureus.
Genomic comparisons between successful S. aureus clones and distantly
related strains have provided insights into MGE contribution to S. aureus adap-
tation [110–112]. For example, genomic studies have pinpointed the dramatic
impact of MGE on the success of MRSA clones, such as USA300, the predomi-
nant CA-MRSA clone in the United States [113]. Notably, they highlighted the
importance of the association between phage-derived PVL with purulent skin
and soft tissue infection and with necrotizing pneumonia and other syndrome-
specific MGE, such as the arginine catabolic mobile element [111, 112]. Other
14 Staphylococcus aureus Adaptation During Infection 449

studies have extensively characterized the contribution of MGE to S. aureus

antibiotic resistance [108, 114].
Recent data suggest that genomic rearrangement through MGE exchange can
contribute to the success during within-host adaptation of S. aureus infection
[115]. An example is given by the regulation of β-toxin expression through bac-
teriophage φSa3 excision. β-toxin is a sphingomyelinase, encoded by virtually
all S. aureus strains, that exhibits human cell cytotoxicity [116]. Despite the
presence of the gene, the majority of human S. aureus isolates are reported not
to express β-toxin due to integration of the bacteriophage φSa3 into the β-toxin
structural gene hlb [117]. φSa3 encodes accessory virulence genes reported to
be involved in immune evasion, including superantigens, staphylokinase, che-
motaxis inhibitory protein, and staphylococcal complement inhibitor [117].
Given the human specificity of these gene products and the high incidence of
φSa3 among human isolates, it has been suggested that the bacteriophage-
encoded virulence factors provide a greater advantage to S. aureus for human
colonization and initial survival than β-toxin production [115]. Recent studies,
however, demonstrated that φSa3 inactivation of β-toxin is reversible [115,
118]. Indeed, S. aureus strains from infective endocarditis patients showed evi-
dence of phage excision and β-toxin restoration compared to isolates from nares
of healthy individuals or to laboratory strains. Collectively, these observations
suggest that during invasive infection, host pressure favors MGE rearrange-
ments with consequent variation in the regulation of genes, possibly benefiting
S. aureus survival and disease progression in vivo.
Identification and discovery of novel MGEs involved in adaptation is compli-
cated by a multitude of repeat regions and mosaic elements that cannot be assem-
bled using traditional “short-read” sequencing instruments. However, de novo
sequence assembly using a combination of short-read and long-read third-­generation
sequencing can provide a comprehensive characterization of MGEs and repeat
regions [119, 120]. Although “hybrid” long- and short-read sequencing is currently
costlier than short-read sequencing alone, the cost is offset by much greater accu-
racy in MGE analysis. Moreover, increases in long-read sequencing accuracy,
throughput, and reduced multiplex sequencing costs will likely eliminate the need
for hybrid analysis in the near future.
Inconsistent annotation and nomenclature in public databases is a complicat-
ing factor. For example, ~40% of S. aureus genes have inconsistent annotation
across S. aureus lineages due to the lack of an adequate framework for ensuring
standardized genome analysis methods and curation of S. aureus meta-data
[105]. This percentage reaches 90% when applied to phages and plasmids,
which represent the main source of MGE diversity among S. aureus public
sequences [105]. As whole-­genome sequencing tools improve and the number
of sequenced genomes multiply, the need for curated databases will be crucial
for rapid sequence analysis across strains. Elimination of the time-constraining
process of creating tools for analyzing strain-specific orthology promises to
produce a significant resource for researchers seeking a greater understanding
of S. aureus and other pathogens.
450 B. Shopsin and R. Copin

14.6.3  otential Role of Metabolic Changes During Adaptation

P
to the Hospital

Even modest genetic changes in a bacterial pathogen can have a dramatic impact on
virulence factor expression and the host-pathogen interaction [121–123]. Although
the presence or absence of a particular gene or polymorphism can be determined,
we do not yet have tools to readily extract information about gene expression in an
organism from the aggregate of its sequence data alone. Thus, given that multiple
mutations and MGEs are involved in the process of host and hospital adaptation, it
is not obvious that an understanding of any one mutation or change will be sufficient
to define what constitutes a hospital-adapted strain.
The research challenge posed by these observations is to reduce the heterogeneity
of hospital adaptation so that we can identify a proxy for hospital adaptation that can
explain epidemiological patterns and clinical outcomes. To augment mutation discov-
ery, we propose a novel approach focused on the phenotypic effects of mutation: anal-
ysis of gene expression, proteomic, and metabolic pathways using information-­rich
datasets derived from low-virulence MRSA populations. Once identified, pathways
can be directly assessed for phenotypic and genetic variation in individual strains.
The metabolic underpinnings of bacterial virulence, host defense, fitness, and
antimicrobial lethality underscore metabolic capacity as an attractive candidate to
determine the ability of MRSA to adapt to the hospital environment. The frequent
occurrence of dysfunctional tricarboxylic acid (TCA) cycle activity in clinical iso-
lates of staphylococci supports the idea that genetic adaptation can redirect metabo-
lism [124], as does the observation that downregulation of TCA cycle activity is
common among small colony-variant phenotypes that are associated with chronic
infections and antimicrobial resistance [125–127].
These observations give rise to the hypothesis that MRSA use mutation to fine-­
tune their metabolism to sustain fitness in response to selective pressures in the
hospital environment. As CA-MRSA become HA-MRSA, metabolic heterogeneity
ensues. Our working model is shown in Fig. 14.6.

14.7 Concluding Remarks

We expect the cancer-biology style systems biology approach to enable character-

ization of the genotypic and phenotypic pathways of low-virulence MRSA popula-
tions that are evolving heterogeneity. The knowledge base generated can be used to
describe the range of features of hospital-associated, low-virulence MRSA that are
associated with complicated infection. Identification of adatpive phenotypes would
open new avenues of further study to identify how mutations may create new vul-
nerabilities that can be used for treatment. This hope arises by analogy with the situ-
ation in typical cancers in which mutational loss of DNA repair checkpoints favors
clonal expansion but also confers sensitivity to DNA-damaging agents.
14 Staphylococcus aureus Adaptation During Infection 451

Fig. 14.6 Differential susceptibility hypothesis of nosocomial adaptation. Current work focuses
on resistant and susceptible S. aureus; we hypothesize that S. aureus displays a spectrum of meta-
bolic states that causes strains to differ in the degree to which they tolerate host and antimicrobial
stress. Hospital-adaptation is accompanied by distinct stress tolerance states that are associated
with distinct metabolic states, and these metabolic states can be identified as signatures that define
specificity for different pathways to metabolize substrates, as well as differences in virulence phe-
notypes and functions. Differential susceptibility of an organism to stress is sometimes called the
orchid hypothesis. In the figure, the boundary between community and hospital-adapted states is
not distinct; however, CA-MRSA are like delicate orchids; and HA-MRSA are like dandelions.
The former are highly virulent, and if left untreated causing more severe disease, they quickly
wither if exposed to antimicrobial stress. The latter prove resilient to the effects of antimicrobials,
but at the same time, they are not particularly virulent. CA-MRSA represents a “high” metabolic
state, although high-metabolic states may be favored during certain hospital environments, which
can account for inter-strain variability. In this scenario, the Goldilocks metaphor is applicable;
extreme states impair fitness and are considered either deleterious or “short-sighted.” Metabolism
refers to growth rate and metabolic “complexity”

Efforts to understand the pathogenesis of infections often assumes that for a

given pathogen, organisms respond equally to management attempts and that
interventions, such as antibiotics, should therefore equally help all patients with a
given infection. During adaptation to the hospital, a variety of strong selective
pressures are exerted on the pathogen. The resulting genetic changes that occur
might influence clinical outcomes and affect diagnosis and epidemiology.
Identification of hospital and host-adaptive distinctions could be used with future
point-of-care diagnostics to screen for characteristics that define patterns of
MRSA epidemiology and clinical outcomes. Adaptations that erode the activities
of bactericidal antimicrobials might also inform the potential use of novel strate-
gies to provide an adjunctive to selectively bolster the effectiveness of antimicro-
bials [82–84, 128–130].
Finally, we note that dentification and characterization of infection-associated
phenotypes is also significant from a basic research perspective: identification of
452 B. Shopsin and R. Copin

changes that have a profound impact on the S. aureus transcriptome and on viru-
lence has important implications for the study of gene regulation in the pathogenic-
ity of CA-MRSA infection. For example, differences in pathology observed in
infections caused by mutants constructed in the laboratory may be obscured or com-
pounded by a failure to undergo mutation of agr to one form or another: only careful
comparison of strains going into and coming out of a host can identify the extent to
which this is a problem. Thus, field work is essential. By unraveling the evolution-
ary steps occurring during infection and characterizing the genetic basis of such
changes, we expect to provide a framework for interpreting the phenotype of newly
emergent clones.
Major Points
• Functional compromise of pleiotropic regulators, which control many genes in
complex genetic networks, provides a broad framework in which alterations in
cellular circuitry promote the emergence of new traits.
• Mutation of agr is an explicit example of a mechanism by which hospital adapta-
tion can alter the relationship between virulence, infectivity, and host and syn-
thetic antimicrobial susceptibility.
• Mechanisms of adaptation-dependent changes in antimicrobial susceptibility
include direct and indirect effects on fitness and antimicrobial tolerance.
• Mutations in genes such as agr may be adaptive for survival in the infected host,
but they are counter-adaptive outside infected host tissues or in situtations that
involve high attack rates, such as outbreaks.
• Hospital adaptations are likely to derive from a range of heritable changes in
virulence attenuation, metabolic activity, and antimicrobial susceptibility, all of
which are linked to the oxidative stress network that participates in host and
synthetic antimicrobial-mediated killing.
• Identification of polymorphisms and targets of selection that affect the cellular
response of MRSA to antimicrobial drugs may increase our understanding of
drug tolerance mechanisms; such identification will also provide an initial step
toward predictive tailoring of drug treatments to individuals to maximize thera-
peutic benefit.

Acknowledgments We thank Karl Drlica for helpful discussions and critical comments. Our
work was supported by NIH grants AI103268 and N272201400019C.

References

1. Dwyer DJ, Belenky PA, Yang JH, et al. Antibiotics induce redox-related physiological altera-
tions as part of their lethality. Proc Natl Acad Sci U S A. 2014;111:E2100–9.
2. Dwyer DJ, Collins JJ, Walker GC. Unraveling the physiological complexities of antibiotic
lethality. Annu Rev Pharmacol Toxicol. 2015;55:313–32.
3. Zhao X, Drlica K. Reactive oxygen species and the bacterial response to lethal stress. Curr
Opin Microbiol. 2014;21:1–6.
14 Staphylococcus aureus Adaptation During Infection 453

4. Zhao X, Hong Y, Drlica K. Moving forward with reactive oxygen species involvement in
antimicrobial lethality. J Antimicrob Chemother. 2015;70:639–42.
5. Lowy FD. Staphylococcus aureus infections. N Engl J Med. 1998;339:520–32.
6. Smith TL, Pearson ML, Wilcox KR, et al. Emergence of vancomycin resistance in
Staphylococcus aureus. Glycopeptide-intermediate Staphylococcus aureus working group.
N Engl J Med. 1999;340:493–501.
7. From the Centers for Disease Control. Staphylococcus aureus resistant to vancomycin–
United States, 2002. JAMA. 2002;288:824–5.
8. From the Centers for Disease Control and Prevention. Vancomycin resistant Staphylococcus
aureus–Pennsylvania, 2002. JAMA. 2002;288:2116.
9. Novick RP. Autoinduction and signal transduction in the regulation of staphylococcal viru-
lence. Mol Microbiol. 2003;48:1429–49.
10. Cirioni O, Giacometti A, Ghiselli R, et al. Prophylactic efficacy of topical temporin A and
RNAIII-inhibiting peptide in a subcutaneous rat Pouch model of graft infection attributable to
staphylococci with intermediate resistance to glycopeptides. Circulation. 2003;108:767–71.
11. Domenico P, Gurzenda E, Giacometti A, et al. BisEDT and RIP act in synergy to prevent graft
infections by resistant staphylococci. Peptides. 2004;25:2047–53.
12. Dell'Acqua G, Giacometti A, Cirioni O, et al. Suppression of drug-resistant staphylococ-
cal infections by the quorum-sensing inhibitor RNAIII-inhibiting peptide. J Infect Dis.
2004;190:318–20.
13. Cirioni O, Giacometti A, Ghiselli R, et al. RNAIII-inhibiting peptide significantly reduces
bacterial load and enhances the effect of antibiotics in the treatment of central venous
catheter-­associated Staphylococcus aureus infections. J Infect Dis. 2006;193:180–6.
14. Wright JS 3rd, Jin R, Novick RP. Transient interference with staphylococcal quorum sensing
blocks abscess formation. Proc Natl Acad Sci U S A. 2005;102:1691–6.
15. Goerke C, Matias y Papenberg S, Dasbach S, et al. Increased frequency of genomic altera-
tions in Staphylococcus aureus during chronic infection is in part due to phage mobilization.
J Infect Dis. 2004;189:724–34.
16. Painter KL, Krishna A, Wigneshweraraj S, Edwards AM. What role does the quorum-­sensing
accessory gene regulator system play during Staphylococcus aureus bacteremia? Trends
Microbiol. 2014;22:676–85.
17. Sakoulas G, Moellering RC Jr, Eliopoulos GM. Adaptation of methicillin-resistant
Staphylococcus aureus in the face of vancomycin therapy. Clin Infect Dis. 2006;42(Suppl
1):S40–50.
18. Traber KE, Lee E, Benson S, et al. agr function in clinical Staphylococcus aureus isolates.
Microbiology. 2008;154:2265–74.
19. Vuong C, Kocianova S, Yao Y, Carmody AB, Otto M. Increased colonization of indwelling
medical devices by quorum-sensing mutants of Staphylococcus epidermidis in vivo. J Infect
Dis. 2004;190:1498–505.
20. Vuong C, Saenz HL, Gotz F, Otto M. Impact of the agr quorum-sensing system on adherence
to polystyrene in Staphylococcus aureus. J Infect Dis. 2000;182:1688–93.
21. Mwangi MM, Wu SW, Zhou Y, et al. Tracking the in vivo evolution of multidrug resis-
tance in Staphylococcus aureus by whole-genome sequencing. Proc Natl Acad Sci U S A.
2007;104:9451–6.
22. Traber K, Novick R. A slipped-mispairing mutation in AgrA of laboratory strains and clin-
ical isolates results in delayed activation of agr and failure to translate delta- and alpha-­
haemolysins. Mol Microbiol. 2006;59:1519–30.
23. Schwan WR, Langhorne MH, Ritchie HD, Stover CK. Loss of hemolysin expression in
Staphylococcus aureus agr mutants correlates with selective survival during mixed infections
in murine abscesses and wounds. FEMS Immunol Med Microbiol. 2003;38:23–8.
24. Fowler VG Jr, Sakoulas G, McIntyre LM, et al. Persistent bacteremia due to methicillin-­
resistant Staphylococcus aureus infection is associated with agr dysfunction and low-­
level in vitro resistance to thrombin-induced platelet microbicidal protein. J Infect Dis.
2004;190:1140–9.
454 B. Shopsin and R. Copin

25. Schweizer ML, Furuno JP, Sakoulas G, et al. Increased mortality with accessory gene regu-
lator (agr) dysfunction in Staphylococcus aureus among bacteremic patients. Antimicrob
Agents Chemother. 2011;55:1082–7.
26. Khan BA, Yeh AJ, Cheung GY, Otto M. Investigational therapies targeting quorum-­
sensing for the treatment of Staphylococcus aureus infections. Expert Opin Investig Drugs.
2015;24:689–704.
27. Shopsin B, Drlica-Wagner A, Mathema B, Adhikari RP, Kreiswirth BN, Novick
RP. Prevalence of agr dysfunction among colonizing Staphylococcus aureus strains. J Infect
Dis. 2008;198:1171–4.
28. Smyth DS, Kafer JM, Wasserman GA, et al. Nasal carriage as a source of agr-defective
Staphylococcus aureus bacteremia. J Infect Dis. 2012;206:1168–77.
29. Shopsin B, Eaton C, Wasserman GA, et al. Mutations in agr do not persist in natural popula-
tions of methicillin-resistant Staphylococcus aureus. J Infect Dis. 2010;202:1593–9.
30. Kernbauer E, Maurer K, Torres VJ, Shopsin B, Cadwell K. Gastrointestinal dissemination and
transmission of Staphylococcus aureus following bacteremia. Infect Immun. 2015;83:372–8.
31. Hunter P. The great leap forward. Major evolutionary jumps might be caused by changes in
gene regulation rather than the emergence of new genes. EMBO Rep. 2008;9:608–11.
32. Wittkopp PJ, Haerum BK, Clark AG. Evolutionary changes in cis and trans gene regulation.
Nature. 2004;430:85–8.
33. Barrier M, Robichaux RH, Purugganan MD. Accelerated regulatory gene evolution in an
adaptive radiation. Proc Natl Acad Sci U S A. 2001;98:10208–13.
34. Quan S, Ray JC, Kwota Z, et al. Adaptive evolution of the lactose utilization network in
experimentally evolved populations of Escherichia coli. PLoS Genet. 2012;8:e1002444.
35. Saxer G, Krepps MD, Merkley ED, et al. Mutations in global regulators lead to metabolic
selection during adaptation to complex environments. PLoS Genet. 2014;10:e1004872.
36. Spencer CC, Bertrand M, Travisano M, Doebeli M. Adaptive diversification in genes that
regulate resource use in Escherichia coli. PLoS Genet. 2007;3:e15.
37. Carter MQ, Parker CT, Louie JW, Huynh S, Fagerquist CK, Mandrell RE. RcsB contrib-
utes to the distinct stress fitness among Escherichia coli O157:H7 curli variants of the 1993
hamburger-associated outbreak strains. Appl Environ Microbiol. 2012;78:7706–19.
38. Das S, Lindemann C, Young BC, et al. Natural mutations in a Staphylococcus aureus viru-
lence regulator attenuate cytotoxicity but permit bacteremia and abscess formation. Proc Natl
Acad Sci U S A. 2016;113:E3101–10.
39. Kisiela DI, Radey M, Paul S, et al. Inactivation of transcriptional regulators during within-­
household evolution of Escherichia coli. J Bacteriol. 2017;199:e00036-17.
40. Smith EE, Buckley DG, Wu Z, et al. Genetic adaptation by Pseudomonas aeruginosa to the
airways of cystic fibrosis patients. Proc Natl Acad Sci U S A. 2006;103:8487–92.
41. Balasubramanian D, Harper L, Shopsin B, Torres VJ. Staphylococcus aureus pathogenesis in
diverse host environments. Pathog Dis. 2017;75(1):ftx005.
42. Novick RP, Geisinger E. Quorum sensing in staphylococci. Annu Rev Genet. 2008;42:541–64.
43. D'Argenio DA, Wu M, Hoffman LR, et al. Growth phenotypes of Pseudomonas aeru-
ginosa lasR mutants adapted to the airways of cystic fibrosis patients. Mol Microbiol.
2007;64:512–33.
44. Feltner JB, Wolter DJ, Pope CE, et al. LasR variant cystic fibrosis isolates reveal an adaptable
quorum-sensing hierarchy in Pseudomonas aeruginosa. MBio. 2016;7:e01513.
45. Hoffman LR, Kulasekara HD, Emerson J, et al. Pseudomonas aeruginosa lasR mutants are
associated with cystic fibrosis lung disease progression. J Cyst Fibros. 2009;8:66–70.
46. Wilder CN, Allada G, Schuster M. Instantaneous within-patient diversity of Pseudomonas
aeruginosa quorum-sensing populations from cystic fibrosis lung infections. Infect Immun.
2009;77:5631–9.
47. Yarwood JM, Paquette KM, Tikh IB, Volper EM, Greenberg EP. Generation of virulence fac-
tor variants in Staphylococcus aureus biofilms. J Bacteriol. 2007;189:7961–7.
48. Pratten J, Foster SJ, Chan PF, Wilson M, Nair SP. Staphylococcus aureus accessory regula-
tors: expression within biofilms and effect on adhesion. Microbes Infect. 2001;3:633–7.
14 Staphylococcus aureus Adaptation During Infection 455

49. Xiong YQ, Bayer AS, Yeaman MR, Van Wamel W, Manna AC, Cheung AL. Impacts of sarA
and agr in Staphylococcus aureus strain Newman on fibronectin-binding protein A gene
expression and fibronectin adherence capacity in vitro and in experimental infective endocar-
ditis. Infect Immun. 2004;72:1832–6.
50. Sakoulas G, Eliopoulos GM, Fowler VG Jr, et al. Reduced susceptibility of Staphylococcus
aureus to vancomycin and platelet microbicidal protein correlates with defective autoly-
sis and loss of accessory gene regulator (agr) function. Antimicrob Agents Chemother.
2005;49:2687–92.
51. DuMont AL, Yoong P, Surewaard BG, et al. Staphylococcus aureus elaborates leukocidin AB
to mediate escape from within human neutrophils. Infect Immun. 2013;81:1830–41.
52. Koziel J, Maciag-Gudowska A, Mikolajczyk T, et al. Phagocytosis of Staphylococcus aureus
by macrophages exerts cytoprotective effects manifested by the upregulation of antiapoptotic
factors. PLoS One. 2009;4:e5210.
53. Kubica M, Guzik K, Koziel J, et al. A potential new pathway for Staphylococcus aureus
dissemination: the silent survival of S. aureus phagocytosed by human monocyte-derived
macrophages. PLoS One. 2008;3:e1409.
54. Lehar SM, Pillow T, Xu M, et al. Novel antibody-antibiotic conjugate eliminates intracellular
S. aureus. Nature. 2015;527:323–8.
55. Surewaard BG, Deniset JF, Zemp FJ, et al. Identification and treatment of the Staphylococcus
aureus reservoir in vivo. J Exp Med. 2016;213:1141–51.
56. Zeng Z, Surewaard BG, Wong CH, Geoghegan JA, Jenne CN, Kubes P. CRIg functions as
a macrophage pattern recognition receptor to directly bind and capture blood-borne gram-­
positive bacteria. Cell Host Microbe. 2016;20:99–106.
57. Fraunholz M, Sinha B. Intracellular Staphylococcus aureus: live-in and let die. Front Cell
Infect Microbiol. 2012;2:43.
58. Soong G, Paulino F, Wachtel S, et al. Methicillin-resistant Staphylococcus aureus adaptation
to human keratinocytes. MBio. 2015;6:e00289-15.
59. Wickersham M, Wachtel S, Wong Fok Lung T, et al. Metabolic stress drives keratinocyte
defenses against Staphylococcus aureus infection. Cell Rep. 2017;18:2742–51.
60. Tsuji BT, MacLean RD, Dresser LD, McGavin MJ, Simor AE. Impact of accessory gene
regulator (agr) dysfunction on vancomycin pharmacodynamics among Canadian community
and health-care associated methicillin-resistant Staphylococcus aureus. Ann Clin Microbiol
Antimicrob. 2011;10:20.
61. Butterfield J, Tsuji B, Brown J, et al. Predictors of agr dysfunction in methicillin-resistant
Staphylococcus aureus (MRSA) isolates among patients with MRSA bloodstream infec-
tions in the “15-20 mg/L” target vancomycin trough era. Antimicrob Agents Chemother.
2011;55:5433.
62. Paulander W, Nissen Varming A, Baek KT, Haaber J, Frees D, Ingmer H. Antibiotic-mediated
selection of quorum-sensing-negative. MBio. 2013;3:e00459–12.
63. Dai Y, Chang W, Zhao C, et al. VraR binding to the promoter region of agr inhibits its
function in vancomycin-intermediate Staphylococcus aureus (VISA) and heterogeneous
VISA. Antimicrob Agents Chemother. 2017;61:e02740.
64. Sakoulas G, Moise PA, Rybak MJ. Accessory gene regulator dysfunction: an advantage for
Staphylococcus aureus in health-care settings? J Infect Dis. 2009;199:1558–9.
65. Mishra NN, Bayer AS, Moise PA, Yeaman MR, Sakoulas G. Reduced susceptibility to host-­
defense cationic peptides and daptomycin coemerge in methicillin-resistant Staphylococcus
aureus from daptomycin-naive bacteremic patients. J Infect Dis. 2012;206:1160–7.
66. Somerville GA, Beres SB, Fitzgerald JR, et al. In vitro serial passage of Staphylococcus
aureus: changes in physiology, virulence factor production, and agr nucleotide sequence.
J Bacteriol. 2002;184:1430–7.
67. Dunman PM, Murphy E, Haney S, et al. Transcription profiling-based identifica-
tion of Staphylococcus aureus genes regulated by the agr and/or sarA loci. J Bacteriol.
2001;183:7341–53.
456 B. Shopsin and R. Copin

68. Zorzet A, Andersen JM, Nilsson AI, Moller NF, Andersson DI. Compensatory mutations in
agrC partly restore fitness in vitro to peptide deformylase inhibitor-resistant Staphylococcus
aureus. J Antimicrob Chemother. 2012;67:1835–42.
69. Sun F, Liang H, Kong X, et al. Quorum-sensing agr mediates bacterial oxidation response via
an intramolecular disulfide redox switch in the response regulator AgrA. Proc Natl Acad Sci
U S A. 2012;109:9095–100.
70. Laabei M, Uhlemann AC, Lowy FD, et al. Evolutionary trade-offs underlie the multi-faceted
virulence of Staphylococcus aureus. PLoS Biol. 2015;13:e1002229.
71. McNamara PJ, Milligan-Monroe KC, Khalili S, Proctor RA. Identification, cloning, and ini-
tial characterization of rot, a locus encoding a regulator of virulence factor expression in
Staphylococcus aureus. J Bacteriol. 2000;182:3197–203.
72. Kumar K, Chen J, Drlica K, Shopsin B. Tuning of the lethal response to multiple stress-
ors with a single-site mutation during clinical infection by Staphylococcus aureus. MBio.
2017;8:e01476.
73. Liu Y, Liu X, Qu Y, Wang X, Li L, Zhao X. Inhibitors of reactive oxygen species accumula-
tion delay and/or reduce the lethality of several antistaphylococcal agents. Antimicrob Agents
Chemother. 2012;56:6048–50.
74. Pader V, Hakim S, Painter KL, Wigneshweraraj S, Clarke TB, Edwards AM. Staphylococcus
aureus inactivates daptomycin by releasing membrane phospholipids. Nat Microbiol.
2016;2:16194.
75. Queck SY, Jameson-Lee M, Villaruz AE, et al. RNAIII-independent target gene control
by the agr quorum-sensing system: insight into the evolution of virulence regulation in
Staphylococcus aureus. Mol Cell. 2008;32:150–8.
76. Keller L, Surette MG. Communication in bacteria: an ecological and evolutionary perspec-
tive. Nat Rev Microbiol. 2006;4:249–58.
77. Sandoz KM, Mitzimberg SM, Schuster M. Social cheating in Pseudomonas aeruginosa quo-
rum sensing. Proc Natl Acad Sci U S A. 2007;104:15876–81.
78. Booth MC, Cheung AL, Hatter KL, Jett BD, Callegan MC, Gilmore MS. Staphylococcal
accessory regulator (sar) in conjunction with agr contributes to Staphylococcus aureus viru-
lence in endophthalmitis. Infect Immun. 1997;65:1550–6.
79. Gillaspy AF, Hickmon SG, Skinner RA, Thomas JR, Nelson CL, Smeltzer MS. Role of
the accessory gene regulator (agr) in pathogenesis of staphylococcal osteomyelitis. Infect
Immun. 1995;63:3373–80.
80. Plata KB, Rosato RR, Rosato AE. Fate of mutation rate depends on agr locus expression
during oxacillin-mediated heterogeneous-homogeneous selection in methicillin-resistant
Staphylococcus aureus clinical strains. Antimicrob Agents Chemother. 2011;55:3176–86.
81. Malik M, Hoatam G, Chavda K, Kerns RJ, Drlica K. Novel approach for comparing the abili-
ties of quinolones to restrict the emergence of resistant mutants during quinolone exposure.
Antimicrob Agents Chemother. 2010;54:149–56.
82. Gusarov I, Shatalin K, Starodubtseva M, Nudler E. Endogenous nitric oxide protects bacteria
against a wide spectrum of antibiotics. Science. 2009;325:1380–4.
83. Liu Y, Zhou J, Qu Y, et al. Resveratrol antagonizes antimicrobial lethality and stimulates
recovery of bacterial mutants. PLoS One. 2016;11:e0153023.
84. Shatalin K, Shatalina E, Mironov A, Nudler E. H2S: a universal defense against antibiotics in
bacteria. Science. 2011;334:986–90.
85. Rose HR, Holzman RS, Altman DR, et al. Cytotoxic virulence predicts mortality in nos-
ocomial pneumonia due to methicillin-resistant Staphylococcus aureus. J Infect Dis.
2015;211:1862–74.
86. Mendes RE, Deshpande LM, Smyth DS, Shopsin B, Farrell DJ, Jones RN. Characterization
of methicillin-resistant Staphylococcus aureus strains recovered from a phase IV clinical trial
for linezolid versus vancomycin for treatment of nosocomial pneumonia. J Clin Microbiol.
2012;50:3694–702.
87. Baines SL, Holt KE, Schultz MB, et al. Convergent adaptation in the dominant global hospi-
tal clone ST239 of methicillin-resistant. MBio. 2015;6:e00080.
14 Staphylococcus aureus Adaptation During Infection 457

88. Cheung GY, Kretschmer D, Duong AC, et al. Production of an attenuated phenol-soluble
modulin variant unique to the MRSA clonal complex 30 increases severity of bloodstream
infection. PLoS Pathog. 2014;10:e1004298.
89. Deleo FR, Kennedy AD, Chen L, et al. Molecular differentiation of historic phage-type
80/81 and contemporary epidemic Staphylococcus aureus. Proc Natl Acad Sci U S A.
2011;108:18091–6.
90. Fowler VG Jr, Nelson CL, McIntyre LM, et al. Potential associations between hematoge-
nous complications and bacterial genotype in Staphylococcus aureus infection. J Infect Dis.
2007;196:738–47.
91. Alegre ML, Chen L, David MZ, et al. Impact of Staphylococcus aureus USA300 colo-
nization and skin infections on systemic immune responses in humans. J Immunol.
2016;197(4):1118–26
92. Malachowa N, Kobayashi SD, Braughton KR, et al. Staphylococcus aureus leukotoxin GH
promotes inflammation. J Infect Dis. 2012;206:1185–93.
93. Malachowa N, Kobayashi SD, Freedman B, Dorward DW, DeLeo FR. Staphylococcus
aureus leukotoxin GH promotes formation of neutrophil extracellular traps. J Immunol.
2013;191:6022–9.
94. Yoong P, Pier GB. Immune-activating properties of Panton-valentine leukocidin improve
the outcome in a model of methicillin-resistant Staphylococcus aureus pneumonia. Infect
Immun. 2012;80:2894–904.
95. Didelot X, Walker AS, Peto TE, Crook DW, Wilson DJ. Within-host evolution of bacterial
pathogens. Nat Rev Microbiol. 2016;14:150–62.
96. Popovich KJ, Weinstein RA, Hota B. Are community-associated methicillin-resistant
Staphylococcus aureus (MRSA) strains replacing traditional nosocomial MRSA strains? Clin
Infect Dis. 2008;46:787–94.
97. Seybold U, Kourbatova EV, Johnson JG, et al. Emergence of community-associated
methicillin-­resistant Staphylococcus aureus USA300 genotype as a major cause of health
care-associated blood stream infections. Clin Infect Dis. 2006;42:647–56.
98. Laabei M, Recker M, Rudkin JK, et al. Predicting the virulence of MRSA from its genome
sequence. Genome Res. 2014;24:839–49.
99. Harris SR, Cartwright EJ, Torok ME, et al. Whole-genome sequencing for analysis of an
outbreak of meticillin-resistant Staphylococcus aureus: a descriptive study. Lancet Infect Dis.
2013;13:130–6.
100. Planet PJ, Diaz L, Kolokotronis SO, et al. Parallel epidemics of community-associated
methicillin-resistant Staphylococcus aureus USA300 infection in North and South America.
J Infect Dis. 2015;212:1874–82.
101. Tong SY, Holden MT, Nickerson EK, et al. Genome sequencing defines phylogeny and
spread of methicillin-resistant Staphylococcus aureus in a high transmission setting. Genome
Res. 2015;25:111–8.
102. Penades JR, Christie GE. The phage-inducible chromosomal islands: a family of highly
evolved molecular parasites. Annu Rev Virol. 2015;2:181–201.
103. Novick RP, Christie GE, Penades JR. The phage-related chromosomal islands of gram-­
positive bacteria. Nat Rev Microbiol. 2010;8:541–51.
104. Malachowa N, DeLeo FR. Mobile genetic elements of Staphylococcus aureus. Cell Mol Life
Sci. 2010;67:3057–71.
105. Copin R, Shopsin B, Torres JV. After de deluge: mining Staphylococcus aureus genomic
data for clinical associations and host pathogen interactions. Curr Opin Microbiol.
2018;41C:43–50.
106. Xia G, Wolz C. Phages of Staphylococcus aureus and their impact on host evolution. Infect
Genet Evol. 2014;21:593–601.
107. Lindsay JA. Genomic variation and evolution of Staphylococcus aureus. Int J Med Microbiol.
2010;300:98–103.
108. Haaber J, Leisner JJ, Cohn MT, et al. Bacterial viruses enable their host to acquire antibiotic
resistance genes from neighbouring cells. Nat Commun. 2016;7:13333.
458 B. Shopsin and R. Copin

109. Nguyen LT, Vogel HJ. Staphylokinase has distinct modes of interaction with antimicrobial
peptides, modulating its plasminogen-activation properties. Sci Rep. 2016;6:31817.
110. Everitt RG, Didelot X, Batty EM, et al. Mobile elements drive recombination hotspots in the
core genome of Staphylococcus aureus. Nat Commun. 2014;5:3956.
111. Gillet Y, Issartel B, Vanhems P, et al. Association between Staphylococcus aureus strains car-
rying gene for Panton-valentine leukocidin and highly lethal necrotising pneumonia in young
immunocompetent patients. Lancet. 2002;359:753–9.
112. Goering RV, McDougal LK, Fosheim GE, Bonnstetter KK, Wolter DJ, Tenover
FC. Epidemiologic distribution of the arginine catabolic mobile element among selected
methicillin-resistant and methicillin-susceptible Staphylococcus aureus isolates. J Clin
Microbiol. 2007;45:1981–4.
113. Diep BA, Gill SR, Chang RF, et al. Complete genome sequence of USA300, an epi-
demic clone of community-acquired meticillin-resistant Staphylococcus aureus. Lancet.
2006;367:731–9.
114. Rolo J, Worning P, Boye Nielsen J, et al. Evidence for the evolutionary steps leading to
mecA-mediated beta-lactam resistance in staphylococci. PLoS Genet. 2017;e1006674:13.
115. Salgado-Pabon W, Herrera A, Vu BG, et al. Staphylococcus aureus beta-toxin production
is common in strains with the beta-toxin gene inactivated by bacteriophage. J Infect Dis.
2014;210:784–92.
116. Herrera A, Kulhankova K, Sonkar VK, et al. Staphylococcal beta-toxin modulates human
aortic endothelial cell and platelet function through sphingomyelinase and biofilm ligase
activities. MBio. 2017;8:e00273.
117. van Wamel WJ, Rooijakkers SH, Ruyken M, van Kessel KP, van Strijp JA. The innate
immune modulators staphylococcal complement inhibitor and chemotaxis inhibitory pro-
tein of Staphylococcus aureus are located on beta-hemolysin-converting bacteriophages.
J Bacteriol. 2006;188:1310–5.
118. Goerke C, Gressinger M, Endler K, et al. High phenotypic diversity in infecting but not in
colonizing Staphylococcus aureus populations. Environ Microbiol. 2007;9:3134–42.
119. Benson MA, Ohneck EA, Ryan C, et al. Evolution of hypervirulence by a MRSA clone
through acquisition of a transposable element. Mol Microbiol. 2014;93:664–81.
120. Altman DR, Sebra R, Hand J, et al. Transmission of methicillin-resistant Staphylococcus
aureus via deceased donor liver transplantation confirmed by whole genome sequencing. Am
J Transplant Off J Am Soc Transplant Am Soc Transplant Surg. 2014;14:2640–4.
121. Beres SB, Richter EW, Nagiec MJ, et al. Molecular genetic anatomy of inter- and intrasero-
type variation in the human bacterial pathogen group A streptococcus. Proc Natl Acad Sci U
S A. 2006;103:7059–64.
122. Kennedy AD, Otto M, Braughton KR, et al. Epidemic community-associated methicillin-­
resistant Staphylococcus aureus: recent clonal expansion and diversification. Proc Natl Acad
Sci U S A. 2008;105:1327–32.
123. Sumby P, Whitney AR, Graviss EA, DeLeo FR, Musser JM. Genome-wide analysis of group
a streptococci reveals a mutation that modulates global phenotype and disease specificity.
PLoS Pathog. 2006;2:e5.
124. Thomas VC, Kinkead LC, Janssen A, et al. A dysfunctional tricarboxylic acid cycle enhances
fitness of Staphylococcus epidermidis during beta-lactam stress. MBio. 2013;4:e00437-13.
125. Kriegeskorte A, Block D, Drescher M, et al. Inactivation of thyA in Staphylococcus aureus
attenuates virulence and has a strong impact on metabolism and virulence gene expression.
MBio. 2014;5:e01447-14.
126. Kriegeskorte A, Grubmuller S, Huber C, et al. Staphylococcus aureus small colony vari-
ants show common metabolic features in central metabolism irrespective of the underlying
auxotrophism. Front Cell Infect Microbiol. 2014;4:141.
127. Kriegeskorte A, Konig S, Sander G, et al. Small colony variants of Staphylococcus aureus
reveal distinct protein profiles. Proteomics. 2011;11:2476–90.
14 Staphylococcus aureus Adaptation During Infection 459

128. Allison KR, Brynildsen MP, Collins JJ. Metabolite-enabled eradication of bacterial persisters
by aminoglycosides. Nature. 2011;473:216–20.
129. Gardner SG, Marshall DD, Daum RS, Powers R, Somerville GA. Metabolic mitigation of
Staphylococcus aureus vancomycin intermediate-level susceptibility. Antimicrob Agents
Chemother. 2018;62:e01608-17.
130. Gutierrez A, Jain S, Bhargava P, Hamblin M, Lobritz MA, Collins JJ. Understanding
and sensitizing density-dependent persistence to quinolone antibiotics. Mol Cell.
2017;68:1147–54. e3.
Chapter 15
Bacterial Signal Transduction Systems
in Antimicrobial Resistance

Andrew T. Ulijasz, Sarah C. Feid, and David G. Glanville

15.1 Introduction

Bacteria must continually sample and assess their environment to survive. Once
pathogens are inside a host, they must detect important external stimuli, including
nutrient availability, oxygen levels, osmolarity, temperature, and even light quality
[1]. Pathogens must also be able to respond to assaults from other competing
microbes within the host, as well as antibacterial activity from the host itself. Indeed,
bacteria produce an entourage of antimicrobial compounds to assure that a particu-
lar niche is maintained in the presence of other bacterial competitors or, in the case
of the host, factors that are produced with the sole intent of clearing the invasive
microbe. It is important that the microbial pathogen responds to such environmental
signals only when necessary. This phenomenon is often referred to as inducible
responses. Bacteria (and most life as we know it) use inducible systems to respond
to a present threat rather than to anticipate one, as the response itself is usually
entergetically costly (e.g., transcription and translation require ATP, etc.).
Consequently, such energy, if its use is not required, is better redirected to another
task or stored for a later time. From these considerations it is thus reasonable to
conclude that antimicrobial resistance and/or tolerance can have a severe fitness
cost, and thus they must be tightly regulated to ensure the long-term survival of an
invading microbe.
In order to properly respond to their changing environment in the most efficient
manner, bacteria have evolved intricate detection mechanisms called signal trans-
duction systems. These systems are based largely on the transfer or “relay” of a
phosphate, as a posttranslational modification, from one protein moiety to another

A. T. Ulijasz (*) · S. C. Feid · D. G. Glanville

Department of Microbiology and Immunology, Loyola University Chicago,
Maywood, IL, USA
e-mail: [email protected]

© Springer International Publishing AG, part of Springer Nature 2018 461

I. W. Fong et al. (eds.), Antimicrobial Resistance in the 21st Century, Emerging Infectious
Diseases of the 21st Century, https://doi.org/10.1007/978-3-319-78538-7_15
462 A. T. Ulijasz et al.

[2], although many other posttranslational modifications are now surfacing as

equally important signaling components (e.g., acetylation [3] and oxidation [4]).
Signal transduction systems sense the threatening antimicrobial and subsequently
elicit a tailored relay to counter the drug’s deleterious effects [5–7]. In addition,
signal transduction systems and other proteins that may not have explicitly evolved
for the sole purpose of conveying antimicrobial resistance (AMR) can incur amino
acid changes that provide an evolutionary advantage over wild-type cells in avoid-
ing the effects of antibiotic treatment [8]. These resistance sensory and response
mechanisms were recognized shortly after the first antibiotic, penicillin, was dis-
covered in 1929. However, it appears that AMR predates the widespread use of
antimicrobials in modern times, suggesting that these molecular mechanisms have
been repurposed through evolutionary pressures in today’s antibiotic era [9]. The
result is that signal transduction systems contribute to one of the most important
medical problems of our time: antimicrobial ineffectiveness [10].
In the most simplistic bacterial signaling system, one protein directly binds the
antimicrobial and elicits a transcriptional response. These systems, commonly
referred to as “one-component” systems, are the most common among the bacterial
signaling family and often take the form of a single transcription factor [11].
However, more complex signaling cascades are usually observed with antibiotic
resistance mechanisms. Those involving two or more components, the most preva-
lent family, are aptly referred to as two-component signaling (TCS) systems [5].
TCS systems usually consist of a membrane-bound sensory histidine kinase that,
upon stimulation by a specific environmental que (e.g., the direct or indirect sensing
of an antimicrobial), dimerizes and autophosphorylates at a conserved histidine
residue (Fig. 15.1). This action leads to a phosphoryl relay from the modified histi-
dine residue to a conserved aspartate residue on a transcription factor called a
response regulator [2, 5]. Although exceptions to this canonical mechanism exist
[12], in general the response regulator then changes conformation to allow its DNA-
binding domain to access specific cis elements within bacterial promoters, which
ultimately leads to transcriptional initiation and an ensuing targeted response to the
threatening antimicrobial. TCS systems usually respond to a specific stimulus, and
they phosphorylate only one target: their cognate response regulator (Fig. 15.1).
Both the histidine kinase and its cognate regulator target are most often found as
co-transcribed operons in bacterial genomes; however exceptions to this rule exist
[12]. The number of TCS systems present in bacteria can range from one to over a
hundred [5].
In recent years, additional bacterial signaling systems have been discovered that
parallel eukaryotic signal transduction systems. These include the so-called bacte-
rial eukaryotic-like serine/threonine kinases (eSTKs) [2, 6], phosphatases (eSTPs),
and tyrosine kinases [13, 14] (Fig. 15.1). In the case of eSTKs, which are most rel-
evant to our discussion in this chapter, the domain architecture usually consists of
an N-terminal intracellular kinase domain, which strongly resembles the eukaryotic
versions, followed by a single transmembrane segment and an extracellular/peri-
plasmic sensory domain that is unique to the microbial systems. Interestingly,
atomic structure comparisons indicate that eSTK kinase domains have a structural
15 Bacterial Signal Transduction Systems in Antimicrobial Resistance 463

Environmental Environmental
Stimuli Stimuli

Tyr
Kinase
HK H eSTK ST ST Y

antibiotics

D eSTP ST Y
REC ST SD

ED ED

Transcriptional response

Fig. 15.1 Bacterial signaling basics. Classical two-component signaling (TCS) involves an initial
sensory input, usually detected from outside the bacterial cell by a membrane-bound histidine
kinase (HK). This signal results in autophosphorylation of the HK on a conserved histidine (H)
residue, whose phosphate group is then transferred to a conserved aspartate residue (D) on a
receiver (REC) domain. This posttranslational modification (PTM) then results in conserved struc-
tural rearrangements that ultimately activate the effector domain (ED) of the receiver, which can
take many forms including DNA binding (most common). Eukaryotic-like serine-threonine
kinases (eSTKs), their cognate phosphatases (eSTPs), and tyrosine (Tyr) kinases can also phos-
pho-regulate TCS systems or one-component systems, the latter which have been shown to directly
bind antibiotics in many cases through a sensory domain (SD). The end result is the sensory input
initiates a transcriptional response to the external threat

fold that is almost identical to their eukaryotic homologs [6, 15], suggesting that the
mechanism by which they act is evolutionarily conserved between kingdoms.
Especially prominent, and relevant to our discussion, are the so-called PASTA (pen-
icillin-binding protein and serine/threonine kinase associated) eSTKs found in
Gram-positive pathogens. PASTA eSTKs contain repeat PASTA domains that are
suspected to directly interact with the peptidoglycan and/or free peptidoglycan frag-
ments. They serve as crucial regulatory players in bacterial cell division, cellular
stress responses and infection [6, 15]. In general, Gram-positive bacteria have only
one eSTK/eSTP pair, or a single tyrosine (Tyr) kinase, although some species have
many more. For example, Mycobacterium tuberculosis possesses 11 eSTKs (but
only a single eSTP [6]). Another difference from the two-component systems is the
pleiotropic nature of the targets for eSTKs and Tyr kinases, where many substrates
have been identified, including transcription factors ([6, 14, 15]; Fig. 15.1).
Surprisingly, these eukaryotic-like signaling proteins in bacteria also directly regu-
late TCS proteins [16]. In addition to the multitude of other signaling proteins
involved in bacterial AMR sensory systems, this observation suggests a complexity
464 A. T. Ulijasz et al.

in bacterial signaling that has been underappreciated. Understanding such complex-

ities should lend insight into how bacteria use their signal transduction systems to
avoid treatment effectiveness and host immunity assaults. Thus, research in this area
has the potential to generate novel therapeutic modalities.
This chapter focuses on these bacterial signaling systems and how they play a
pivotal role in AMR. Examples of specific signaling systems that lead to resistance
will first be discussed, followed by signaling systems that respond to host cell anti-
microbial peptides (AMPs), systems involved in biofilm regulation, and finally
mechanisms of bacterial “persistence.” It should be noted that these topics are not
mutually exclusive, and therefore they overlap. For example, bacterial persisters are
a large component of biofilms and greatly contribute to the tolerance of biofilms to
antimicrobial treatment [17].

15.2 Examples of Signaling and Antibiotic Resistance

15.2.1 Signaling Mechanisms of Vancomycin Resistance

One of the first examples of inducible signal transduction-regulated AMR emerged

from vancomycin-resistant enterococci isolates that were obtained in 1988 in
Europe, 34 years after this antibiotic’s introduction [18]. Vancomycin-resistant and
vancomycin-tolerant isolates were subsequently found in the USA and, unfortu-
nately, have now spread throughout the world to include diverse microbial species
[18]. Vancomycin is a “drug of last resort” and is on the World Health Organization’s
list of essential medicines, as it is used to treat Gram-positive infections that are
unresponsive to other treatments, in particular methicillin-resistant Staphylococcus
aureus (MRSA).
Vancomycin was isolated from the soil-dwelling Actinomycete: Amycolatopsis
orientalis [18]. The drug targets the bacterial cell wall, binding with high affinity to
the D-Ala-D-Ala portion of the pentapeptide component, effectively preventing
transglycosylation to the nascent peptidoglycan chain. This binding results in a sub-
sequent blocking of transpeptidation cross-linking [19]. Resistance to vancomycin
is due to replacement of the D-Ala-D-Ala vancomycin target with D-Ala-D-Lac or
D-Ala-D-Ser, both of which exhibit an approximately 1000-fold lower affinity for
the antibiotic. Several different variations of aminoglycoside resistance are con-
ferred by this same mechanism, depending on the precise microbe and genes
involved (VanA-, VanB-, VanD-, VanE-, and VanG-type resistance).

15.2.1.1 Enterococcus sp.

The classic example of inducible vancomycin resistance is observed with the entero-
cocci (e.g., E. faecium and E. faecalis), where it was originally observed in the late
1980s [18]. Among the enterococci, VanA-type strains display high levels of
15 Bacterial Signal Transduction Systems in Antimicrobial Resistance 465

Capsule

muropeptides?

PASTA

PASTA

PASTA
Cell wall

PASTA

PASTA

PASTA
PASTA

PASTA

PASTA
vancomycin lipid II LL-37 and indolicidin ?

PASTA

PASTA

PASTA
Membrane

VanS WalK GraS VraS

eSTK eSTK eSTK

H H ST ST eSTP H eSTP ST ST H eSTP ST ST
Cytoplasm

D D D D

REC REC GraX REC REC CcpA SarA MgrA

VanR WalR GraR VraR
ST
ST eSTP DBD
ST ST ST
DBD DBD DBD DBD DBD DBD

CAMP efflux
Sythesis and cell wall VraF VraG cell wall simulon
Chromosome

-Reduced autolysis
incorporation of D-ala-D-lac -Increase in cell wall thickness
D-alanylation of cell wall teichoic acid murein hydrolases
vanH vanA vanX cell wall regulatory operon dlt operon

Lys additon to PG
MprF
central metabolism
enterococci/
Actinomycetes staphylococci

Fig. 15.2 Vancomycin resistance and tolerance signaling mechanisms in the enterococci and S.
aureus. In the enterococci, a single two-component signaling (TCS) system conveys vancomycin
resistance through the vanHAX operon. This histidine kinase VanS has been shown to likely bind
vancomycin directly to initiate a D-Ala-D-lac modification of the cell wall pentapeptide. In S.
aureus there are at least three TCS systems responsible for the VISA phenotype, WalKR, GraSR,
and VraSR, all of which are controlled through the S. aureus serine-threonine kinase Stk1 (an
eSTK). The serine-threonine phosphatase (eSTP) acts on Stk1 and its substrates for counter-regu-
lation. The histidine kinase WalK responds through contact with lipid II, and GraS binds and
responds to the human CAMP LL-37. The VraS ligand remains unknown. The ligand for Stk1
is still under debate, but is suspected to be cell wall muropeptides. Dotted lines represent direct
regulation by phosphorylation; dashed lines represent regulatory interactions with GraX. ST, ser-
ine-threonine phosphorylation; D, aspartate phosphorylation; H, histidine phosphorylation; REC,
receiver domain; DBD, DNA-binding domain; PASA, penicillin-binding and serine-threonine
kinase associated domain

inducible resistance to both vancomycin and teicoplanin aminoglycosides, whereas

VanB-type strains are only resistant to vancomycin [19]. In the more common
VanA-type strains, the resistance operon generally contains three core genes
required for peptidoglycan modification and resistance, namely, vanH, vanA, and
vanX. The VanH protein is a dehydrogenase that reduces cellular pyruvate to D-Lac,
VanA is a ligase that catalyzes the formation of the new D-Ala-D-Lac dipeptide that
is incorporated into the cell wall, and VanX is a dipeptidase that assists in eliminat-
ing the wild-type D-Ala-D-Ala dipeptide pool by cleaving the D-Ala-D-Ala peptide
bond. Accessory proteins VanY and VanZ participate in the hydrolysis of the termi-
nal pentapeptide D-Ala residue. VanZ confers resistance to teicoplanin by an as-of-
yet unknown mechanism ([19]; Fig. 15.2).
466 A. T. Ulijasz et al.

Adjacent to the vanHAXYZ operon is the TCS pair VanRS, which controls the
synthesis of vanHAXYZ and therefore renders vancomycin resistance inducible
[19]. Upon addition of vancomycin, the histidine kinase VanS phosphorylates the
response regulator VanR. VanR then binds the vanHAXYZ promoter, inducing the
operon, thereby conferring resistance to the drug [20]. Deletion of VanS results in a
constitutively active phenotype [19]. To date, the precise mechanism for how van-
comycin interacts with the Gram-positive versions of VanS to induce resistance has
remained enigmatic. However, in recent years progress has been made with the
study of the VanS histidine kinase-driven system from Actinomycetes, the bacteria
that naturally produce vancomycin and therefore require their own inducible resis-
tance mechanism when manufacturing the antibiotic. Wright and colleagues were
the first to show that the Actinomycete version of VanS directly binds vancomycin
[21]. The antibiotic binds within the first 41 N-terminal residues of the protein,
which comprise the first transmembrane region plus a predicted short extracellular
sensory peptide containing the sequence DQGW. Vancomycin is predicted to inter-
act with this peptide based on biochemical data [21] (Fig. 15.3). These data hold
true for Actinomycete versions of VanS, which directly bind vancomycin and confer
intrinsic resistance (i.e., resistance to vancomycin by these native producers [21]).
On the other hand, the mechanism by which the enterococci, staphylococci, and
Gram-positive pathogens in general sense the presence of vancomycin through the
VanS histidine kinase receptor has yet to be determined. The primary amino acid
sequences of the Actinomycete VanS are divergent enough from the enterococcal
and staphylococcal versions to suggest a possible alternative mechanism (Fig. 15.3)
that might include more indirect signaling mechanisms, such as interactions with
cell envelope components. However, a recent study indicates that vancomycin and
teicoplanin might bind directly to the receptor at aromatic residues, similar to what
was observed with Actinomycetes [22]. Indeed, an alignment of VanS receptors
shows that some aromatic residues, such as the tryptophan residue proposed to
directly interact with vancomycin [21], are conserved among VanS versions
(Fig. 15.3). An interesting topic of future research would be to resolve exactly how
VanS from pathogenic Gram-positive bacteria sense the presence of vancomycin.

15.2.1.2 Staphylococcus aureus

Although it is clear how vancomycin resistance in both the Actinomycetes and

pathogenic enterococci rely on the specific vanRS/vanHAX operons, the signaling
cascade and ensuing mechanism of resistance in S. aureus and other staphylococci
appear more diverse and are therefore far less clear [23]. In rare cases, the
Enterococcus VanA genes have been found within vancomycin-resistant S. aureus
strains. These S. aureus strains, which have presumably acquired the enterococcal
resistance by horizontal transfer, produce the high levels of resistance observed in
vancomycin-resistant Enterococcus strains; they are collectively referred to as high-
level vancomycin-resistant Staphylococcus aureus, or VRSA strains. A more com-
mon form of S. aureus vancomycin resistance seen in the clinic results from acquired
15 Bacterial Signal Transduction Systems in Antimicrobial Resistance 467

TM1 Vancomycin loop

VanS(sc) .....MDRRPGLSVRLKLTLSYAGFLTLAGVLLLVAVGVFLLDQG.. W LLTNERGAVRAT
VanS(ef ) MVIKLKNKKNDYSKLERKLYMYIVAIVVVAIVFVLYIRSMIRGKLGD WILSILENKYDLN
consensus

TM2
VanS(sc) PGTVFLRSFAPTAAWVMAFLLVFGLVGGWFLAGRMLAPLDRITEATRTAATGSLS...HR
VanS(ef ) HLDAMKLYQYSIRNNIDIFIYVAIVISILILCRVMLSKFAKYFDEINTGIDVLIQNEDKQ
consensus

VanS(sc) IRLPGRRDEYRELADAFDEMLAR..LEAHVAEQRR..FAANASHELRTPLAVSK...AIL
VanS(ef ) IELSAEMDVMEQKLNTLKRTLEKREQDAKLAEQRKNDVVMYLAHDIKTPLTSIIGYLSLL
consensus

VanS(sc) DVARTDPHQDPGEIIDRLHAVNTRAIDLTEALLLLSRAGQR..SFTREQVDMSLLAEEAT
VanS(ef ) DEAPDMPVDQKAKYVHITLDKAYRLEQLIDEFFEITRYNLQTITLTKTHIDLYYMLVQMT
consensus

VanS(sc) ETLLPFAEKHGVTLETRGHVTLALG.SPALLLQLTTNLVHNAIVHNLPGRGRVWIHTAAG
VanS(ef ) DEFYPQLSAHGKQAVIHAPEDLTVSGDPDKLARVFNNILKNAAAYSE.DNSIIDITAGLS
consensus

HATPase domain
VanS(sc) PHTTRLVVENTGDLISPHQASTLPEPFQRGTERIHTDHPGVGLGLAIVNTITQAHDGTLT
VanS(ef ) GDVVSIEFKNTG.SIPKDKLAAIFEKFYRLDNARSSDTGGAGLGLAIAKEIIVQHGGQIY
consensus

VanS(sc) LTPRHSGGLRVTVELPAAAPHTGR...
VanS(ef ) AESNDN.YTTFRVELPAMPDLVDKRRS
consensus

Fig. 15.3 Alignment of VanS from Streptomyces coelicolor (VanS(sc)) and VanS from
Enterococcus faecium (VanS(ef)). Yellow bars indicate transmembrane (TM) regions, the green
bar the vancomycin sensory loop, and the red bar the histidine kinase ATPase domain. The arrow
indicates the phosphorylated histidine. The purple highlighted area indicates the residues (DQGW)
proposed to bind vancomycin directly in VanS(sc). Black-colored residues are identical and gray
similar. For consensus,! is identical residues and * similar

mutations that lead to in an intermediate resistance phenotype. These strains are

commonly referred to as vancomycin-intermediate Staphylococcus aureus, or VISA
[8, 23] (see Chap. 9 for additional discussion of VISA). The VISA mechanism
appears to involve decreased cell wall cross-linking and, importantly, a thicker cell
wall that is thought to act as a mechanical barrier to vancomycin uptake by acting as
a “vancomycin sponge” (a phenomenon sometimes referred to as the “cell wall
clogging effect”). Genomic comparisons between VRSA and VISA strains indicate
that many genes are involved in producing VISA traits, although only a handful of
specific contributions have been experimentally verified (see Refs [8] and [23], and
references therein). Not surprisingly, many of these mutations occur in signal trans-
duction proteins that control the intermediate-resistance phenotype.
468 A. T. Ulijasz et al.

Signal transduction system genes that produce S. aureus VISA phenotypes

include three TCS pairs (VraSR, GraSR, WalKR) plus the S. aureus eSTK/eSTP
signaling pair [6, 8] (Fig. 15.2). Although complex, it is now becoming more clear
how these signaling proteins might act to reduce susceptibility. The VraS (histidine
kinase) and VraR (response regulator) pair regulate a cell wall synthesis stimulon in
S. aureus. Mutations within the VraSR TCS pair, found in a clinical isolate, corre-
late with an overactive or constitutively active cell wall synthesis phenotype [23].
However, VraSR mutations alone do not always confer VISA – only after additional
mutations arise in the GraSR TCS pair is a VISA phenotype likely to emerge [24].
The GraSR system is responsible for host defense cationic antimicrobial peptide
(CAMP) resistance [25]. Upon stimulation by direct binding of host CAMP LL-37,
the GraS histidine kinase phosphorylates the GraR response regulator transcription
factor. The phosphorylated GraR then initiates transcription of GraX and an ABC
transporter called VraFG. Together, these five proteins change the net surface charge
of the microbe, which results in repulsion of the CAMP [25]. Alone, mutations in
the GraSR system cause only a marginal effect on susceptibility to the antibiotic
[23, 24] (Fig. 15.2).
A third TCS system that contributes to the VISA phenotype is the WalKR TCS
cognate pair (also referred to as YycGF). The WalK histidine kinase and WalR
response regulator are rare among the TCS family of signaling proteins as they are
essential for bacterial viability. Interestingly, their essential nature is conserved in
many other Gram-positive pathogens, as is their important function in controlling
the regulatory connection between cell wall biosynthesis and cell division (for
review see Ref. [26] and references therein). Mutations resulting in downregulation
of the WalKR function result in a marked decrease in vancomycin effectiveness,
whereas mutations resulting in WalKR upregulation result in cells becoming more
susceptible to treatment [8, 23]. Unlike the VraSR and GraSR systems, it appears
that WalKR mutations can autonomously control vancomycin susceptibility in S.
aureus [23].
Mutations within the S. aureus eSTK (also referred to as PknB or Stk1) and eSTP
(also called Stp1) pair also contribute to the VISA phenotype (Fig. 15.2). Indeed,
VraR is a substrate for the S. aureus eSTK, such that phosphorylation negatively
impacts DNA binding and thus VraR activity [27]. Furthermore, eSTK phosphory-
lates several other transcription factors associated with vancomycin susceptibility,
including catabolite control protein A (CcpA) and the murine hydrolase regulators
SarA and MgrA [28] (Fig. 15.2). Each of these transcription factors has been associ-
ated with a loss of susceptibility to vancomycin [6]. Interestingly, the S. aureus
eSTK (Stk1) directly regulates SarA and MgrA by phosphorylating a cysteine resi-
due on each. This observation represents the first example of a cysteine posttransla-
tional modification having a clear regulatory role [28]. Finally, deletion of the S.
aureus eSTP (Stp1), which would result in a constitutive eSTK phosphorylation of
the GraR, VraR, CcpA, and SarA transcription factors, produces the thickened cell
wall phenotype associated with vancomycin and teicoplanin intermediate resistance
[29]. Although not yet demonstrated in S. aureus, it has recently been shown that an
eSTK directly phosphorylates the WalK histidine kinase to regulate antibiotic sus-
15 Bacterial Signal Transduction Systems in Antimicrobial Resistance 469

ceptibility [16]. This novel discovery demonstrates that eSTKs and TCS can work
together to co-regulate essential cellular processes and antibiotic susceptibility, and
opens new opportunities and strategies for bacterial signaling perturbation to facili-
tate novel antibiotic treatments.

15.2.1.3 Streptococcus pneumoniae

S. pneumoniae is the most common cause of community-acquired bacterial

­pneumonia and a major cause of many other serious infections, such as meningitis
and otitis media [30]. Despite an available vaccine, pneumococcal infections con-
tinue to be a problem, as the immunization covers only a small subset of strains (see
Chap. 2). After introduction of a pneumococcal vaccine, an epidemiology shift
occurs away from the vaccine serotypes. This “vaccine escape” phenomenon, com-
bined with an increase in antibiotic-resistant strains, has placed S. pneumoniae,
along with Enterococcus and Staphylococcus species, on the list of the Center for
Disease Control’s (CDC) “biggest threats” to public health due to AMR [10].
Although vancomycin resistance in S. pneumoniae has yet to be described, tolerant
clinical isolates have been reported.
S. pneumoniae possesses 13 TCS pairs and 1 orphan response regulator [31] (i.e.,
a TCS response regulator that lacks a known cognate kinase). The VncRS TCS sys-
tem was initially reported to be responsible for vancomycin tolerance in S. pneu-
moniae [32]. However, with the exception of the ABC transporter (Vex123) that was
described as the VncRS target-regulated gene in the original report [33], these find-
ings were later proven to be artifactual [34]. More recently it was found by another
group that the CiaHR TCS pair helps confer the vancomycin tolerance phenotype,
although CiaHR is better known for controlling beta-lactam resistance in this patho-
gen (see below) [35]. Surprisingly, only a single mutation within the CiaH histidine
kinase gene (within Ser198) is enough to convey tolerance, but only in the presence
of the pneumococcal polysaccharide capsule and in the absence of the pneumococ-
cal autolysin LytA [36]. Another recent report implicates a PadR family transcrip-
tion factor called PtrR and its vancomycin-inducible control over a four-gene operon
(ptvABC) in vancomycin tolerance. The ptvABC operon reportedly encodes mem-
brane-associated proteins whose precise function remains enigmatic [37]. The
mechanism with which these seemingly unrelated transporters, signaling proteins,
and the capsule components act to confer loss of susceptibility to vancomycin in S.
pneumoniae remains to be deciphered.

15.2.2 Signaling Mechanisms of Beta-Lactam Resistance

Beta-lactams, which are among the most widely prescribed antibiotics, were the
first to be discovered (in the 1920s by Alexander Fleming at St. Mary’s Hospital,
now part of Imperial College London) [38]. Members of this antibiotic class target
470 A. T. Ulijasz et al.

the aptly named penicillin-binding proteins (PBPs), enzymes that are responsible
for catalyzing the later steps in the assembly of the bacterial cell wall. When applied,
the beta-lactams can cause malformation of the cell wall, cell lysis, and death. Due
to the early introduction and overuse of beta-lactams, especially with Gram-positive
pathogens, resistance is now widespread. As with many cell envelope-acting antibi-
otics, beta-lactam susceptibility is controlled through bacterial signaling proteins.

15.2.2.1 Streptococcus pneumoniae

A classic example of a signal transduction system controlling beta-lactam resistance

is found with the S. pneumoniae CiaRH TCS pair, which has also been implicated
in competence (DNA uptake) regulation and cell lysis [35]. This TCS system moni-
tors cell envelope integrity; however, the mechanism by which this occurs, includ-
ing the exact inducer/molecule sensed by the CiaH kinase, remains unknown. What
is understood is that CiaRH responds to a diverse array of beta-lactams, including
advanced generation derivatives, such as cefotaxime and the Gram-negative-acting
piperacillin. Using these antibiotics, Hackenbeck and colleagues have isolated
resistant laboratory mutants [39]. Additionally, the genomes of several clinical iso-
lates have been sequenced and the influence of the mutations on beta-lactam resis-
tance and CiaRH promoter-regulated activity assessed (11 mutations in total [39]).
Data from these studies demonstrated that the mutations generally result in an
increase in CiaRH activity and ensuing regulatory gene expression, which ulti-
mately leads to a loss of beta-lactam susceptibility. A more in-depth survey of 3085
Thai and 616 US pneumococcal isolates as independent datasets demonstrated a
high mutation frequency within the ciaH kinase allele, resulting in its hyperactiva-
tion and ensuing beta-lactam resistance phenotype [40]. Combined with previous
data, these findings indicate that signal transduction genes can mutate extensively
during treatment, presumably to rapidly accommodate new environmental condi-
tions and avoid the effects of antibiotics.

15.2.2.2 Staphylococcus aureus

S. aureus presents one of the most serious problems with respect to beta-lactam-
resistant hospital infections [10]. After the introduction of penicillin and the ensuing
development of resistance provided by penicillinase, a semisynthetic penicillinase-
resistant beta-lactam called methicillin was introduced in 1959. Within a year resis-
tant strains emerged, and MRSA was born as a problematic nosocomial pathogen
[41]. Although methicillin is no longer in clinical use, MRSA strains of S. aureus
continue to become resistant to new beta-lactams. Resistance is conferred by intro-
duction of an alternative PBP target (called PBP2a), which lowers the affinity of the
antibiotic. PBP2a is encoded by the mecA gene, which is in a mobile genetic ele-
ment that is referred to as SCCmec. MecA is under control of an inducible signaling
system, as depicted in Fig. 15.4.
 15 Bacterial Signal Transduction Systems in Antimicrobial Resistance 471

BL

BL
BL
Cell wall

BL
MecA BL
BL
BL

MecA BL BL
BL
MecA BL
BL
PBD
MecR1
Membrane

MecI
Cytoplasm

DBD

MecR2
MecI MecI
Chromosome

DBD DBD

SCCmec mecA mecR1 mecI mecR2

Fig. 15.4 MecA (PBP2a)-mediated regulation in S. aureus. MecI is a transcription factor that
binds to the promoter of mecA (PBP2a) and represses its transcription until it is degraded via
MecR1-mediated proteolysis. MecR1 contains a penicillin-binding domain (PBD) that senses the
beta-lactam (BL, green octagons). The binding of cell wall fragments (specifically g-D-Glu-L-Lys)
and MecR2 to MecI destabilizes its binding to DNA and enhances MecR1-mediated degradation.
Once mecA repression is remediated, MecA is produced and binds the beta-lactam to lessen drug
susceptibility

In contrast to the inducible resistance pathways discussed above, beta-lactam

resistance in MRSA is induced via a proteolytic signaling system encoded within
SCCmec. This resistance mechanism parallels those associated with the “bla” sys-
tems found in wild-type S. aureus and Bacillus licheniformis. In MRSA, the sig-
naling is carried out by an integral membrane receptor, MecR1, and its intracellular
relay target, the winged helix-turn-helix transcription factor MecI. MecR1 senses
the presence of beta-lactams though its extracellular penicillin-binding domain
(PBD). Intriguingly, upon activation though drug binding, MecR1 acts as a
Zn-dependent metalloendopeptidase zymogen, autocatalytically cleaving its
intracellular signaling domain, which then, directly or indirectly, binds/inactivates
by cleavage of the MecI transcription factor. This action then results in the induc-
tion of MecA and resistance (Fig. 15.4). Interestingly, cell wall fragments, specifi-
472 A. T. Ulijasz et al.

cally γ-D-Glu-L-Lys, bind the C-terminal domain of the MecI repressor and act as
a cofactor that mediates MecI cleavage and inactivation by MecR1 (Fig. 15.4).
Once produced, the mecA (PBP2a) gene product interacts with beta-lactams.
Other reports show that MecA is more complicated, as it is apparently also con-
trolled allosterically through binding the D-Ala-D-Ala terminus of the pentapep-
tide stem, likely generated by the inhibition of cell wall synthesis by beta-lactams.
The allosteric site requires occupation to then unveil the active binding pocket for
the beta-lactam [42, 43].

15.2.2.3 Enterococcus sp.

The Gram-positive opportunistic nosocomial pathogen Enterococcus faecalis is a

major cause of urinary tract infections and endocarditis, and it has now been widely
associated with inflammatory bowel disease [44] and more recently with colon can-
cer [45]. E. faecalis and other enterococci are unfortunately naturally resistant to the
cephalosporin beta-lactam antibiotics and many other antimicrobials including bile,
a natural human-produced antimicrobial [46]. In 2007 it was determined that resis-
tance is largely conferred by the sole eSTK in this organism [46].
Kristich et al. found that the E. faecalis eSTK (now renamed IreK) is responsible
for controlling both antimicrobial resistance and intestinal persistence in this patho-
gen. An IreK mutant exhibits increased susceptibility to cephalosporins, as well as
to bile; the mutation rendered E. faecalis unable to colonize the mouse intestine
[46]. Although such a pronounced effect as seen with cephalosporins was not
observed, resistance through IreK was also extended to other cell envelope-acting
antibiotics, including ampicillin, bacitracin, and vancomycin. These data suggest a
broad-based sensory mechanism. This broad specificity is likely due to unlinked
peptidoglycan fragments that result from antibiotic-induced cell wall disruption
[47]. Later work by the Kristich lab showed that IreK modulates its antimicrobial
activity through phosphoryl regulation of a small protein substrate, IreB, and the
IreK cognate eSTP, IreP [48]. Interestingly, identification of IreB in many other
Gram-positive organisms having low GC content in their DNA suggests that this
signaling system may be conserved outside of enterococci species [48].
Another two-component system, CroRS, was also found to be required for intrin-
sic beta-lactam resistance in E. faecalis [49]. A deletion of the CroRS system pres-
ents a dramatic phenotype, with a 4000-fold reduction in the MIC for the new,
third-generation cephalosporin ceftriaxone. This MIC reduction is facilitated by the
induction of PBP5 [47]. As with the IreK signaling system, resistance to structurally
unrelated cell wall-targeting antibiotics bacitracin and vancomycin was also
observed in the CroRS mutant. An expanded study determined that another TCS
system, CisRS, could compensate for the loss of CroRS signaling. Interestingly, this
system also compensates for the absence of the VanG-type resistance system in E.
faecalis, suggesting that CisRS might act to compensate as a “surrogate” TCS cell
wall stress response system [50]. The molecules that bind and activate the CroRS
15 Bacterial Signal Transduction Systems in Antimicrobial Resistance 473

and CisRS systems and the molecular mechanism underlying antibiotic resistance
remain to be deciphered. Based on the differing targets and structures of the antibi-
otics that CroRS/CisRS respond to, it appears that the mechanism may involve gen-
eral sensing of cell envelope stresses.

15.2.2.4 Gram-Negative Pathogens

Although beta-lactam resistance is widespread in Gram-negative bacteria, unlike

Gram-positive microbes, the focus has generally been on the mechanism of the
beta-lactam-degrading enzymes (beta-lactamases) that are responsible for resis-
tance, rather than on the sensory systems involved. Nevertheless, like many antibi-
otic resistance mechanisms, beta-lactam resistance in Gram-negative bacteria is
also controlled by signal transduction systems. In this section we discuss two of
these signaling cascades.
In the Enterobacteriaceae family, the beta-lactam resistance determinant AmpC
(a beta-lactamase [51]) is induced by the AmpG-AmpR (AmpRG) TCS pair. As a
consequence of a beta-lactam being present, anhydrous N-acetylglucosamine-N-
acetylmuramic acid (GlcNAc-MurNAc) oligopeptides from peptidoglycan accumu-
late in the periplasm. These oligopeptides are then transported into the cytoplasm by
the inner membrane-associated AmpG transporter. Once in the cytoplasm, the
GlcNAc moiety is removed by the glycoside hydrolase NagZ, liberating anhydro-
MurNAc. Anhydro-MurNAc then binds the LysR-type transcription factor AmpR,
resulting in a conformational change that enables promoter binding and activation
of AmpC transcription (Ref. [52] and references therein). As production of AmpC
taxes the overall energy requirements of the cell, another protein, the
N-acetylmuramoyl-L-alanine amidase (AmpD), mitigates AmpC synthesis by
cleaving the muropeptides and reducing accumulation of the inducing anhydro-
MurNAc molecule.
Variations of the AmpG-AmpR-AmpC resistance theme are also found within
other Gram-negative bacteria. For example, E. coli and Shigella species, as well as
Acinetobacter baumannii, lack AmpR, which results in a low level of constitutive
AmpC production. Introduction of a heterologous AmpR regulator into these
AmpR-deficient species results in a reinstatement of the inducible system. In
another example of AmpC operon divergence, the problematic cystic fibrosis patho-
gen P. aeruginosa genome harbors three redundant copies of AmpD, which results
in a hyper suppression of AmpC transcription/translation. As successive AmpD cop-
ies are mutated during, for example, chronic infection with P. aeruginosa, beta-
lactam susceptibility drops in a stepwise fashion [52].
In addition to the AmpG-AmpR system, TCS systems have also been implicated
in AmpC induction by the BlrAB signaling cascade. The BlrAB histidine kinase
response regulator phosphoryl relay has been studied in Aeromonas species, which
are facultative anaerobes that cause a variety of human diseases (Ref. [52] and refer-
ences therein). In Aeromonas, overexpression of the BlrB histidine kinase results in
a marked increase in AmpC, presumably due to enhanced phosphorylation of the
474 A. T. Ulijasz et al.

BlrA response regulator. Standard bioinformatics analysis reveals that the closest
homolog to BlrAB is the CreBC TCS system. Interestingly, the Aeromonas CreBC
TCS, when introduced into in a heterologous host (E. coli), regulates the E. coli
beta-lactamases. This regulation occurs through the E. coli CreC response regulator
binding a conserved CreC-binding motif (TTCACnnnnnnTTCAC), which activates
gene expression. A recent report confirmed this same binding motif for CreC in
Aeromonas, and it importantly demonstrated that the CreBC system is specifically
responsive to inhibition of PBP4 by beta-lactams [53]. The authors hypothesize that
the BlrAB and CreBC TCS systems respond to peptidoglycan recycling and there-
fore levels of AmpC-inducing muropeptides that control beta-lactam susceptibility
[53]. Future work will need to elucidate the precise mechanisms underlying the
BlrAB and CreBC systems and determine whether their roles are ubiquitous among
Gram-negative pathogens.

15.2.3  ignaling Mechanisms of Polymyxin Resistance

S
in Gram-Negative Bacteria

Polymyxins are antimicrobials classified as cyclic peptides that contain a hydropho-

bic tail; the two classic examples are polymyxin B and colistin [54]. They are pre-
dominantly used against Gram-negative infections, but sometimes they are
administered in combination with other antibiotics to treat Gram-positive infec-
tions. Because this antibiotic class was originally found to be both neuro- and neph-
rotoxic, their use dwindled in the wake of newer, less harmful choices (e.g.,
beta-lactams and aminoglycosides). A further complication for their clinical use is
that polymyxins are not absorbed in the gastrointestinal tract, and they must there-
fore be administered intravenously, by inhalation or by application topically for skin
infections, for example. However, the recent rise of untreatable infections due to
multidrug resistance has led to reintroduction of the polymyxins as a “new” drug of
last resort.
Unlike the beta-lactams and aminoglycosides, the polymyxins target the bacte-
rial cell membrane. There they bind to the lipopolysaccharide (LPS) through their
cyclic peptide moiety and then disrupt both inner and outer membranes of the
Gram-negative cell envelope. This mechanism of action is facilitated by the hydro-
phobic “tail” of the polymyxins, which is suggestive of detergent-like qualities [55].
Unfortunately, with the increase in polymyxin use has come the emergence of
resistance. In general, a common theme among the polymyxins is that they work by
ultimately changing the charge, or electrostatic repulsion properties, of the LPS
such that the initial (and required for efficacy) cyclic peptide binding is blocked.
This is accomplished by substituting moieties, such as 4-amino-4-deoxy-L-arabi-
nose (L-Ara-4N), phosphoethanolamine (pEtN), or galactosamine enzymatically
into Lipid A or the LPS core (Fig. 15.5). In some cases, the LPS is simply lost [54].
15 Bacterial Signal Transduction Systems in Antimicrobial Resistance 475

Host CAMPs,
Polymyxins, pEtN,
other antibiotics L-Ara4N,
galactosamine
Lipid A Core O-Antigen
LPS
Cell wall Membrane
Outer

CAMPs CAMPs Polymyxin

low Mg2+ high Fe3+ Zn2+ indolicidin
Colistin CAMPs
Membrane

low pH
CcrB PhoQ PmrB ColS ParS CprS indolicidin
Inner

MgrB

H H H H H H
Cytoplasm

D CcrA D PhoP D PmrA D ColR D ParR D CprR

REC micA REC REC REC REC REC
CcrC
DBD DBD PmrD DBD DBD DBD DBD

pEtN
ccrC pmrD eptA oprD
Chromosome

mgrR L-Ara4N
pEtN (E. coli) arn operon
mexX mexY
eptB
galactosamine
(A. baumannii) pEtN (E. coli)
eptA
naxD
pEtN (K. pneumoniae, A. baumannii)
mgrB pmrC pmrA pmrB

Fig. 15.5 The complex signaling pathways of Gram-negative polymyxin and CAMP resistance.
TCS systems shown in gray (ColR, ParR, and CprR) are found in P. aeruginosa. The CcrABC
system and MgrB are found in K. pneumoniae. Small regulatory RNAs micA and mgrR are found
in E. coli. Sensory inputs vary depending on the pathogen. Here we show some common histidine
kinase stimulants (e.g., Mg2+, Fe3+) among the Gram-negative pathogens. For a complete list, refer
to the text and references therein. Phosphoethanolamine (pEtN), galactosamine, and 4-amino-
4-deoxy-L-arabinose (L-Ara-4N) molecules are all added to Lipid A of the LPS to modify the
outer membrane charge. A. baumannii lacks a L-Ara-4N-modifying system and instead has galac-
tosamine and pEtN enzymes. The mexXY/oprD multidrug resistance efflux system is found in P.
aeruginosa

Since resistance costs valuable energy, signal transduction networks have been
adapted by bacteria to control the antibiotic response on an as-needed basis.

15.2.3.1 Enterobacteriaceae

With the Enterobacteriaceae species, Salmonella species (Salmonella), E. coli, and

Klebsiella pneumoniae, polymyxin resistance is controlled via two main TCS sys-
tems: PhoPQ and PmrAB (for comprehensive reviews see Ref. [54] and Ref. [56]).
Induction of these systems has been investigated extensively, especially in
476 A. T. Ulijasz et al.

Salmonella, the pathogen species in which they were originally discovered [56–58].
Although much is still not understood about the complexities of how polymyxin
resistance arises, resistance appears to be induced nonspecifically by the presence
of cationic compounds through the action of TCS networks. Inducing agents include
polymyxins, low magnesium (Mg2+) concentrations, high ferric iron (Fe3+) concen-
trations, and acidic pH [54, 56]. In some cases metals such as aluminum and zinc
induce the polymyxin signaling cascade. Specifically, magnesium or cationic drugs
stimulate phosphorylation of the histidine kinase PhoQ by direct interaction with an
acidic patch within its sensory domain. PhoQ then dimerizes and phosphorylates
the intracellular response regulator PhoP, which then modulates transcriptional
activity [59]. In K. pneumoniae, PhoP directly activates the arn operon to initiate a
chemical modification that changes the net charge of LPS. In E. coli and Salmonella
species, signaling is accomplished through an intermediary relay protein PmrD,
which signals between the PhoPQ and a second TCS cognate pair, PmrAB [58]. The
PhoPQ/PmrAB systems have common activators and are also able to respond to
their own individualized signals. Both signaling systems respond to cationic pep-
tides (i.e., polymyxins) and low pH; however, they differ in that the PhoQ histidine
kinase responds to low magnesium, while the PmrB histidine kinase responds to
high ferric iron (Fig. 15.5). Interestingly, in E. coli PmrB responds to other metals,
such as zinc and aluminum [54].
Downstream signaling can differ among Enterobacteriaceae species. However,
it is important to note that independent of species, the end result of PhoPQ/PmrAB
induction is the same phenotypic change: the enzymatic restructuring of the LPS as
a charge switch through upregulation of 4-amino-4-deoxy-L-arabinose (L-Ara-4N),
phosphoethanolamine (pEtN), or galactosamine additions. Nevertheless, there are
slight variations among Enterobacteriaceae that are exploited to achieve this com-
mon goal. For example, in K. pneumoniae, aside from activating LPS-modifying
operons, PhoP also activates synthesis of a membrane protein called MgrB, which
inhibits the PhoP kinase to complete a negative feedback regulatory loop (Fig. 15.5).
To further complicate matters, in K. pneumoniae a third TCS, CrrAB, can respond
to polymyxins by activating the transcription of an intermediary signaling protein,
CrrC, which then activates the PmrB response regulator to induce the PmrC LPS-
modifying enzyme (Fig. 15.5). E. coli PhoPQ/PmrAB signaling differs from K.
pneumoniae regulation by adding another layer of regulatory complexity in which
two small regulatory RNAs, mgrR and micA, are involved in inhibition of the phos-
phoethanolamine LPS-modifying operon and PhoP, respectively (Fig. 15.5).
The genomes of many polymyxin-resistant clinical isolates have been sequenced,
thereby verifying that mutations within the Enterobacteriaceae PhoPQ/PmrAB
TCS relays and associated signaling systems are sufficient to result in polymyxin
resistance. In particular, MgrB has been the subject of several reports, as a mutation
within this gene is sufficient to result in colistin resistance by strongly activating the
PhoPQ signaling system [60]. These single mutations within the K. pneumoniae
mgrB gene have been identified in clinical isolates from globally sampled patients,
indicating that this form of resistance is common and possibly arises from indepen-
dent mutagenic events. In fact, one study reported that over 40% of colistin-resistant
isolates, collected from several countries worldwide, had an mgrB mutation [60]. In
15 Bacterial Signal Transduction Systems in Antimicrobial Resistance 477

addition to MgrB, independent mutations within the PmrB kinase also produce
resistance to colistin. This observation was noted in a study of globally collected
and sequenced K. pneumoniae colistin-resistant genomes in which a single con-
served amino acid change in the PmrB regulator (threonine-157 to proline) resulted
in (up to) a 170-fold induction in the PmrC LPS-modifying enzyme [61].

15.2.3.2 Pseudomonas aeruginosa and Acinetobacter baumannii

As in Enterobacteriaceae, P. aeruginosa activates its polymyxin resistance system

through upregulating an operon that adds the L-Ara-4N moiety to LPS (Fig. 15.5).
However, in this pathogen the signaling that controls this process is complex, with
at least five TCS networks being involved in resistance, including PhoPQ and
PmrAB (Fig. 15.5). In addition to the presence of polymyxins, in P. aeruginosa
PhoPQ and PmrAB can both respond to low magnesium and calcium levels. A third
TCS system, ColRS, responds to zinc as it activates an operon that adds phospho-
ethanolamine groups to the LPS. Lastly, two other systems, ParRS and CprRS,
respond to a variety of polymyxins and cationic peptides (Fig. 15.5; see section
below on cationic antimicrobial peptides).
Clinical isolates of Pseudomonas that exhibit resistance to polymyxin have
mutations within the histidine kinases PmrB (i.e., similar to Enterobacteriaceae),
PhoQ, ParS, or the response regulator ParR that are associated with MICs that range
from 2 mg/L up to 512 mg/L. Higher MICs can be reached when additional muta-
tions accumulate in genes encoding CprS, CprR, ColS, or ColR TCS proteins in a
PhoQ-negative genetic background [62]. These findings suggest that, in many cases,
resistance is a result of the accumulation of mutations within several alleles, rather
than a single mutation in one allele. Interestingly, the ParRS TCS pair also posi-
tively controls the important MexXY multidrug transporter operon (Fig. 15.5). Thus
clinical isolates with mutations in ParRS not only lead to LPS modification but also
result in low-to-moderate loss of susceptibility to a broad spectrum of antimicrobi-
als, including polymyxins, aminoglycosides, fluoroquinolones, and beta-lactams,
contributing to the broad-spectrum resistance/tolerance P. aeruginosa is known for.
In contrast, A. baumannii possesses only a PmrAB TCS system that when activated
by polymyxins can alter the LPS with pEtN, or alternatively with galactosamine,
through the action of the NaxD enzyme [54] (Fig. 15.5). Why some bacterial spe-
cies have evolved more or less complex signaling networks to lower susceptibility
to antimicrobials is an open question.

15.2.4 Daptomycin Resistance in Staphylococcus aureus

Daptomycin (DAP) belongs to the same family of antimicrobials as the polymyxins;

however, unlike polymyxins, DAP is a cyclic lipopeptide that specifically targets
Gram-positive bacteria. Since it received approval from the US Food and Drug
Administration for the treatment of soft tissue infections (2003) and S. aureus
478 A. T. Ulijasz et al.

bacteremia (2006), it has been reserved for the most serious infections caused by
methicillin-resistant S. aureus (MRSA) and vancomycin-resistant E. faecium (VRE)
[63]. The mechanism by which DAP exerts its bactericidal activity is not fully
understood; however, it is thought to bind and insert itself into the Gram-positive
bacterial membrane via a process that is enhanced by the presence of calcium and is
dependent upon interaction with the anionic membrane phospholipid phosphatidyl-
glycerol (PG) [64–67]. How this results in cell death is still unclear, but it may
involve the formation of oligomeric DAP pore-like structures that results in ion
leakage and disruption of membrane potential [66]. Another possibility is the
recently proposed lipid extracting effect in which the accumulation of DAP in the
cell membrane leads to the aggregation and subsequent release of lipid from the cell
membrane [67], both of which would affect permeabilization, metabolism, and cell
division.
Due to its distinct mechanism of action, resistance to DAP is rare. Nevertheless,
treatment failure is a serious concern [68, 69]. Some of the mechanisms that under-
lie DAP resistance (DAP-R) involve enzymes participating in phospholipid metab-
olism and membrane homeostasis (Fig. 15.2) [70]. Regarding signal transduction
genes, the previously mentioned VraSR-regulated cell wall synthesis stimulon and
the essential TCS, WalKR, have been shown to contribute to the DAP-R phenotype
in clinical and laboratory isolates [71, 72] (see section above regarding Vancomycin
resistance/Staphylococcus aureus). VraSR is involved in the regulation of cell wall
biosynthesis via transcription of a number of genes including pbp2 (penicillin-bind-
ing protein 2) [73] and is upregulated in the presence of DAP. Differential gene
expression analysis between DAP-R and DAP-susceptible (DAP-S) clinical iso-
lates reveals an upregulation of the VraSR TCS system. In addition, deletion of
VraSR from a DAP-R isolate conferred a DAP-S phenotype, pointing to the impor-
tance of these signaling systems in producing daptomycin AMR. In S. aureus, the
WalKR TCS system is involved in the control of peptidoglycan biosynthesis and
can influence peptidoglycan turnover, cross-linking, and chain length by sensing
membrane fluidity, likely through lipid II (Fig. 15.2) [74, 75]. Due to the similari-
ties between a DAP-R and a WalKR-deficient phenotype that includes thickened
cell walls, increased membrane fluidity, and resistance to membrane disruption, it
has been proposed that mutations in WalKR may lead to the downregulation of cell
wall homeostasis, which leads to an increase in bacterial survival in the presence of
DAP [70].
The success of S. aureus as a pathogen can be attributed largely to its ability to
produce a wide range of virulence factors and accessory genes, many of which are
under the control of the Agr quorum-sensing system, a classical TCS module (see
also Chap. 14). Importantly, dysfunction of this system has recently been implicated
in a transient defense mechanism that protects against daptomycin activity [76].
Quorum sensing is a form of intercellular communication that enables bacteria to
initiate density-dependent changes in gene expression, allowing populations of bac-
teria to restrict the expression of genes whose resulting phenotypes are most benefi-
cial at high cell densities. Examples are biofilm production, bioluminescence, and
15 Bacterial Signal Transduction Systems in Antimicrobial Resistance 479

virulence factor secretion (for review see Ref. [77]). Quorum sensing involves the
production and secretion of small molecules that are sensed by neighboring cells.
Once the concentration of these small molecules, termed “autoinducers,” exceeds a
certain threshold, population-wide responses are initiated. Broadly speaking, S.
aureus uses the Agr system to coordinate the upregulation of exotoxin expression
and downregulation of surface proteins, such as adhesion molecules, at high cell
densities [78].
The Agr quorum-sensing system is encoded by a four-gene operon (agrBDCA)
and a regulatory RNA, RNAIII, which are expressed from two divergent promoters,
P2 and P3, respectively (Fig. 15.6). AgrA and AgrC comprise a classical TCS sys-
tem. AgrD is the activating ligand (or autoinducer) which is N- and C-terminally
processed in the bacterial cell and then secreted via AgrB into the extracellular
milieu to produce the final autoinducing peptide (AIP) (Fig. 15.6). AIP is sensed by
the transmembrane histidine kinase AgrC, which induces phosphorylation of the
cytoplasmic HPK domain upon AIP binding. This phosphate is transferred to AgrA,
the response regulator TCS transcription factor, thereby activating AgrA and tran-
scription of the two agr promoters, P2 and P3. The Agr system is an example of an
autoactivating system; agr autoactivation leads to an exponential increase in expres-
sion of the two agr promoters. At high cell densities, RNAIII, whose transcription
is under the control of the P3 promoter, is a regulatory molecule that is responsible
for the differential expression of many genes. These large changes in gene expres-
sion inflict a significant metabolic burden upon the cell and have been hypothesized
to partly explain the selective enrichment of Agr-defective mutants among seriously
ill, hospitalized patients [78] (see also Chap. 14).
The development of Agr-defective mutants during invasive infection and the
hypothesis that these mutations may incur a survival advantage in the presence of
antibiotics led the Edwards lab to investigate the role of the Agr system in daptomy-
cin susceptibility [76]. Somewhat counterintuitively, they discovered that the loss
of AgrA or AgrC, the TCS module, allowed for survival of S. aureus in the presence
of daptomycin. Investigation into the mechanism by which this occurred revealed
that Agr-defective mutants can survive antibiotic exposure by actively releasing
membrane phospholipids that bind to and inactivate daptomycin. This process also
occurs in wild-type bacteria; however, a set of Agr-regulated genes that are
expressed in wild-type bacteria mitigate this observed inactivating effect. The
genes encode molecules called phenol-soluble modulins (PSMs), small cytolytic
toxins that promote binding of daptomycin to the bacteria by sequestering the shed
membrane phospholipid, likely via their surfactant (or detergent-like) properties
[76]. Pader et al. also found that the enhanced survival of Agr-defective mutants in
the presence of daptomycin could be mitigated by the addition of the β-lactam
antibiotic oxacillin, which reduced the rate of lipid release from the bacterial mem-
brane and therefore the inactivation of daptomycin. This result suggests addition of
oxacillin as an immediate clinical remedy for daptomycin-tolerant S. aureus. This
mechanism has been extended to include the enterococci and streptococci in a
recent report [79].
 480 A. T. Ulijasz et al.

DAP DAP

DAP
Extracellular

DAP
DAP

DAP
PSMa

DAP DAP
PSMa PSMa

AIP C
S PSMa

AgrC
Membrane

AgrB

D AgrA
REC
Cytoplasm

DBD

Virulence factors

ArgD
Chromosome

RNAIII

P2 P3
argA argC argD argB RNAIII PSMα1-4

Fig. 15.6 The Agr signaling pathway and daptomycin resistance. AgrC and AgrA comprise a TCS
pair that senses the AIP quorum-sensing autoinducer peptide (AIP) signal. AIP is made from ArgD
peptide, which is processed and secreted by ArgB. The ArgCA TCS system also regulates (alpha
1–4 type) phenol-soluble modulin (PSM) production, which are excreted and bind daptomycin
(DAP) to inactivate it. agrBDCA is transcribed from the P2 promoter, and the virulence regulatory
RNAIII is transcribed from the P3 promoter. Both are controlled by AgrA
15 Bacterial Signal Transduction Systems in Antimicrobial Resistance 481

15.2.5 The CpxR TCS System and Fosfomycin Resistance

The drug fosfomycin, discovered in the late 1960s, has lately reemerged as a sec-
ond-line treatment for urinary tract infections due to the increasing prevalence of
resistance to commonly prescribed antibiotics, such as trimethoprim-sulfamethoxa-
zole and ciprofloxacin [80]. Fosfomycin is primarily taken up by the glycerol-
3-phosphate transport system (GlpT) but may also be internalized by the hexose
phosphate uptake transport system (UhpT) when glucose-6-phosphate is present.
Once inside the bacterium, fosfomycin inactivates cytosolic N-acetylglucosamine
enolpyruvyl transferase (MurA). This impairs bacterial cell wall formation by
inhibiting the first step in peptidoglycan synthesis, the formation of N-acetylmuramic
acid from N-acetylglucosamine and phosphoenolpyruvate [81–83]. Since its clini-
cal launch, several mechanisms for fosfomycin resistance have emerged, including
inducible modulation of the GlpT and UhpT transporters by the Cpx TCS system
[84]. Although this section will focus on the Cpx system as it is understood in E. coli
and as it relates to fosfomycin, the Cpx system has also been implicated in resis-
tance to other antibiotics in E. coli, such as aminoglycosides and beta-lactams, as
well as forms of resistance in Salmonella and P. aeruginosa [85–87].
The Cpx system is comprised by the genes cpxA and cpxR, which encode the
inner membrane sensor histidine kinase CpxA and the response regulator transcrip-
tion factor CpxR [88]. CpxA responds to membrane stressors such as unfolded or
misfolded proteins, as well as to changes in pH. In the absence of stressors, CpxA
functions as a phosphatase, keeping CpxR in its unphosphorylated, inactive form.
The activation of CpxA is modulated by the periplasmic protein CpxP, which inter-
acts with the sensing domain of CpxA; CpxP is displaced by misfolded proteins [89,
90]. Free CpxA is activated by additional signals that are currently not well under-
stood. In typical TCS fashion, once activated, CpxA autophosphorylates and then
transfers the phosphate to CpxR [89, 91]. P-CpxR then acts as a transcription factor
for a variety of genes, including the cpx regulon, degP, glpT, and uhpT (Fig. 15.7)
[84, 88–91].
Interestingly, the Cpx response is shut off by feedback inhibition modulated by
the amount of unfolded protein present in the bacterial cell. P-CpxR increases the
transcription levels of cpxP, and CpxP subsequently inhibits the CpxA kinase, while
also carrying out its function in binding unfolded proteins. Increases in P-CpxR also
increase transcription and translation of the periplasmic protease DegP [88], whose
function is to relieve membrane stress by degrading misfolded or unfolded proteins
in the periplasmic space. DegP recognizes unfolded proteins bound to CpxP, and the
CpxP complex is then degraded by DegP. Importantly, the digestion of CpxP does
not occur in the absence of unfolded proteins. This careful balance suggests a mech-
anism whereby unfolded proteins displace CpxP from CpxA, allowing for auto-
phosphorylation and the subsequent phosphorylation of CpxR, which, in turn, leads
to increased levels of both CpxP and DegP through transcriptional activation
(Fig. 15.7). While unfolded protein remains in the periplasmic space, CpxP will
bind to it and target it for degradation by DegP. As unfolded protein levels fall, more
482 A. T. Ulijasz et al.

Cell wall

DegP
CpxP
Periplasm

(active)
CpxP
(free)
CpxP misfolded
protein G3P F F G6P

quinone
CpxA ArcB pool GlpT UhpT
Membrane
Inner

H H

F F G6P
Cytoplasm

G3P
G3P G6P
G3P
G3P G6P

D CpxR D ArcA G6P

REC REC

DBD DBD
F MurA

Cell wall synthesis

ROS
Chromosome

cpxP degP MurA

Metabolic respiratory
genes, ROS
glpT

uhpT

Fig. 15.7 Fosfomycin resistance and the Cpx regulatory system. Fosfomycin (F) acts on the early
stages of cell wall synthesis by binding MurA and inhibiting its function. The drug enters the cell
via GlpT and UhpT transporters that normally transport glycerol-3-phosphate (G3P) and glycerol-
6-phosphate (G6P) sugars, respectively. The CpxAR TCS system represses glpT and uhpT tran-
scription and activates degP transcription. Free CpxP inhibits CpxA activity. CpxP functions to
bind unfolded protein and deliver it to the DegP protease in the periplasm. The TCS system ArcAB
senses cellular quinone pools and, in turn, regulates aerobic metabolism and resultant reactive
oxygen species (ROS) that is proposed to induce fosfomycin antibiotic lethality

CpxP is free to bind CxpA; this change in the CpxP/CpxA stoichiometry shifts
activity from kinase to phosphatase function in CpxA, increasing the levels of CpxR
relative to P-CpxR (i.e., phosphorylated and active CpxR (Fig. 15.7)). The altering
of the CpxR phosphorylation equilibrium ultimately results in a decrease in the
transcription of membrane stress-response genes [92].
The Cpx system promotes fosfomycin resistance through P-CpxR by negatively
regulating the transcription of glpT and uhpT. This limits fosfomycin entry into the
cell via the GlpT and UhpT transporters (current data suggests that the GlpT trans-
porter is the more physiologically relevant [84]). It is worth noting that the role of
the Cpx system in mediating a response to antibiotics is still under debate. Some
work suggests that the Cpx system activation is involved in antibiotic-mediated
accumulation of reactive oxygen species (ROS), which then can lead to cell death
[93]. This conclusion was drawn by observing that a null cpxA strain of E. coli
reduced the lethality of several antibiotic classes and accumulated less ROS. Other
studies show that mutations turning CpxA into a constitutively active kinase lacking
15 Bacterial Signal Transduction Systems in Antimicrobial Resistance 483

phosphatase activity confer a decrease in susceptibility to some antibiotics, while

leaving the bacteria susceptible to others. These results indicate that some antibiot-
ics might work through the CpxAR TCS system. Kohanski et al. [93] hypothesize
that the ArcBA TCS may be a signaling pair that CpxAR communicates with to
facilitate intracellular ROS accumulation and thus antibiotic-induced ROS-mediated
killing (Fig. 15.7) [93]. Careful distinction between blocking growth and killing
cells is required to assess the contribution of Cpx to ROS [94] (see Chap. 20).
The Cpx system as a means of decreasing fosfomycin susceptibility is of interest
because other responses to fosfomycin challenge involve permanently altering the
activity of important carbon transporters, such as MurA. This enzyme, which is
necessary for bacterial cell wall production [95], is thought to be associated with a
high biological cost to the cell and thus may explain why fosfomycin resistance
rates in clinical practice remain consistently low [95, 96]. However rare, the Cpx
mechanism of resistance presents an inducible means of decreasing GlpT activity
and the potential for these mechanisms to contribute to decreased drug susceptibil-
ity, which is an important consideration in the development of new antibiotics.

15.3 Resistance to Host Antimicrobial Peptides (AMPs)

Although we usually focus on bacteria when contemplating the natural producers of

antimicrobials, these compounds are produced by virtually every life form on earth,
including humans. Human-produced antibiotics are mainly protein-based and are
commonly referred to as antimicrobial peptides (AMPs; they have also been referred
to as host defense peptides or HDPs; for review see Ref. [97]). AMPs are critical
components of the host immune response to infections; however, recently they have
also been implicated in many immunological roles, including modulation of pro-
and anti-inflammatory responses, chemoattraction, enhancement of extracellular
and intracellular bacterial killing, cellular differentiation and activation of the innate
and adaptive immune compartments, wound healing, modulation of autophagy,
apoptosis, and pyroptosis [98]. The mammalian versions of AMPs generally fall
into two major categories: (i) defensins, which are structurally defined by the pres-
ence of a beta-sheet component, and (ii) cathelicidins, which are more heteroge-
neous and are characterized by a conserved and a highly variable cathelicidin
peptide domain. AMPs are effective against microbial invaders, as they are gener-
ally positively charged (cationic), which enables them to more specifically target
bacterial cell surfaces, as the bacterial cell envelope is naturally more negatively
charged than the cell surface of eukaryotic cells. This phenomenon has earned
AMPs the extended name of cationic AMPs (or CAMPs).
CAMPs are a major host defense against bacterial pathogens; they are especially
exploited by innate immune cells, such as neutrophils, macrophages, epithelial
cells, and specialized secretory cells (e.g., Paneth cells that are important for gut
microbiota composition [99]). In phagocytes (e.g., neutrophils and macrophages),
CAMPs are found within the granules where they aid in phagosomal killing once
484 A. T. Ulijasz et al.

bacteria are engulfed. However, CAMPs can also be released into the medium after
phagosomal digestion to kill pathogens that have a more extracellular lifestyle, such
as streptococci, staphylococci, and Pseudomonas species. In the case of epithelial-
type cells, CAMPs can be secreted and act as an intrinsic barrier to pathogens enter-
ing organs or tissues where they are unwelcome. These activities place CAMPs at
the forefront of maintaining the normal gut flora and general microbiome composi-
tion of the host [100].
In order to successfully colonize and infect their host, pathogenic bacteria have
evolved a multitude of CAMP-resistance mechanisms. Among these mechanisms
are (1) repulsion via recharging their cell envelope, (2) sequestration, (3) export via
transporters, and (4) direct enzymatic-driven breakdown of the peptides [101].
Since these collective processes require energy, bacteria must have ways to detect
and respond quickly to host CAMP assaults without compromising precious energy
needed for colonization and pathogenicity. In this section we describe several spe-
cific examples of how bacterial pathogens sense and respond to CAMPs.

15.3.1 Salmonellae

Bacterial CAMP-resistance mechanisms often use the same general signaling and
resistance determinants that they use for bacteria-produced antimicrobials. Probably
the most well-studied CAMP-resistance signaling cascade is PhoPQ TCS pair
[102]. Aside from being an integral signaling component of polymyxin resistance
(see the above section on polymyxins), after engulfment of the bacteria by the host,
PhoPQ is required for sensing the acidic shift in pH and an accompanying drop in
manganese concentrations within the phagosome (see Fig. 15.5). This detection is
crucial to the Salmonella species life cycle, since they are intracellular pathogens
and must therefore escape and/or subvert innate immunity to remain viable and
replicate within host cells, including innate immune cells such as macrophages.
Since phagocytes produce CAMPs to lyse the engulfed Salmonella, once PhoPQ is
activated it initiates expression of several genes that modify the cell envelope charge
and repel CAMPs. To accomplish this, PhoPQ controls a large regulon containing
more than twenty genes, which together change several chemical moieties within
lipopolysaccharides, glycerophospholipids, and outer membrane proteins that alter
the net charge of the cell envelope from negative to positive (Fig. 15.5). As described
in the previous section on polymyxins, these changes are also influenced by the
PhoPQ induction of a second TCS, PmrAB. For a more extensive description of the
chemical basis for PhoPQ-induced CAMP resistance see Ref. [102] and references
therein.
The collective action of PhoPQ/PmrAB induction results in CAMP resistance
and a concomitant decrease in innate immune recognition that ensures the survival
of Salmonella within macrophages and the host environment. One interesting aspect
of the many outer membrane modifications is that PhoPQ/PmrAB may make spe-
cific changes that depend on the particular infected host tissue and environment. For
15 Bacterial Signal Transduction Systems in Antimicrobial Resistance 485

example, using a murine model of infection, it was found that PhoPQ/PmrAB-

mediated induction of the specific aminoarabinose (L-Ara-4N) modification of
Lipid A is required for full colonization of the lumen of the small intestine (Fig. 15.5)
[101]. This modification is likely due to PhoPQ sensing host CAMPs found within
the lumen and then activating PmrAB and downstream resistance regulons that code
for enzymes necessary to resist CAMP assaults.
The association of PhoPQ signaling and CAMP resistance has been extended to
several additional Gram-negative pathogens, including Legionella pneumophilia,
Bordetella species, Francisella species, Yersinia pestis, and P. aeruginosa [103]. In
the case of P. aeruginosa, chemically modified LPS was recovered from bacteria
cultured from the cystic fibrosis lung, suggesting that PhoPQ/PmrAB signaling is
important for this pathogen’s ability to cause cystic fibrosis chronic infection [104].
Collectively, these findings suggest that the Gram-negative cell envelope is not
static but rather chemically dynamic, since modifications occur through PhoPQ/
PmrAB signaling to accommodate different host cell environments and threats in an
energy-efficient manner.

15.3.2 Streptococci

Gram-positive bacteria are also subject to CAMPs from phagocytes and other host
cells. Differing from Gram-negative microbes, Gram-positive bacteria lack an outer
membrane and instead possess an expanded cell wall for fortification, often sur-
rounded by a protective polysaccharide capsule (see Fig. 15.2). These layers col-
lectively act as barriers to phagocytosis and unwanted host molecules. Taking these
differences into consideration, it is not surprising that Gram-positive bacteria have
evolved systems of CAMP resistance that may differ considerably from those of
Gram-negatives, in both signaling and mechanistic outcomes. For example, strepto-
cocci (e.g., S. pneumoniae, S. agalactiae, and S. pyogenes) and staphylococci spe-
cies often rely on an L-lysinylation-protective strategy facilitated by MprF (for
multiple peptide resistance factor). MprF conveys its phenotype by adding posi-
tively charged L-lysine groups to the membrane lipid phosphatidylglycerol, thereby
enabling repulsion of host CAMPs [105–108]. Other strategies include the
D-alanylation of cell wall teichoic acid by the dlt operon, a mechanism conserved
among several streptococcal and other Gram-positive species. This chemical modi-
fication results in increased density and surface charge of the Gram-positive cell
wall, which then acts to absorb and repel CAMPs. D-alanylation generated by the
Dlt proteins also confers resistance to host-produced acid, lysozyme, neutrophil-
produced neutrophil extracellular traps (NETs), and antimicrobial peptides known
as bacteriocins [106]. Such broad resistance capabilities suggest that the dlt operon
and other antimicrobial systems act as versatile signaling and host-responsive
mechanisms, rather than specifically targeting a single antibiotic. This reoccurring
theme is emphasized below.
486 A. T. Ulijasz et al.

The streptococci and staphylococci often possess a polysaccharide capsule that

creates a natural barrier to CAMPs and therefore provides intrinsic resistance [106,
107]. Also present in the Gram-positive arsenal to counterbalance the host CAMP
threat is a litany of: (i) efflux pumps dedicated to the export of host CAMPs, (ii)
proteases that inactive CAMPs, and (iii) other proteins that sequester CAMPs within
the extracellular milieu [106, 107]. Due to the environment that Gram-positive
pathogens, such as the streptococci, encounter within the host, they must have a
means to detect CAMP threats and quickly respond. This is accomplished by many
signaling systems.
Unlike the PhoQ histidine kinase found in Gram-negative bacteria, no bona fide
receptor for CAMPs has been discovered in the streptococci. However, several sig-
naling systems that respond to these peptides have been identified with both labora-
tory-generated and clinical mutant collections. In S. agalactiae, S. pyogenes, and
many related pathogens, the CovRS TCS system has been one of the most studied
signaling systems [109]; it controls many virulence factors and, relevant to this sec-
tion, a moderate level of CAMP resistance. Paralleling the Gram-negative PhoPQ
TCS pair, the S. pyogenes CovRS signaling system activates genes in response to
host CAMP LL-37, a major mammalian CAMP that is also responsible for repres-
sion of genes in response to magnesium. Collectively, the broad response network
controlled by CovRS regulates induction of virulence factors during infection,
including production of the capsule (Ref. [106] and references therein). Although
the mechanism is largely unknown, CovRS CAMP resistance could result from
some of the virulence factors it controls, which include the SpeB protease that
directly cleaves the LL-37 CAMP. SpeB also regulates the S. pyogenes hyaluronic
acid capsule locus (has). However, mutations that arise during infection within
covRS, speB, and the has locus do not result in drastic reduction in susceptibility to
CAMPs. Instead, it appears that multiple, cumulative mutations are necessary for a
pronounced effect. An interesting regulatory twist is that in both S. agalactiae and
S. pyogenes CovR function is also regulated through phosphorylation by the sole
eSTK and eSTP cognate pair found in these pathogens [6, 110]. eSTKs and eSTPs
(bacterial serine-threonine kinases and phosphatases) have been implicated in anti-
biotic resistance in many Gram-positive pathogens [29, 46, 111]. Whether eSTK/
eSTP and CovRS integrate their sensory properties to confer resistance to antibiot-
ics requires further investigation.
In addition to CovRS, other signaling mechanisms have been discovered that
regulate streptococcal resistance to CAMPs. For example, S. agalactiae upregulates
several CAMP-resistance factors in response to subinhibitory CAMP concentra-
tions (specifically LL-37), the response to which is carried out by two TCS systems:
CiaRS (discussed above) and LiaRS [112]. Although CiaRS/LiaRS homologs are
found in many other streptococcal species, at present it is not known whether these
systems directly bind CAMPs to facilitate resistance/tolerance. Based on their broad
range of inducing signals, the CiaRS/LiaRS sensory mechanisms are likely to be
indirect, functionally paralleling sensory attributes of other two-component respon-
sive systems (e.g., PhoPQ).
15 Bacterial Signal Transduction Systems in Antimicrobial Resistance 487

In another example of TCS regulation and responsiveness to the presence of

CAMPs, S. agalactiae and other streptococci utilize LisRS (also referred to as
LisRK). Originally discovered in L. monocytogenes for its role in nicin and cepha-
losporin resistance [113], the LisRS TCS pair appears to be present in many other
Gram-positive pathogens and is seemingly functionally conserved. Similar to other
TCS-induced CAMP resistance/tolerance strategies, LisRS controlled responses
function to chemically alter the bacterial cell envelope to avoid a broad range of
antimicrobials, including vancomycin, bacitracin, and polymyxins. As above, a
common theme is the alteration of cell wall chemical bonds to repel the CAMP
positive charge and their binding to the cell envelope. Interestingly, LisRS is also
postulated to control the expression of antibiotic targets, such as PBPs. In doing so,
the bacterial cell is able to withstand higher concentrations of certain antibiotics by
downregulating their PBPs, resulting in less targets for the drugs to act upon (Ref.
[112] and references therein).
As detailed above, the S. pneumoniae CiaRS system has been implicated in peni-
cillin resistance [39]. CiaRS has also been implicated in conveying responsive pro-
tection to ROS produced by phagocytes, including hypochlorite and hydrogen
peroxide, and also regulates several proteases that could be involved in the degrada-
tion of CAMPs. In contrast, another transcription factor called CrgR, found in S.
pyogenes, likely controls a more specific CAMP-resistance/tolerance response.
Evidence comes from an interesting study by Nizet et al. where they found that
CrgR confers a competitive advantage to S. pyogenes only when infecting the skin
of wild-type mice but not of CAMP knockout mice [114]. Additional work is
required to understand the regulatory profile and mechanism of CrgR, but it is
clearly important in avoiding CAMPs generated by the vertebrate skin during
infection.

15.3.3 Staphylococci

The main CAMP sensory system in S. aureus and related species is the GraRS/
VraFG signaling cascade, which modifies the overall charge of the cell envelope
(described above with reference to its contribution to vancomycin resistance [25];
Fig. 15.2). Thus, this signaling cascade is a general, nonspecific means to facilitate
resistance based on electrostatic repulsion of host-produced CAMPs. However, the
resistance conveyed by GraRS/VraFG can differ greatly depending on the staphylo-
coccus species. For example, GraRS/VraFG recognizes and responds to host LL-37
and indolicidin in both S. aureus and S. epidermidis, but only S. epidermidis GraRS/
VraFG responds to human beta-defensin-3, which protects against skin infections.
S. epidermidis, as its name suggests, is part of the normal flora of human skin and
causes opportunistic skin infections. The species-specific recognition is facilitated
by a short extracellular loop within the GraS histidine kinase sensory domain [115].
Induction of the GraRS signaling cascade (histidine kinase phosphorylation of the
response regulator; Fig. 15.1) then initiates expression of the dlt operon, which is
488 A. T. Ulijasz et al.

responsible for enzymatic incorporation of D-alanine into the cell wall teichoic
acid. Simultaneously, GraRS upregulates MprF, which adds a lysine moiety onto
phosphatidylglycerol (Fig. 15.2). These modifications collectively alter the overall
cell wall and membrane charge of the bacterial envelope, thereby electrostatically
repelling host CAMPs. In addition, GraRS activates the VraFG TCS system, which
has been proposed to govern the efflux of CAMPs from the bacterial cell [107]. The
requirement of the GraRS and VraFG signaling systems for CAMP resistance there-
fore explains experimental evidence showing that these TCS pairs are required to
survive neutrophil attacks and are critical to the success of staphylococcal infections
[107, 116].
In addition to the all-important and very ubiquitous GraRS/VraFG systems, the
staphylococci also use other signaling cascades to counteract CAMP activity. The
global regulators Agr (see section above on daptomycin resistance) and SarA, as
well as the TCS system SaeRS, have been implicated in CAMP resistance through
activation of the controlled expression and release of CAMP-degrading proteases,
such as the exoprotease SepA (Ref. [116]; for review see Ref. [107]). Finally, a
report from 2013 describes another TCS pair, LytSR, which confers staphylococcal
CAMP resistance. Interestingly, the authors proposed that LytSR senses subtle
changes in membrane potential, alerting the bacterial cell of early exposure to host
CAMPs that perturb the normal electrical gradient [117]. An appropriate response
can then be elicited in time. These studies collectively suggest that S. aureus and
other staphylococci have evolved complex regulatory networks to survive host
onslaughts from innate immune cells, and specifically from CAMPs.

15.4 Biofilms and Antimicrobial Resistance

Bacteria are capable of growing either planktonically as free autonomous entities,

or as a largely sessile, immobile community commonly referred to as a biofilm
[118]. The latter state has been the subject of considerable research in recent years,
and it is now well understood that once a biofilm is established within host tissues,
bacteria become far less susceptible to host immunity and antimicrobial treatments
(as much as a thousand times less susceptible [118]). As bacterial infections usually
have a large biofilm component, knowing how to better treat biofilm infections,
which include catheters, surgical implants, and chronic infections such as the cystic
fibrosis lung, remain a consistent and unresolved clinical problem [119].
The mechanism responsible for reduced susceptibility of biofilms to antibiotic
treatments is under debate. One hypothesis is that there is simply more of a physical
and/or mechanical barrier to the drugs. This idea depicts the surface of the biofilm
as being exposed to lethal doses, and as one moves deeper into the biofilm substrata,
antibiotics are diluted to concentrations where they are less effective. However,
there is now compelling evidence that the situation is more complex. For example,
ciprofloxacin, a clinically relevant fluoroquinolone antibiotic, fully penetrates bio-
15 Bacterial Signal Transduction Systems in Antimicrobial Resistance 489

films rapidly, as is the case for several other antimicrobial classes such as the
polyketide tetracycline (see Ref. [120] and references therein).
One commonality to all biofilms is the upregulation of genes producing a protec-
tive glycocalyx or capsular sugary matrix once the bacteria are attached to a host
surface. This physiological change upon surface attachment is often accompanied
by the initiation of a number of AMR resistance mechanisms, including enzymati-
cally driven antibiotic degradation (e.g., by beta-lactamases) and the production of
multidrug resistance efflux pumps. Indeed, both are hypothesized to contribute sig-
nificantly to the staying power of the bacterial biofilm [119, 120].
Biofilms also contain a large DNA component, aptly named “extracellular DNA”
or eDNA, which was found to act as a protectant against innate immunity and some
antibiotics by adsorption [121]. A large fraction of the bacteria within the biofilm
community are also metabolically less active, especially within the deeper layers of
the substratum. This phenotype has now been associated with bacterial stress
responses that include the ppGpp-driven stringent response, or in other cases the
SOS response. Both are involved in controlling bacterial cell death and longevity.
This slowing of metabolism is a hallmark of the so-called “persister” cell pheno-
type, or the subset of cells that “persist” in the wake of antimicrobial or other envi-
ronmental stress [17, 120]. Data now show that persister cells comprise a
considerable proportion of the biofilm and thus might contribute significantly to the
longstanding question of why biofilms render antibiotics less effective (for an
expanded description on persisting microbes see the “Persisters” section below).
The transition from a planktonic state to initial adherence to gradual, stepwise
increments toward a mature and resistant biofilm is a highly complex process that
requires bacterial cells to undergo dramatic physiological, metabolic, and pheno-
typic changes [118]. This process is reversibly controlled through a variety of sig-
naling mechanisms that include quorum-sensing systems (i.e., cell-cell
communication) and signal transduction cascades. In many cases, biofilms are
induced by subinhibitory concentrations of the antibiotic itself, strongly suggesting
that signal transduction systems are involved at sensing and responding to the threat
and play an integral role in tolerance and resistance [122]. Here we will discuss
specific examples from both Gram-positive and Gram-negative bacteria concerning
how pathogens undergo biofilm formation and, importantly, the relevant signaling
proteins involved in biofilm-mediated resistance and tolerance to antibiotics.

15.4.1 Pseudomonas aeruginosa

P. aeruginosa is arguably the most well-studied microorganism with respect to bio-

film formation and its contribution to pathogenicity. This problematic pathogen has
a relatively large genome (5.5–6.8 million base pairs depending on the strain) and
an arsenal of virulence factors that the cell controls through multiple and diverse
signaling cascades. Although its natural habitat is soil and water, its versatility
enables it to colonize and infect a range of animal and plant tissues. In humans, P.
490 A. T. Ulijasz et al.

aeruginosa is problematic with implanted and indwelling devices, as well as with

skin wounds and many internal organ infections, including infections of the urinary
tract (most common) and kidneys. This microbe is most infamous for infecting the
lungs of patients with chronic obstructive pulmonary disorder (COPD) or cystic
fibrosis. For the latter, P. aeruginosa is the major cause of death [120]. One of the
reasons for its disproportionate morbidity and mortality rates is that P. aeruginosa
is naturally resistant to several classes of antibiotics, which is largely associated
with its ability to form biofilms in the chronically infected cystic fibrosis lung. To
regulate biofilm formation, many signaling proteins are involved, some key exam-
ples of which we describe in this section. We focus on cell signaling and AMR
rather than on the in-depth fundamentals of biofilm formation and clinical conse-
quences. For a comprehensive review on biofilms and how they are regulated in P.
aeruginosa and related microbes, see Ref. [118].
As previously mentioned, P. aeruginosa possesses an arsenal of multidrug resis-
tance (MDR) pumps and antibiotic-degrading enzymes to combat both host
onslaughts and therapeutic treatments [120]. As with most resistance mechanisms,
these actions are energetically costly, and they must therefore be coordinated with
environmental ques so they are used only when explicitly required (e.g., in the pres-
ence of an antimicrobial). The so-called antimicrobial “resistome” used by the P.
aeruginosa pathogen was recently shown to be controlled by a biofilm-specific tran-
scription factor called BrlR (for biofilm resistance locus regulator [123, 124]). BrlR
accomplishes this by upregulating the well-known multidrug resistance transporter
mexAB-oprM and mexEF-oprN operons. Interestingly, BrlR does not appear to
respond to specific antibiotics directly, as seen in canonical MerR-type transcription
factors, but rather to the cell-cell (or quorum-sensing) messenger cyclic-di-GMP
(c-di-GMP; for review see Ref. [118] and references therein). This finding revealed
the important connection between regulation of antibiotic-induced biofilms and
quorum sensing, the latter of which had already been associated with the general
transition from a planktonic to sessile state. c-di-GMP directs key phenotypic
changes in P. aeruginosa biofilm formation that include the production of the pro-
tective extracellular matrix, a hallmark of the established biofilm. This matrix con-
tributes loss of susceptibility due to its electrostatic repulsion and absorption
properties [118]. Another report by Sauer and colleagues connects the activity of the
histidine kinase SagS with control of the levels of cellular c-di-GMP and therefore
BrlR responsiveness and biofilm-associated antimicrobial resistance. A sagS dele-
tion mutant was found to have lower c-di-GMP levels, and therefore was more sus-
ceptible to antibiotic treatment [125].
BrlR has also been shown to confer tolerance to host CAMPs, as well as to the
polymyxin colistin and the aminoglycoside tobramycin, the latter of which is typi-
cally used to treat Gram-negative infections, especially in cystic fibrosis patients.
Sauer and colleagues also demonstrated that BrlR regulates the PhoPQ and PmrAB
TCS networks of P. aeruginosa, which change the cell envelope charge confer-
ring tolerance to CAMPs and polymyxins (Fig. 15.5). When expressed, BrlR
represses PhoPQ/PmrAB network (described in detail in the Polymyxin section)
and therefore its signaling. An interesting finding from these studies revealed that
15 Bacterial Signal Transduction Systems in Antimicrobial Resistance 491

many P. aeruginosa clinical isolates from cystic fibrosis patients’ sputum contained
mutations within BrlR, suggesting that in chronic infections BrlR might be inacti-
vated to enhance tolerance to host-produced CAMPs [126]. Indeed, recent clinically
relevant cystic fibrosis models developed in swine have shown that P. aeruginosa
and other bacteria are better able to colonize the diseased cystic fibrosis lung due to
the drastic change in lung pH. It was hypothesized that the pH change alters the
charge of host-produced CAMPs, making the lung far less effective at clearing
microbes from the usually more or less “sterile lung” environment [127]. On the
other hand, when present in its wild-type form, BrlR is responsible for activation of
biofilms and, as a result, tolerance to the aminoglycoside tobramycin. Collectively,
these studies suggest that tobramycin is more effective with isolates that have lost
BrlR function in the more chronic, later stages of the cystic fibrosis disease and,
conversely, might be less effective if given early on.
Almost a decade before the discovery of BrlR the association of a signaling path-
way connecting aminoglycoside resistance and biofilm formation was discovered.
Miller and colleagues described the aminoglycoside response regulator Arr as being
responsible for induction of biofilm formation at subinhibitory concentrations of
aminoglycosides, such as tobramycin, in both P. aeruginosa and E. coli [128]. Arr
contains a c-di-GMP phosphodiesterase (or EAL) domain that directly regulates
biofilm formation and tobramycin resistance. This regulation is accomplished by
responding to subinhibitory levels of antibiotic via an as-of-yet unknown mecha-
nism and, in turn, regulating c-di-GMP levels through breakdown by the Arr EAL
domain [128].
Other regulators that affect the levels of c-di-GMP have also been implicated in
antibiotic resistance, such as the PvrR response regulator, originally described as
controlling P. aeruginosa phenotypic variance and antibiotic resistance. With a pvrR
mutant, the aminoglycoside kanamycin induced small-colony variants, which are
hyper-adhesive and better resist antimicrobials [129]. Similar to Arr, PvrR contains
an EAL domain and is therefore involved in responding to and controlling cellular
c-di-GMP levels. Results from these studies collectively point to bacterial cell-cell
communication systems responding to often subinhibitory concentrations of antibi-
otics and then changing the cell phenotype to enable resistance/tolerance. In this
way the bacteria can communicate the threat to their community and respond in a
timely fashion. The concept of subinhibitory levels of antibiotic resulting in biofilm
formation has now been demonstrated with many diverse pathogens, including
Gram-positive bacteria such as S. aureus. The general response to antimicrobial
treatment and induction of the protective biofilm state is a result of the ability of
signal transduction systems to detect and quickly eliminate the threat. For a compre-
hensive review on antimicrobial induction of biofilms see Ref. [122].
As mentioned in the introduction of this section, biofilms also contain an exten-
sive amount of eDNA. These web-like structures are either secreted by dedicated
export systems in the bacteria or, more indirectly, can be a consequence of cell lysis
[121]. Once outside the cell, eDNA can act as neutrophil extracellular traps (or
NETs) to enable immune system evasion. Interestingly, within the context of the
Pseudomonas biofilm, eDNA was recently shown to play a role in signaling CAMP
492 A. T. Ulijasz et al.

and other AMR mechanisms [121]. Due to the highly anionic nature of eDNA, it
was shown to act as an absorber of cationic metal ions, such as magnesium. As a
consequence, eDNA depletes extracellular magnesium and activates the PhoPQ/
PmrAB TCS, which chemically modifies the P. aeruginosa polysaccharide to repel
host CAMPs (see Fig. 15.5). The same effect holds true with other Gram-negative
bacteria, such as Salmonella species [121]. A surprising discovery was that addition
of eDNA to planktonic cultures of P. aeruginosa cells induced the expression of a
three-gene cluster adjacent to PmrAB that is responsible for the production and
export to the outer membrane of the cationic molecule spermidine. It was hypothe-
sized that spermidine then acts as a positively charged surrogate to magnesium,
repelling CAMPs and positively charged antibiotics, including polymyxins and
aminoglycosides (e.g., tobramycin and gentamicin) [121].

15.4.2 Staphylococci

Gram-positive bacteria are naturally more susceptible to many cell envelope-acting

antimicrobials, as they lack the outer membrane possessed by Gram-negatives.
Therefore, to aid in antibiotic resistance and immune avoidance, it would seem
especially advantageous for them to form a protective, structured biofilm commu-
nity. Although it is well established that Gram-positives also form biofilms, in com-
parison with their Gram-negative counterparts, Gram-positive biofilm formation
and mechanisms are currently less clear and understudied. One difference is that
many Gram-positive biofilm phenotypes tend to be strain-specific rather than a spe-
cies-wide attribute, as in the case of S. pyogenes and S. aureus, for example [130].
Although many species of Gram-positive bacteria form biofilm structures (e.g., oral
pathogen dental plaques and S. pneumoniae colonization of the nasopharynx [131]),
in this section we focus mainly on the staphylococci, as their biofilm formation and
associated signaling have been studied more extensively than with other Gram-
positive microbes.
S. aureus is an animal and human bacterial pathogen capable of causing a wide-
range of infections and known for its formation of biofilms aiding in its success at
inflicting substantial morbidity and mortality in the USA and abroad. S. aureus
infections that are assisted by the biofilm matrix include infective endocarditis and
implant-associated disease. In addition, the skin-dwelling opportunistic pathogen,
S. epidermidis, contributes to many of these same difficult-to-treat biofilm-related
infections [132].
The S. aureus biofilm consists of eDNA and protein and is also largely composed
of the polysaccharide poly-N-acetyl-β-(1–6)-glucosamine (PNAG). The latter is
produced from the icaADBC operon found in most strains. However, expression of
this operon varies among S. aureus strains, and thus the composition of the biofilm
also varies among different isolates [133]. Although the icaADBC operon is respon-
sible for creating the fundamental building blocks of the biofilm, it is curiously not
under control (at least under the conditions tested to date) of the signaling system(s)
15 Bacterial Signal Transduction Systems in Antimicrobial Resistance 493

that direct biofilm formation in S. aureus, namely, the agr quorum-sensing signaling
cascade ([107]; see the section on daptomycin resistance above). Maturation of the
protective S. aureus biofilm structure is facilitated by phenol-soluble modulins
(PSMs), which has been confirmed to be a signaling requirement for catheter-related
biofilm infections in animal infection models ([132, 134]). Despite a plethora of
examples of antibiotic-induced biofilm formation in staphylococcal species, most of
the regulatory connections remain tenuous at best (for review see Ref. [122]).
The direct association between induction of staphylococcal biofilm formation
and antibiotic treatment, especially at subinhibitory concentrations, was first
observed in 1940 by Arthur Gardner with the Gram-positive Clostridium perfrin-
gens (for review Ref. [122]). This effect was then established in several other Gram-
positive and Gram-negative pathogens with many additional antibiotics. However,
the signaling responsible for induction remains unknown for most Gram-positive
cases, with a few recent exceptions with staphylococci. One of these exceptions
comes from a report from Michael Otto and colleagues, which demonstrates that
in vivo PSMs are key contributors to the S. aureus biofilm maturation process [132].
The authors found that PSM placement within the matrix dictates the local struc-
tures of the biofilm. Otto and colleagues propose a model by which this local varia-
tion is controlled by targeted activity of the agr quorum-sensing system (described
in the daptomycin resistance section and displayed in Fig. 15.6) [132]. Agr and
phenol-soluble modulins (PSMs) also control biofilm detachment. Another report
from Schilcher et al. builds upon previous knowledge that at subinhibitory concen-
trations clindamycin induces higher eDNA content in the S. aureus biofilm matrix
[135]. This effect was then determined to be triggered by the alternative sigma fac-
tor B (σB) and its upregulation of known biofilm-associated factors. This report is
important because it provided the missing link in staphylococcal signaling between
inducible biofilm formation and biofilm-driven resistance. The extent to which σB is
integrated within other Gram-positive antibiotic-induced biofilm resistance signal-
ing networks remains to be determined.

15.5 Persisters

Despite the application of antibiotics, a subpopulation of bacteria almost always

survive, only to regrow and again establish infection [136]. This phenomenon,
referred to as “persistence,” was first observed in 1944 by Joseph Bigger after it was
discovered that penicillin was incapable of sterilizing S. aureus cells in culture
[137]. As it turns out persistence is ubiquitous among bacterial pathogens, as antibi-
otic treatments fail to kill a small portion of cells. This problem has contributed
greatly to the tolerance of a variety of bacterial infections in the hospital setting,
notably those involving P. aeruginosa, E. coli, M. tuberculosis, Clostridium diffi-
cile, Salmonella species, and several other Gram-positive and Gram-negative patho-
gens [136, 138].
494 A. T. Ulijasz et al.

Persister cells appear to be genotypically identical to their susceptible counter-

parts, which strongly suggests there are specific and reversible signaling mecha-
nisms in place to “weather the antibiotic storm” at the onset of treatment. We have
briefly touched on this subject in the biofilm section above. We pointed out that a
major reason biofilms are able to resist treatment is that they are comprised of a
large percentage of persister cells. Similar to biofilms, the persister state is revers-
ible and therefore regulated through existing signaling mechanisms. Thus, one
major question researchers have been struggling to answer is what are the environ-
mental ques that are responsible for this transient state of tolerance? So far, studies
have shown that most signals are related to environmental stresses, such as: (i) nutri-
ent limitation – a state that would be observed within the center of biofilms or after
engulfment by phagocytes, (ii) diauxie (the lag between metabolizing two different
energy sources – usually sugars), (iii) extreme shifts in pH, and lastly (iv) DNA
damage [138]. Here we discuss some key bacterial signal transduction cascades that
enable pathogens to respond to these environmental ques, enabling the switch into
and back out of the dormant-like persister state.

15.5.1 Toxin-Antitoxin (TA) Signaling Systems and Persistence

Bacterial toxin-antitoxin (TA) systems were the first signaling systems discovered
to induce persistence through environmental ques. TA systems are ubiquitous in
bacteria, with many species possessing multiple versions. TA systems are usually
encoded within an operon that consists of two genes, one encoding a toxin that regu-
lates cell growth in some manner and the other encoding a cognate antitoxin that
regulates the levels of the toxin (for review see Ref. [139]). In the most simplistic
system, the antitoxin, which is normally a DNA-binding domain-containing tran-
scription factor, directly binds to the toxin. In doing so the toxin then acts as a core-
pressor with the antitoxin to bind DNA and self-regulate repression of the TA
operon [138]. Because the antitoxin is usually produced in excess of toxin, the toxin
then controls the TA relationship through the toxin-to-antitoxin ratio. In this manner
the TA ratio can fine-tune and control the cellular growth and therefore the persis-
tence phenotype. It is interesting to note that the “toxin” in most cases is not actually
a toxin in the classical sense, but rather a regulatory protein or RNA whose action
indirectly results in cellular toxicity (e.g., through posttranslationally modifying
other proteins that actually confer the toxicity). To date there is only one example in
which the toxin component of the TA pair directly influences cellular toxicity (Ref.
[140] and see below). To add another layer of complexity, it is known that with
many TA systems a protease (Lon protease) degrades the antitoxins, which is, in
turn, controlled by cellular phosphate levels [136, 138, 139, 141]. One hypothesis
for how TA systems are able to direct reversible phenotypic heterogeneity involves
the balance the TA module provides between cell growth and arrest. In doing so, it
enables a responsive subpopulation to occur only when required. The mechanism
for how this balance is controlled is still the subject of much debate [141].
15 Bacterial Signal Transduction Systems in Antimicrobial Resistance 495

The first identification of a bacterial genetic locus connected to persistence was

the discovery of the high-persistence (Hip) mutations within the Hip TA system in
E. coli [141].
The hipAB locus consists of the toxin HipA and antitoxin HipB, HipB being a
transcriptional repressor that binds HipA directly. HipA is not a toxin itself, per se,
but is a bacterial eukaryotic-like serine-threonine kinase (eSTK) that acts indirectly
by phosphorylating glutamyl tRNA synthetase to inhibit its function [141]. The accu-
mulation of phosphorylated, uncharged glutamyl tRNA then results in an increase in
ppGpp, the cellular alarmone that controls the stringent response and is a general
controller of persister formation [141]. ppGpp accomplishes this reprogramming by
changing the expression of approximately 500 genes through direct binding of RNA
polymerase. Importantly, ppGpp stimulates RpoS expression, which is the master
stress-response regulator in many bacteria. Because the ppGpp signaling pathway is
a general cellular stress-response signaling pathway, persister-induced signaling
often culminates in ppGpp accumulation to produce the persister phenotype [142].

15.5.2 The SOS Response System and Persistence

In addition to the ppGpp signaling system, other general stress-response mecha-

nisms have been associated with persistence. One such signaling system is the SOS
response and associated genes (e.g., lexA, recA, and recB), which act in response to
DNA damage. When DNA is damaged, single-stranded DNA (ssDNA) accumulates
and activates RecA, which then binds to the transcriptional repressor of the SOS
response, LexA. LexA then becomes inactivated by self-cleavage, and the SOS
response is initiated. Not surprisingly, the SOS signaling system, in particular from
studies done in E. coli, has been implicated in resistance/tolerance and persister
formation in response to DNA-damaging antibiotics such as fluoroquinolones [136].
As it turns out, a TA system called TisAB is responsible. Dorr et al. showed that the
TisB toxin component is a small membrane-acting peptide that influences the pro-
ton motive force of the bacterial cell. When ciprofloxacin (a fluoroquinolone) is
present, the SOS system activates canonical DNA repair enzymes, but it also acts to
produce TisB. TisB then binds to the bacterial membrane, decreasing the proton
motive force and therefore cellular ATP levels, which initiate growth arrest and
thereby the metabolically less active persister phenotype [140]. As TisB is actually
a toxin and not a signaling protein, the TisAB TA system is the only TA system cur-
rently known to directly influence persister formation.

15.5.3 Quorum Sensing and Persistence

In some bacterial species, there is evidence to suggest that persistence is controlled

by quorum-sensing mechanisms. In a rare example of a Gram-positive persister
model of induction, Leung et al. have shown that competence-stimulating peptide
496 A. T. Ulijasz et al.

(CPS), a pheromone produced by the dental pathogen Streptococcus mutans and

related streptococci species, is a stress-induced alarmone, which activates two TA
systems (MazEF and RelEB) to increase the persister population. The CPS compe-
tence pheromone appears to work by inducing a LexA homolog in S. mutans [143].
These studies are important because they connect the DNA uptake mechanism
(induced by CSP) with the SOS response system and persistence in a Gram-positive
organism. How ubiquitous this signaling system is in Gram-positive microbes
remains to be determined. In a particularly intriguing example of bona fide quorum-
sensing induction of persister formation, studies by Vega et al. have shown that
when added to the medium, the bacterial cell-cell communication aromatic com-
pound indole induces persister cell formation in E. coli, as well as in Salmonella
Typhimurium. Remarkably, S. Typhimurium is not known to produce indole; thus,
these data suggest that indole and possibly other low-molecular-weight molecules
could act as cross-species inducers of persistence in, for example, polymicrobial
biofilms [141]. These recent discoveries illustrate the exciting prospect of quorum
sensing controlling persister formation and suggest a feasible path forward to design
new antimicrobials.

15.5.4 Persistence Studies In Vivo

Most studies investigating the mechanisms of persister formation have been accom-
plished using laboratory strains of E. coli in test tubes (i.e., in vitro). Although these
studies have yielded a plethora of information [136], it is also important, perhaps
even more so, to study persister formation in actual pathogens and in the context of
their native host environment. Using S. Typhimurium single-cell analysis with a
murine model of infection, Helaine and colleagues have done just this with surpris-
ing results. They found that within 30 minutes after macrophages engulf S.
Typhimurium in an animal, as much as 20% of the population changes to the per-
sister state, becoming tolerant to antibiotic treatments [144]. This surprising result
was in stark contrast to all previous reports from in vitro experiments, which showed
that the fraction of persisters is normally no greater than 1% of the total population
[138]. The persistence was shown to be aided by Lon protease/ppGpp-dependent TA
modules and, interestingly, triggered by the drastic change in pH and nutrient depra-
vation when the bacteria enter the macrophage vacuole [144]. A second paper by the
same group describes the mechanism by which this occurs through a new class of
uncharacterized TA modules harboring Gcn5 N-acetyltransferase (i.e., acetylation)
activity [145]. It was determined that the acetyl transferase, dubbed TacT for “tRNA
acetylating toxin,” signals for initiation of the persister state by acetylating tRNA,
thereby inhibiting tRNA function and simultaneously inducing lon-ppGpp-medi-
ated cell growth arrest [145]. Furthermore, this acetylation and therefore growth
could be reversed by the S. Typhimurium deacetylase CobB [145]. These studies are
important, as they show a more comprehensive picture of how TA system signaling
can facilitate reversible persister formation that is independent of genotype.
15 Bacterial Signal Transduction Systems in Antimicrobial Resistance 497

15.6 Concluding Remarks

In this chapter we discussed how bacterial cells sense antimicrobials and respond in
a timely manner to tolerate/resist them using signal transduction systems. Although
many systems are specific to a particular species or genus, there are also examples
of conservation that might elicit the targeting of these pathways for new broad-
spectrum antimicrobials. One of these research areas that has recently gained atten-
tion involves drugs that inhibit quorum-sensing pathways, which are shared by both
Gram-positive and Gram-negative organisms and generally control biofilm forma-
tion in many pathogens [146]. In this regard, the most progress has been made with
P. aeruginosa and related pathogens [147]. Other attempts to subvert antibiotic
resistance by inhibition of signaling systems have been made in the area of two-
component signaling. Although two-component signaling systems have been
deemed too diverse and prone to mutagenesis for the design of serious broad-spec-
trum inhibitors, recent efforts to target the conserved WalRK system have resusci-
tated this area of research [148]. An important consideration favoring these
approaches is that both quorum-sensing and two-component signaling are com-
pletely absent from the mammalian genome. Promising data have also been gener-
ated with a recent screening effort for antibiotics that specifically inhibit persister
formation [149]. Thus, this work could have broad implications for preventing bio-
films and ensuing antibiotic resistance/tolerance. This idea seems rational, as persis-
tence signaling seems to eventually culminate in a finite number of stress-response
pathways that are common to many bacterial pathogens. Recent studies have shown
that some antibiotics derived from previously “unculturable” bacteria demonstrate
great potential to provide a new class of persister-targeting antibiotics [149, 150].
Perhaps the future holds a more multifaceted approach to the problem of resistance
and tolerance, such that one antibiotic is given to clear an infection and another to
clear the anticipated tolerant subpopulation.

Major Points
• Tolerance and resistance mechanisms can be energetically costly; thus bacteria
must have a means to sense a threat, induce a response, and terminate a response.
• Bacterial signal transduction systems provide a means to sense and respond to
both their extracellular and intracellular environments.
• Bacterial signaling systems are largely comprised of two-component signaling
(TCS) systems, but they can also be serine-threonine kinase (eSTK) and phos-
phatase (eSTP) systems, along with more specialized signaling such as the S.
aureus Agr cascade.
• A general response to many cell-envelope targeting antimicrobials is the modifi-
cation of the envelope charge to electrostatically repel the drug. Some signaling
systems that detect and respond to antibiotics in this way, such as the PhoQP and
PmrAB TCS systems, are conserved among many bacterial species.
• Bacterial signaling systems, especially cell-cell communication systems (or quo-
rum sensing), are required to regulate biofilm formation and antibiotic tolerance.
498 A. T. Ulijasz et al.

• Within the biofilm, a subpopulation of persister cells exist that are metabolically
less active than their majority counterparts and whose state in many cases is
controlled by toxin-antitoxin (TA) modules.
• Bacterial signaling systems accumulate mutations that contribute to resistance
and tolerance; therefore, they are an important component of antimicrobial
resistance.

References

1. Swartz TE, Tseng TS, Frederickson MA, Paris G, Comerci DJ, Rajashekara G, et al.
Blue-light-activated histidine kinases: two-component sensors in bacteria. Science.
2007;317(5841):1090–3.
2. Kobir A, Shi L, Boskovic A, Grangeasse C, Franjevic D, Mijakovic I. Protein phosphoryla-
tion in bacterial signal transduction. Biochim Biophys Acta. 2011;1810(10):989–94.
3. Hu LI, Lima BP, Wolfe AJ. Bacterial protein acetylation: the dawning of a new age. Mol
Microbiol. 2010;77(1):15–21.
4. Antelmann H, Helmann JD. Thiol-based redox switches and gene regulation. Antioxid Redox
Signal. 2011;14(6):1049–63.
5. Hoch JA, Silhavy TJ, editors. Two-component signal transduction. Washington, D.C.:
American Society for Microbiology; 1995.
6. Wright DP, Ulijasz AT. Regulation of transcription by eukaryotic-like serine-threonine
kinases and phosphatases in gram-positive bacterial pathogens. Virulence. 2014;5:863–85.
7. Burnside K, Rajagopal L. Regulation of prokaryotic gene expression by eukaryotic-like
enzymes. Curr Opin Microbiol. 2012;15(2):125–31.
8. Howden BP, Davies JK, Johnson PD, Stinear TP, Grayson ML. Reduced vancomycin sus-
ceptibility in Staphylococcus aureus, including vancomycin-intermediate and heterogeneous
vancomycin-intermediate strains: resistance mechanisms, laboratory detection, and clinical
implications. Clin Microbiol Rev. 2010;23(1):99–139.
9. D'Costa VM, King CE, Kalan L, Morar M, Sung WW, Schwarz C, et al. Antibiotic resistance
is ancient. Nature. 2011;477(7365):457–61.
10. CDC. Antibiotic resistance threats in the United States. US Department of Health and Human
Services; 2013.
11. Ulrich LE, Koonin EV, Zhulin IB. One-component systems dominate signal transduction in
prokaryotes. Trends Microbiol. 2005;13(2):52–6.
12. Maule AF, Wright DP, Weiner JJ, Han L, Peterson FC, Volkman BF, et al. The aspar-
tate-less receiver (ALR) domains: distribution, structure and function. PLoS Pathog.
2015;11(4):e1004795.
13. Grangeasse C, Cozzone AJ, Deutscher J, Mijakovic I. Tyrosine phosphorylation: an emerging
regulatory device of bacterial physiology. Trends Biochem Sci. 2007;32(2):86–94.
14. Grangeasse C, Nessler S, Mijakovic I. Bacterial tyrosine kinases: evolution, biological func-
tion and structural insights. Philos Trans R Soc Lond Ser B Biol Sci. 2012;367(1602):2640–55.
15. Pereira SF, Goss L, Dworkin J. Eukaryote-like serine/threonine kinases and phosphatases in
bacteria. Microbiol Mol Biol Rev. 2011;75(1):192–212.
16. Libby EA, Goss LA, Dworkin J. The eukaryotic-like Ser/Thr kinase PrkC regulates the essen-
tial WalRK two-component system in Bacillus subtilis. PLoS Genet. 2015;11(6):e1005275.
17. Conlon BP, Rowe SE, Lewis K. Persister cells in biofilm associated infections. Adv Exp Med
Biol. 2015;831:1–9.
18. Levine DP. Vancomycin: a history. Clin Infect Dis. 2006;42(Suppl 1):S5–12.
19. Courvalin P. Vancomycin resistance in gram-positive cocci. Clin Infect Dis. 2006;42(Suppl
1):S25–34.
15 Bacterial Signal Transduction Systems in Antimicrobial Resistance 499

20. Holman TR, Wu Z, Wanner BL, Walsh CT. Identification of the DNA-binding site for the
phosphorylated VanR protein required for vancomycin resistance in Enterococcus faecium.
Biochemistry. 1994;33(15):4625–31.
21. Koteva K, Hong HJ, Wang XD, Nazi I, Hughes D, Naldrett MJ, et al. A vancomycin pho-
toprobe identifies the histidine kinase VanSsc as a vancomycin receptor. Nat Chem Biol.
2010;6(5):327–9.
22. Hughes CS, Longo E, Phillips-Jones MK, Hussain R. Characterisation of the selective bind-
ing of antibiotics vancomycin and teicoplanin by the VanS receptor regulating type A vanco-
mycin resistance in the enterococci. Biochim Biophys Acta. 2017;1861(8):1951–9.
23. Hu Q, Peng H, Rao X. Molecular events for promotion of vancomycin resistance in vanco-
mycin intermediate Staphylococcus aureus. Front Microbiol. 2016;7:1601.
24. Cui L, Neoh HM, Shoji M, Hiramatsu K. Contribution of vraSR and graSR point mutations
to vancomycin resistance in vancomycin-intermediate Staphylococcus aureus. Antimicrob
Agents Chemother. 2009;53(3):1231–4.
25. Falord M, Karimova G, Hiron A, Msadek T. GraXSR proteins interact with the VraFG
ABC transporter to form a five-component system required for cationic antimicrobial
peptide sensing and resistance in Staphylococcus aureus. Antimicrob Agents Chemother.
2012;56(2):1047–58.
26. Dubrac S, Bisicchia P, Devine KM, Msadek T. A matter of life and death: cell wall homeo-
stasis and the WalKR (YycGF) essential signal transduction pathway. Mol Microbiol.
2008;70(6):1307–22.
27. Fridman M, Williams GD, Muzamal U, Hunter H, Siu KW, Golemi-Kotra D. Two unique
phosphorylation-driven signaling pathways crosstalk in Staphylococcus aureus to modulate
the cell-wall charge: Stk1/Stp1 meets GraSR. Biochemistry. 2013;52(45):7975–86.
28. Sun F, Ding Y, Ji Q, Liang Z, Deng X, Wong CC, et al. Protein cysteine phosphorylation
of SarA/MgrA family transcriptional regulators mediates bacterial virulence and antibiotic
resistance. Proc Natl Acad Sci U S A. 2012;109(38):15461–6.
29. Beltramini AM, Mukhopadhyay CD, Pancholi V. Modulation of cell wall structure and anti-
microbial susceptibility by a Staphylococcus aureus eukaryote-like serine/threonine kinase
and phosphatase. Infect Immun. 2009;77(4):1406–16.
30. Kadioglu A, Weiser JN, Paton JC, Andrew PW. The role of Streptococcus pneumoniae viru-
lence factors in host respiratory colonization and disease. Nat Rev Microbiol. 2008;6(4):
288–301.
31. Ulijasz AT, Andes DR, Glasner JD, Weisblum B. Regulation of iron transport in Streptococcus
pneumoniae by RitR, an orphan response regulator. J Bacteriol. 2004;186(23):8123–36.
32. Novak R, Charpentier E, Braun JS, Tuomanen E. Signal transduction by a death signal pep-
tide: uncovering the mechanism of bacterial killing by penicillin. Mol Cell. 2000;5(1):49–57.
33. Haas W, Sublett J, Kaushal D, Tuomanen EI. Revising the role of the pneumococcal vex-
vncRS locus in vancomycin tolerance. J Bacteriol. 2004;186(24):8463–71.
34. Robertson GT, Zhao J, Desai BV, Coleman WH, Nicas TI, Gilmour R, et al. Vancomycin
tolerance induced by erythromycin but not by loss of vncRS, vex3, or pep27 function in
Streptococcus pneumoniae. J Bacteriol. 2002;184(24):6987–7000.
35. Hakenbeck R, Grebe T, Zahner D, Stock JB. beta-lactam resistance in Streptococcus pneu-
moniae: penicillin-binding proteins and non-penicillin-binding proteins. Mol Microbiol.
1999;33(4):673–8.
36. Moscoso M, Domenech M, Garcia E. Vancomycin tolerance in clinical and laboratory
Streptococcus pneumoniae isolates depends on reduced enzyme activity of the major LytA
autolysin or cooperation between CiaH histidine kinase and capsular polysaccharide. Mol
Microbiol. 2010;77(4):1052–64.
37. Liu X, Li JW, Feng Z, Luo Y, Veening JW, Zhang JR. Transcriptional repressor PtvR
­regulates phenotypic tolerance to vancomycin in Streptococcus pneumoniae. J Bacteriol.
2017;199(14):e00054–17.
38. Fleming A. Classics in infectious diseases: on the antibacterial action of cultures of a peni-
cillium, with special reference to their use in the isolation of B. influenzae by Alexander
500 A. T. Ulijasz et al.

Fleming, Reprinted from the British Journal of Experimental Pathology 10:226-236, 1929.
Rev Infect Dis. 1980;2(1):129–39.
39. Muller M, Marx P, Hakenbeck R, Bruckner R. Effect of new alleles of the histidine kinase
gene ciaH on the activity of the response regulator CiaR in Streptococcus pneumoniae R6.
Microbiology. 2011;157(Pt 11):3104–12.
40. Chewapreecha C, Marttinen P, Croucher NJ, Salter SJ, Harris SR, Mather AE, et al.
Comprehensive identification of single nucleotide polymorphisms associated with beta-lac-
tam resistance within pneumococcal mosaic genes. PLoS Genet. 2014;10(8):e1004547.
41. Peacock SJ, Paterson GK. Mechanisms of methicillin resistance in Staphylococcus aureus.
Annu Rev Biochem. 2015;84:577–601.
42. Lim D, Strynadka NC. Structural basis for the beta lactam resistance of PBP2a from methi-
cillin-resistant Staphylococcus aureus. Nat Struct Biol. 2002;9(11):870–6.
43. Otero LH, Rojas-Altuve A, Llarrull LI, Carrasco-Lopez C, Kumarasiri M, Lastochkin
E, et al. How allosteric control of Staphylococcus aureus penicillin binding protein 2a
enables methicillin resistance and physiological function. Proc Natl Acad Sci U S A.
2013;110(42):16808–13.
44. Balish E, Warner T. Enterococcus faecalis induces inflammatory bowel disease in interleu-
kin-10 knockout mice. Am J Pathol. 2002;160(6):2253–7.
45. Amarnani R, Rapose A. Colon cancer and enterococcus bacteremia co-affection: a dangerous
alliance. J Infect Public Health. 2017;10:681.
46. Kristich CJ, Wells CL, Dunny GM. A eukaryotic-type Ser/Thr kinase in enterococcus fae-
calis mediates antimicrobial resistance and intestinal persistence. Proc Natl Acad Sci U S A.
2007;104(9):3508–13.
47. Squeglia F, Marchetti R, Ruggiero A, Lanzetta R, Marasco D, Dworkin J, et al. Chemical
basis of peptidoglycan discrimination by PrkC, a key kinase involved in bacterial resuscita-
tion from dormancy. J Am Chem Soc. 2011;133(51):20676–9.
48. Hall CL, Tschannen M, Worthey EA, Kristich CJ. IreB, a Ser/Thr kinase substrate, influ-
ences antimicrobial resistance in Enterococcus faecalis. Antimicrob Agents Chemother.
2013;57(12):6179–86.
49. Comenge Y, Quintiliani R Jr, Li L, Dubost L, Brouard JP, Hugonnet JE, et al. The CroRS two-
component regulatory system is required for intrinsic beta-lactam resistance in Enterococcus
faecalis. J Bacteriol. 2003;185(24):7184–92.
50. Kellogg SL, Kristich CJ. Functional dissection of the CroRS two-component system required
for resistance to cell wall stressors in enterococcus faecalis. J Bacteriol. 2016;198(8):1326–36.
51. Jaurin B, Grundstrom T. ampC cephalosporinase of Escherichia coli K-12 has a different
evolutionary origin from that of beta-lactamases of the penicillinase type. Proc Natl Acad Sci
U S A. 1981;78(8):4897–901.
52. Zeng X, Lin J. Beta-lactamase induction and cell wall metabolism in Gram-negative bacteria.
Front Microbiol. 2013;4:128.
53. Zamorano L, Moya B, Juan C, Mulet X, Blazquez J, Oliver A. The Pseudomonas aeruginosa
CreBC two-component system plays a major role in the response to beta-lactams, fitness,
biofilm growth, and global regulation. Antimicrob Agents Chemother. 2014;58(9):5084–95.
54. Jeannot K, Bolard A, Plesiat P. Resistance to polymyxins in Gram-negative organisms. Int
J Antimicrob Agents. 2017;49(5):526–35.
55. Tsubery H, Ofek I, Cohen S, Fridkin M. Structure activity relationship study of polymyxin B
nonapeptide. Adv Exp Med Biol. 2000;479:219–22.
56. Trimble MJ, Mlynarcik P, Kolar M, Hancock RE. Polymyxin: alternative mechanisms of
action and resistance. Cold Spring Harb Perspect Med. 2016;6(10):pii: a025288.
57. Wosten MM, Kox LF, Chamnongpol S, Soncini FC, Groisman EA. A signal transduction
system that responds to extracellular iron. Cell. 2000;103(1):113–25.
58. Kato A, Latifi T, Groisman EA. Closing the loop: the PmrA/PmrB two-component system
negatively controls expression of its posttranscriptional activator PmrD. Proc Natl Acad Sci
U S A. 2003;100(8):4706–11.
15 Bacterial Signal Transduction Systems in Antimicrobial Resistance 501

59. Bader MW, Sanowar S, Daley ME, Schneider AR, Cho U, Xu W, et al. Recognition of anti-
microbial peptides by a bacterial sensor kinase. Cell. 2005;122(3):461–72.
60. Olaitan AO, Diene SM, Kempf M, Berrazeg M, Bakour S, Gupta SK, et al. Worldwide emer-
gence of colistin resistance in Klebsiella pneumoniae from healthy humans and patients in Lao
PDR, Thailand, Israel, Nigeria and France owing to inactivation of the PhoP/PhoQ regulator
mgrB: an epidemiological and molecular study. Int J Antimicrob Agents. 2014;44(6):500–7.
61. Jayol A, Poirel L, Brink A, Villegas MV, Yilmaz M, Nordmann P. Resistance to colistin asso-
ciated with a single amino acid change in protein PmrB among Klebsiella pneumoniae iso-
lates of worldwide origin. Antimicrob Agents Chemother. 2014;58(8):4762–6.
62. Gutu AD, Sgambati N, Strasbourger P, Brannon MK, Jacobs MA, Haugen E, et al. Polymyxin
resistance of Pseudomonas aeruginosa phoQ mutants is dependent on additional two-compo-
nent regulatory systems. Antimicrob Agents Chemother. 2013;57(5):2204–15.
63. Steenbergen JN, Alder J, Thorne GM, Tally FP. Daptomycin: a lipopeptide antibiotic for the
treatment of serious Gram-positive infections. J Antimicrob Chemother. 2005;55(3):283–8.
64. Jung D, Rozek A, Okon M, Hancock RE. Structural transitions as determinants of the action
of the calcium-dependent antibiotic daptomycin. Chem Biol. 2004;11(7):949–57.
65. Muraih JK, Pearson A, Silverman J, Palmer M. Oligomerization of daptomycin on mem-
branes. Biochim Biophys Acta. 2011;1808(4):1154–60.
66. Silverman JA, Perlmutter NG, Shapiro HM. Correlation of daptomycin bactericidal activ-
ity and membrane depolarization in Staphylococcus aureus. Antimicrob Agents Chemother.
2003;47(8):2538–44.
67. Chen YF, Sun TL, Sun Y, Huang HW. Interaction of daptomycin with lipid bilayers: a lipid
extracting effect. Biochemistry. 2014;53(33):5384–92.
68. Stefani S, Campanile F, Santagati M, Mezzatesta ML, Cafiso V, Pacini G. Insights and clini-
cal perspectives of daptomycin resistance in Staphylococcus aureus: a review of the available
evidence. Int J Antimicrob Agents. 2015;46(3):278–89.
69. Seaton RA, Menichetti F, Dalekos G, Beiras-Fernandez A, Nacinovich F, Pathan R, et al.
Evaluation of effectiveness and safety of high-dose daptomycin: results from patients
included in the European Cubicin((R)) outcomes registry and experience. Adv Ther.
2015;32(12):1192–205.
70. Miller WR, Bayer AS, Arias CA. Mechanism of action and resistance to daptomycin in
Staphylococcus aureus and Enterococci. Cold Spring Harb Perspect Med. 2016;6(11):pii:
a026997.
71. Muthaiyan A, Silverman JA, Jayaswal RK, Wilkinson BJ. Transcriptional profiling reveals
that daptomycin induces the Staphylococcus aureus cell wall stress stimulon and genes
responsive to membrane depolarization. Antimicrob Agents Chemother. 2008;52(3):980–90.
72. Mehta S, Cuirolo AX, Plata KB, Riosa S, Silverman JA, Rubio A, et al. VraSR two-compo-
nent regulatory system contributes to mprF-mediated decreased susceptibility to daptomycin
in in vivo-selected clinical strains of methicillin-resistant Staphylococcus aureus. Antimicrob
Agents Chemother. 2012;56(1):92–102.
73. Kuroda M, Kuroda H, Oshima T, Takeuchi F, Mori H, Hiramatsu K. Two-component
system VraSR positively modulates the regulation of cell-wall biosynthesis pathway in
Staphylococcus aureus. Mol Microbiol. 2004;49(3):807–21.
74. Dubrac S, Msadek T. Identification of genes controlled by the essential YycG/YycF two-
component system of Staphylococcus aureus. J Bacteriol. 2004;186(4):1175–81.
75. Turck M, Bierbaum G. Purification and activity testing of the full-length YycFGHI proteins
of Staphylococcus aureus. PLoS One. 2012;7(1):e30403.
76. Pader V, Hakim S, Painter KL, Wigneshweraraj S, Clarke TB, Edwards AM. Staphylococcus
aureus inactivates daptomycin by releasing membrane phospholipids. Nat Microbiol.
2016;2:16194.
77. Rutherford ST, Bassler BL. Bacterial quorum sensing: its role in virulence and possibilities
for its control. Cold Spring Harb Perspect Med. 2012;2(11):a012427.
78. Novick RP, Geisinger E. Quorum sensing in staphylococci. Annu Rev Genet. 2008;42:541–64.
502 A. T. Ulijasz et al.

79. Ledger EVK, Pader V, Edwards AM. Enterococcus faecalis and pathogenic streptococci inac-
tivate daptomycin by releasing phospholipids. Microbiology. 2017;163(10):1502–8.
80. Sastry S, Doi Y. Fosfomycin: resurgence of an old companion. J Infect Chemother.
2016;22(5):273–80.
81. Kahan FM, Kahan JS, Cassidy PJ, Kropp H. The mechanism of action of fosfomycin (phos-
phonomycin). Ann N Y Acad Sci. 1974;235(0):364–86.
82. Silhavy TJ, Hartig-Beecken I, Boos W. Periplasmic protein related to the sn-glycerol-3-phos-
phate transport system of Escherichia coli. J Bacteriol. 1976;126(2):951–8.
83. Kadner RJ, Winkler HH. Isolation and characterization of mutations affecting the transport of
hexose phosphates in Escherichia coli. J Bacteriol. 1973;113(2):895–900.
84. Kurabayashi K, Hirakawa Y, Tanimoto K, Tomita H, Hirakawa H. Role of the CpxAR two-
component signal transduction system in control of fosfomycin resistance and carbon sub-
strate uptake. J Bacteriol. 2014;196(2):248–56.
85. Mahoney TF, Silhavy TJ. The Cpx stress response confers resistance to some, but not all,
bactericidal antibiotics. J Bacteriol. 2013;195(9):1869–74.
86. Huang H, Sun Y, Yuan L, Pan Y, Gao Y, Ma C, et al. Regulation of the two-component regu-
lator CpxR on aminoglycosides and beta-lactams resistance in Salmonella enterica serovar
typhimurium. Front Microbiol. 2016;7:604.
87. Tian ZX, Yi XX, Cho A, O'Gara F, Wang YP. CpxR activates MexAB-OprM efflux pump
expression and enhances antibiotic resistance in both laboratory and clinical nalB-type iso-
lates of Pseudomonas aeruginosa. PLoS Pathog. 2016;12(10):e1005932.
88. Danese PN, Snyder WB, Cosma CL, Davis LJ, Silhavy TJ. The Cpx two-component signal
transduction pathway of Escherichia coli regulates transcription of the gene specifying the
stress-inducible periplasmic protease, DegP. Genes Dev. 1995;9(4):387–98.
89. Raivio TL, Popkin DL, Silhavy TJ. The Cpx envelope stress response is controlled by ampli-
fication and feedback inhibition. J Bacteriol. 1999;181(17):5263–72.
90. Tschauner K, Hornschemeyer P, Muller VS, Hunke S. Dynamic interaction between the
CpxA sensor kinase and the periplasmic accessory protein CpxP mediates signal recognition
in E. coli. PLoS One. 2014;9(9):e107383.
91. Raivio TL, Silhavy TJ. Transduction of envelope stress in Escherichia coli by the Cpx two-
component system. J Bacteriol. 1997;179(24):7724–33.
92. Isaac DD, Pinkner JS, Hultgren SJ, Silhavy TJ. The extracytoplasmic adaptor protein CpxP
is degraded with substrate by DegP. Proc Natl Acad Sci U S A. 2005;102(49):17775–9.
93. Kohanski MA, Dwyer DJ, Wierzbowski J, Cottarel G, Collins JJ. Mistranslation of mem-
brane proteins and two-component system activation trigger antibiotic-mediated cell death.
Cell. 2008;135(4):679–90.
94. Zhao X, Hong Y, Drlica K. Moving forward with reactive oxygen species involvement in
antimicrobial lethality. J Antimicrob Chemother. 2015;70(3):639–42.
95. Takahata S, Ida T, Hiraishi T, Sakakibara S, Maebashi K, Terada S, et al. Molecular mecha-
nisms of fosfomycin resistance in clinical isolates of Escherichia coli. Int J Antimicrob
Agents. 2010;35(4):333–7.
96. Nilsson AI, Berg OG, Aspevall O, Kahlmeter G, Andersson DI. Biological costs and
mechanisms of fosfomycin resistance in Escherichia coli. Antimicrob Agents Chemother.
2003;47(9):2850–8.
97. Mookherjee N, Hancock RE. Cationic host defence peptides: innate immune regulatory pep-
tides as a novel approach for treating infections. Cell Mol Life Sci. 2007;64(7–8):922–33.
98. Mansour SC, Pena OM, Hancock RE. Host defense peptides: front-line immunomodulators.
Trends Immunol. 2014;35(9):443–50.
99. Bauer ME, Shafer WM. On the in vivo significance of bacterial resistance to antimicrobial
peptides. Biochim Biophys Acta. 2015;1848(11 Pt B):3101–11.
100. Muniz LR, Knosp C, Yeretssian G. Intestinal antimicrobial peptides during homeostasis,
infection, and disease. Front Immunol. 2012;3:310.
15 Bacterial Signal Transduction Systems in Antimicrobial Resistance 503

101. Richards SM, Strandberg KL, Conroy M, Gunn JS. Cationic antimicrobial peptides serve as
activation signals for the Salmonella typhimurium PhoPQ and PmrAB regulons in vitro and
in vivo. Front Cell Infect Microbiol. 2012;2:102.
102. Dalebroux ZD, Miller SI. Salmonellae PhoPQ regulation of the outer membrane to resist
innate immunity. Curr Opin Microbiol. 2014;17:106–13.
103. Needham BD, Trent MS. Fortifying the barrier: the impact of lipid A remodelling on bacterial
pathogenesis. Nat Rev Microbiol. 2013;11(7):467–81.
104. Ernst RK, Yi EC, Guo L, Lim KB, Burns JL, Hackett M, et al. Specific lipopolysaccharide
found in cystic fibrosis airway Pseudomonas aeruginosa. Science. 1999;286(5444):1561–5.
105. Ernst CM, Staubitz P, Mishra NN, Yang SJ, Hornig G, Kalbacher H, et al. The bacterial
defensin resistance protein MprF consists of separable domains for lipid lysinylation and
antimicrobial peptide repulsion. PLoS Pathog. 2009;5(11):e1000660.
106. LaRock CN, Nizet V. Cationic antimicrobial peptide resistance mechanisms of streptococcal
pathogens. Biochim Biophys Acta. 2015;1848(11 Pt B):3047–54.
107. Joo HS, Otto M. Mechanisms of resistance to antimicrobial peptides in staphylococci.
Biochim Biophys Acta. 2015;1848(11 Pt B):3055–61.
108. Peschel A, Jack RW, Otto M, Collins LV, Staubitz P, Nicholson G, et al. Staphylococcus
aureus resistance to human defensins and evasion of neutrophil killing via the novel viru-
lence factor MprF is based on modification of membrane lipids with l-lysine. J Exp Med.
2001;193(9):1067–76.
109. Graham MR, Smoot LM, Migliaccio CA, Virtaneva K, Sturdevant DE, Porcella SF, et al.
Virulence control in group A Streptococcus by a two-component gene regulatory system:
global expression profiling and in vivo infection modeling. Proc Natl Acad Sci U S A.
2002;99(21):13855–60.
110. Burnside K, Rajagopal L. Aspects of eukaryotic-like signaling in Gram-positive cocci: a
focus on virulence. Future Microbiol. 2011;6(7):747–61.
111. Pensinger DA, Aliota MT, Schaenzer AJ, Boldon KM, Ansari IU, Vincent WJ, et al. Selective
pharmacologic inhibition of a PASTA kinase increases Listeria monocytogenes susceptibility
to beta-lactam antibiotics. Antimicrob Agents Chemother. 2014;58(8):4486–94.
112. Klinzing DC, Ishmael N, Dunning Hotopp JC, Tettelin H, Shields KR, Madoff LC, et al. The
two-component response regulator LiaR regulates cell wall stress responses, pili expression
and virulence in group B Streptococcus. Microbiology. 2013;159(Pt 7):1521–34.
113. Cotter PD, Guinane CM, Hill C. The LisRK signal transduction system determines the sensi-
tivity of Listeria monocytogenes to nisin and cephalosporins. Antimicrob Agents Chemother.
2002;46(9):2784–90.
114. Nizet V, Ohtake T, Lauth X, Trowbridge J, Rudisill J, Dorschner RA, et al. Innate antimicrobial
peptide protects the skin from invasive bacterial infection. Nature. 2001;414(6862):454–7.
115. Cheung AL, Bayer AS, Yeaman MR, Xiong YQ, Waring AJ, Memmi G, et al. Site-specific
mutation of the sensor kinase GraS in Staphylococcus aureus alters the adaptive response to
distinct cationic antimicrobial peptides. Infect Immun. 2014;82(12):5336–45.
116. Cheung GY, Rigby K, Wang R, Queck SY, Braughton KR, Whitney AR, et al.
Staphylococcus epidermidis strategies to avoid killing by human neutrophils. PLoS Pathog.
2010;6(10):e1001133.
117. Yang SJ, Xiong YQ, Yeaman MR, Bayles KW, Abdelhady W, Bayer AS. Role of the LytSR
two-component regulatory system in adaptation to cationic antimicrobial peptides in
Staphylococcus aureus. Antimicrob Agents Chemother. 2013;57(8):3875–82.
118. Valentini M, Filloux A. Biofilms and cyclic di-GMP (c-di-GMP) signaling: lessons from
Pseudomonas aeruginosa and other bacteria. J Biol Chem. 2016;291(24):12547–55.
119. Lebeaux D, Ghigo JM, Beloin C. Biofilm-related infections: bridging the gap between clini-
cal management and fundamental aspects of recalcitrance toward antibiotics. Microbiol Mol
Biol Rev. 2014;78(3):510–43.
120. Mulcahy LR, Isabella VM, Lewis K. Pseudomonas aeruginosa biofilms in disease. Microb
Ecol. 2014;68(1):1–12.
504 A. T. Ulijasz et al.

121. Lewenza S. Extracellular DNA-induced antimicrobial peptide resistance mechanisms in

Pseudomonas aeruginosa. Front Microbiol. 2013;4:21.
122. Kaplan JB. Antibiotic-induced biofilm formation. Int J Artif Organs. 2011;34(9):737–51.
123. Liao J, Schurr MJ, Sauer K. The MerR-like regulator BrlR confers biofilm tolerance
by activating multidrug efflux pumps in Pseudomonas aeruginosa biofilms. J Bacteriol.
2013;195(15):3352–63.
124. Chambers JR, Liao J, Schurr MJ, Sauer K. BrlR from Pseudomonas aeruginosa is a c-di-
GMP-responsive transcription factor. Mol Microbiol. 2014;92(3):471–87.
125. Gupta K, Liao J, Petrova OE, Cherny KE, Sauer K. Elevated levels of the second messenger
c-di-GMP contribute to antimicrobial resistance of Pseudomonas aeruginosa. Mol Microbiol.
2014;92(3):488–506.
126. Chambers JR, Sauer K. The MerR-like regulator BrlR impairs Pseudomonas aeruginosa bio-
film tolerance to colistin by repressing PhoPQ. J Bacteriol. 2013;195(20):4678–88.
127. Pezzulo AA, Tang XX, Hoegger MJ, Abou Alaiwa MH, Ramachandran S, Moninger TO,
et al. Reduced airway surface pH impairs bacterial killing in the porcine cystic fibrosis lung.
Nature. 2012;487(7405):109–13.
128. Hoffman LR, D'Argenio DA, MacCoss MJ, Zhang Z, Jones RA, Miller SI. Aminoglycoside
antibiotics induce bacterial biofilm formation. Nature. 2005;436(7054):1171–5.
129. Drenkard E, Ausubel FM. Pseudomonas biofilm formation and antibiotic resistance are
linked to phenotypic variation. Nature. 2002;416(6882):740–3.
130. Fiedler T, Koller T, Kreikemeyer B. Streptococcus pyogenes biofilms-formation, biology, and
clinical relevance. Front Cell Infect Microbiol. 2015;5:15.
131. Gilley RP, Orihuela CJ. Pneumococci in biofilms are non-invasive: implications on nasopha-
ryngeal colonization. Front Cell Infect Microbiol. 2014;4:163.
132. Periasamy S, Joo HS, Duong AC, Bach TH, Tan VY, Chatterjee SS, et al. How
Staphylococcus aureus biofilms develop their characteristic structure. Proc Natl Acad Sci U
S A. 2012;109(4):1281–6.
133. Hobley L, Harkins C, MacPhee CE, Stanley-Wall NR. Giving structure to the biofilm matrix:
an overview of individual strategies and emerging common themes. FEMS Microbiol Rev.
2015;39(5):649–69.
134. Wang R, Khan BA, Cheung GY, Bach TH, Jameson-Lee M, Kong KF, et al. Staphylococcus
epidermidis surfactant peptides promote biofilm maturation and dissemination of biofilm-
associated infection in mice. J Clin Invest. 2011;121(1):238–48.
135. Schilcher K, Andreoni F, Dengler Haunreiter V, Seidl K, Hasse B, Zinkernagel AS. Modulation
of Staphylococcus aureus biofilm matrix by subinhibitory concentrations of clindamycin.
Antimicrob Agents Chemother. 2016;60(10):5957–67.
136. Lewis K. Persister cells. Annu Rev Microbiol. 2010;64:357–72.
137. Bigger JW. Treatment of staphyloccal infections with penicillin by intermittent sterilisation.
Lancet. 1944;ii:497–500.
138. Fisher RA, Gollan B, Helaine S. Persistent bacterial infections and persister cells. Nat Rev
Microbiol. 2017;15:453.
139. Gerdes K, Maisonneuve E. Bacterial persistence and toxin-antitoxin loci. Annu Rev
Microbiol. 2012;66:103–23.
140. Dorr T, Vulic M, Lewis K. Ciprofloxacin causes persister formation by inducing the TisB
toxin in Escherichia coli. PLoS Biol. 2010;8(2):e1000317.
141. Germain E, Castro-Roa D, Zenkin N, Gerdes K. Molecular mechanism of bacterial persis-
tence by HipA. Mol Cell. 2013;52(2):248–54.
142. Maisonneuve E, Castro-Camargo M, Gerdes K. (p)ppGpp controls bacterial persistence by
stochastic induction of toxin-antitoxin activity. Cell. 2013;154(5):1140–50.
143. Leung V, Ajdic D, Koyanagi S, Levesque CM. The formation of Streptococcus mutans per-
sisters induced by the quorum-sensing peptide pheromone is affected by the LexA regulator.
J Bacteriol. 2015;197(6):1083–94.
15 Bacterial Signal Transduction Systems in Antimicrobial Resistance 505

144. Helaine S, Cheverton AM, Watson KG, Faure LM, Matthews SA, Holden DW. Internalization
of Salmonella by macrophages induces formation of nonreplicating persisters. Science.
2014;343(6167):204–8.
145. Cheverton AM, Gollan B, Przydacz M, Wong CT, Mylona A, Hare SA, et al. A Salmonella
toxin promotes persister formation through acetylation of tRNA. Mol Cell. 2016;63(1):86–96.
146. Grandclement C, Tannieres M, Morera S, Dessaux Y, Faure D. Quorum quenching: role in
nature and applied developments. FEMS Microbiol Rev. 2016;40(1):86–116.
147. O'Loughlin CT, Miller LC, Siryaporn A, Drescher K, Semmelhack MF, Bassler BL. A quo-
rum-sensing inhibitor blocks Pseudomonas aeruginosa virulence and biofilm formation. Proc
Natl Acad Sci U S A. 2013;110(44):17981–6.
148. Worthington RJ, Blackledge MS, Melander C. Small-molecule inhibition of bacterial
two-component systems to combat antibiotic resistance and virulence. Future Med Chem.
2013;5(11):1265–84.
149. Ling LL, Schneider T, Peoples AJ, Spoering AL, Engels I, Conlon BP, et al. A new antibiotic
kills pathogens without detectable resistance. Nature. 2015;517(7535):455–9.
150. Lewis K. New approaches to antimicrobial discovery. Biochem Pharmacol. 2017;134:87–98.
Chapter 16
Bacterial Type II Topoisomerases
and Target-Mediated Drug Resistance

Elizabeth G. Gibson, Rachel E. Ashley, Robert J. Kerns, and Neil Osheroff

16.1 Introduction

Fluoroquinolones are among the most efficacious and broad-spectrum oral antibac-
terials currently in clinical use [1–4]. They are used as frontline treatments for a
wide variety of infections caused by Gram-negative and Gram-positive bacteria [5].
Among the diseases treated with fluoroquinolones are urinary tract infections and
pyelonephritis, sexually transmitted diseases, prostatitis, skin and tissue infections,
chronic bronchitis, community-acquired and nosocomial pneumonia, and intra-­
abdominal and pelvic infections [5]. Fluoroquinolones are also the first line of pro-
phylactic treatment for anthrax, the “biological agent most likely to be used” in a
bioterrorist attack, according to the Centers for Disease Control and Prevention
(CDC) [6]. Furthermore, they are commonly used to treat tuberculosis in cases of
resistance or patient intolerance to established regimens [7]. Tuberculosis recently
overtook HIV/AIDS as the deadliest disease in the world caused by a single infec-
tive agent [8].
Fluoroquinolones kill bacteria by increasing levels of double-stranded DNA
breaks generated by enzymes known as type II topoisomerases [2, 9–12]. The vast

E. G. Gibson
Vanderbilt University School of Medicine, Department of Pharmacology,
Nashville, TN, USA
R. E. Ashley
Vanderbilt University School of Medicine, Department of Biochemistry, Nashville, TN, USA
R. J. Kerns
University of Iowa College of Pharmacy, Department of Pharmaceutical Sciences and
Experimental Therapeutics, Iowa City, IA, USA
N. Osheroff (*)
Vanderbilt University School of Medicine, Departments of Biochemistry and Medicine, VA
Tennessee Valley Healthcare System, Nashville, TN, USA
e-mail: [email protected]

© Springer International Publishing AG, part of Springer Nature 2018 507

I. W. Fong et al. (eds.), Antimicrobial Resistance in the 21st Century, Emerging Infectious
Diseases of the 21st Century, https://doi.org/10.1007/978-3-319-78538-7_16
508 E. G. Gibson et al.

majority of bacteria encode two type II enzymes, gyrase and topoisomerase IV [10,
13–15]. These enzymes are essential for cell survival, and both appear to be physi-
ological targets for fluoroquinolones [2, 9, 10, 14]. In contrast, a handful of species
encode only gyrase. This group includes a number of disease-causing organisms,
including Treponema pallidum (syphilis) [16], Helicobacter pylori (stomach and
intestinal ulcers) [17], Campylobacter jejuni (gastroenteritis) [18], Mycobacterium
leprae (leprosy) [19], and Mycobacterium tuberculosis (tuberculosis) [20]. In these
species, gyrase takes on dual characteristics and can fulfill its own functions as well
as those of topoisomerase IV [21].
Unfortunately, fluoroquinolone usage is being threatened by an increasing preva-
lence of resistance, which extends to every bacterial infection treated by this drug
class [1, 2]. The most common and clinically relevant form of resistance is target-­
mediated, which is caused by specific mutations in gyrase and topoisomerase IV
[22, 23]. Therefore, it is critically important to understand how this drug class inter-
acts with and alters the activity of its enzyme targets to better guide drug develop-
ment and to overcome resistance [11]. In this chapter, we will discuss fluoroquinolone
action and targeting, resistance mechanisms, and efforts to overcome this
resistance.

16.2 Fluoroquinolones

The history of the fluoroquinolones began in 1962, when Lesher et al. made the
accidental discovery of nalidixic acid (Fig. 16.1) as a by-product of the synthesis of
the antimalarial compound chloroquine [24]. This first-generation quinolone dis-
played limited efficacy and was used mainly for the treatment of uncomplicated
urinary tract infections caused by Gram-negative enteric bacteria [25]. In the 1980s,
the second generation of quinolones was established when norfloxacin (Fig. 16.1)
was synthesized [1, 25, 26]. This drug featured a fluorine at the C6 position, making
it the first true fluoroquinolone, and a cyclic diamine piperazine at the C7 position.
The fluorine at the C6 position increased tissue penetration and has been included
in every subsequent clinically relevant member of this drug class [1, 25–27].
Even with improved tissue penetration, norfloxacin was still confined to the uri-
nary tract and displayed low serum concentrations [1, 25–27]. However, it broad-
ened the use of quinolones to include sexually transmitted diseases [1, 25–27].
Ciprofloxacin (Fig. 16.1) was the first fluoroquinolone to display efficacy toward
both Gram-positive and Gram-negative bacterial species and was the first with suf-
ficiently high tissue penetration and serum concentration to be used outside the
urinary tract [1, 25, 26]. The clinical success of ciprofloxacin spawned the develop-
ment of third-generation fluoroquinolones that include moxifloxacin, gatifloxacin,
and levofloxacin (Fig. 16.1) [1, 25–27]. These drugs all exhibit improved half-lives
compared to ciprofloxacin [28]. Moreover, they have extended the spectrum of
fluoroquinolone activity to include a broader array of Gram-positive bacteria
­
(including a number of respiratory infections), as well as atypical pathogens such as
Legionella pneumophila and Chlamydia pneumoniae [2–5].
16 Bacterial Type II Topoisomerases and Target-Mediated... 509

Fig. 16.1 Fluoroquinolone structures. Nalidixic acid, a first-generation quinolone, is the founding
member of this drug class. This drug displayed limited efficacy for systemic infections and had a
narrow antibacterial spectrum. Norfloxacin and ciprofloxacin, second-generation fluoroquino-
lones, had improved efficacy, with ciprofloxacin being the more efficacious of the two.
Ciprofloxacin displayed an improved antibacterial profile that included additional Gram-positive
bacterial infections and improved Gram-negative coverage. Moxifloxacin, gatifloxacin, and levo-
floxacin, third-generation fluoroquinolones, are the most efficacious and broad-spectrum fluoro-
quinolones in clinical use today
510 E. G. Gibson et al.

16.3 Bacterial Type II DNA Topoisomerases

Type II topoisomerases control the topological state of the DNA in the cell [9–11,
29–31]. These enzymes modulate DNA under- and overwinding (i.e., negative and
positive supercoiling, respectively) and remove tangles and knots from the genome
[11, 12, 31–35]. As discussed earlier, most bacterial species encode two type II
topoisomerases, gyrase and topoisomerase IV [9–11, 29–31]. Gyrase was the first
type II topoisomerase to be described in any species and was originally reported in
1976 [36]. It is an A2B2 heterotetramer in which the two subunit types are GyrA and
GyrB [9–11, 29–31]. The A subunits contain the active site tyrosine residues that
cleave the DNA (shown in blue in Fig. 16.2). The B subunits form the N-terminal
gate of the enzyme and contain the sites of ATP binding and hydrolysis (shown in
green in Fig. 16.2) [9–11, 29–31].
The subunits of topoisomerase IV were first identified in Gram-negative species
as being required for chromosome partitioning and were named ParC and ParE (blue
and green in Fig. 16.2, respectively) [10–13, 31, 37–39]. Sequence analysis revealed
that these proteins were homologous to GyrA and GyrB, respectively. In 1990, it
was determined that the ParC/ParE complex was a heterotetramer that functioned as
a distinct type II topoisomerase [10–13, 31, 37–39]. The enzyme was subsequently
named topoisomerase IV. Whereas the subunits of topoisomerase IV are denoted as
ParC and ParE in Gram-negative species because of their historic roles in chromo-
some partitioning, they are called GrlA and GrlB, respectively, (which comes from
their initial name gyrase-like proteins) in Gram-positive species.
Gyrase and topoisomerase IV regulate DNA topology by using a double-stranded
DNA passage mechanism [10–13, 31, 34]: the enzymes generate a double-stranded
break in the gate or G-segment (green in Fig. 16.2) and pass the transport or
T-segment (red) through the open DNA gate. The transport helix eventually exits the
enzymes when the two subunits at the bottom of the enzyme open to form the exit
gate. This reaction takes place at the expense of ATP binding, which opens the DNA
gate and induces a conformational change that moves the T-segment through the
open gate, and ATP hydrolysis, which drives enzyme turnover.
As a prerequisite for opening the DNA gate, gyrase and topoisomerase IV gener-
ate a double-stranded break in the G-segment [10–13, 31, 34]. The scissile bonds on
the two strands of the double helix are located across the major groove from one
another. Cleavage results in 5′-overhanging termini with a four-base cohesive stag-
ger. In order to maintain genomic integrity during the DNA cleavage event, the
enzymes form a covalent phosphotyrosine linkage between active site residues and
the newly generated 5′ termini. This covalent enzyme-cleaved DNA complex is criti-
cal for the actions of quinolones and is called the cleavage complex [10–13, 31, 34].
Despite the sequence and structural similarities between gyrase and topoisomer-
ase IV, differences in the C-terminal domains of GrlA/ParC and GyrA confer these
enzymes with unique catalytic activities [10–13, 31, 34]. The C-terminal domain of
GrlA/ParC allows topoisomerase IV to interact with two distal DNA segments.
Thus, the enzyme uses a “canonical” strand-passage mechanism in which it cap-
16 Bacterial Type II Topoisomerases and Target-Mediated... 511

Fig. 16.2 Cellular functions of bacterial type II topoisomerases. Topoisomerase IV (bottom left)
uses a canonical double-stranded DNA passage mechanism. The enzyme can remove positive
supercoils but acts primarily behind the replication fork (middle) to remove precatenanes and
unlink daughter chromosomes. Gyrase (bottom right) uses a DNA-wrapping mechanism that is
superimposed upon the double-stranded DNA passage reaction. The enzyme removes positive
DNA supercoils ahead of transcription (top) and replication (middle) complexes and maintains the
negative superhelicity of the genome. (Artwork by Ethan Tyler, NIH Medical Arts)

tures existing intra- or intermolecular DNA crossovers (Fig. 16.2, left) [10–13, 31,
34]. This allows the enzyme to relax (i.e., remove) positive or negative DNA super-
coils and to remove DNA tangles and knots in a highly efficient manner [10–13, 31,
34]. Although topoisomerase IV is able to alleviate torsional stress ahead of DNA
tracking systems and appears to play a role in regulating genomic superhelicity, its
major function is to remove the precatenanes that form behind DNA replication
forks (Fig. 16.2 middle), separate daughter chromosomes following replication, and
512 E. G. Gibson et al.

Intact DNA Cleaved DNA

Fluoroquinolones

Decreasing Cleavage Increasing Cleavage

Slow growth rates SOS response
(gyrase) (gyrase, topoisomerase IV)
Mitotic failure Balanced Cleavage/Religation Mutagenesis
(topoisomerase IV) Appropriatechromosome segregation
Genome maintenance Cell death
Cell death
Normal cell growth

Fig. 16.3 Bacterial type II topoisomerases: the critical balance of DNA cleavage and ligation. The
DNA cleavage/ligation activity of bacterial type II topoisomerases must be regulated in the cell.
When an appropriate level of cleavage complexes is maintained, topological problems within DNA
are resolved, and the cell can grow normally. If the levels of cleavage complexes decrease, slow
growth rates and mitotic failure can cause cell death. Conversely, if the levels of cleavage com-
plexes are too high, the resulting strand breaks can block essential nucleic acid functions and
induce the SOS response, generate mutations, and lead to cell death. Compounds that increase
levels of gyrase or topoisomerase IV cleavage complexes, such as fluoroquinolones, act as topoi-
somerase “poisons” and convert the proteins to cellular toxins that have the potential to fragment
the genome. Compounds that inhibit the catalytic activity of gyrase or topoisomerase IV without
increasing levels of DNA cleavage work by robbing the cell of essential enzyme function leading
to slow growth rates, mitotic failure, and cell death. (Adapted from Pendleton et al. [46])

remove DNA knots that form during DNA recombination [13, 40–45]. If topoisom-
erase IV activity drops below threshold levels, cells die of mitotic failure (Fig. 16.3)
[10–13, 31, 34].
In contrast to the canonical mechanism used by topoisomerase IV, gyrase uses a
mechanism in which the C-terminal domain of the GyrA subunit (gray, Fig. 16.2)
wraps DNA, inducing a positive crossover between the G- and T-segments that
mimics a positive supercoil [13, 47–49]. Because of this “wrapping” mechanism,
the captured G- and T-segments are proximal to one another [50]. As a result, gyrase
greatly favors the catalysis of intra- over intermolecular strand-passage reactions.
Consequently, the enzyme can efficiently alter superhelical density but is very poor
at removing tangles and knots [50, 51]. In addition, because gyrase always acts on
the induced positive crossover, it works in a unidirectional manner [13, 52]; in the
presence of ATP, the enzyme can remove positive, but not negative supercoils.
Furthermore, gyrase is able to induce negative supercoils into relaxed DNA [36,
53]. The major cellular roles of gyrase stem from its DNA-wrapping mechanism.
Gyrase functions ahead of replication forks and transcription complexes to alleviate
the torsional stress induced by DNA overwinding (Fig. 16.2, middle and bottom)
[31, 54]. Furthermore, in conjunction with the ω protein (commonly called topoi-
somerase I), a type I topoisomerase, gyrase modulates the superhelicity of the bac-
16 Bacterial Type II Topoisomerases and Target-Mediated... 513

terial chromosome and allows the organism to maintain its genetic material in an
underwound state [55]. If gyrase activity in the cell drops, rates of replication/tran-
scription are severely impacted (Fig. 16.3) [9–13, 53]. Furthermore, a number of
pleiotropic effects on gene expression are observed due to changes in superhelicity
of the bacterial chromosome [56].
Although gyrase and topoisomerase IV are essential enzymes, they also pose a
threat to the bacterial cell. Indeed, if a replication fork, transcription complex, or
DNA tracking system encounters and attempts to pass through a gyrase- or topoi-
somerase IV-mediated DNA cleavage complex, it can disrupt the complex and ren-
der the enzyme unable to ligate the DNA [40]. This event generates double-stranded
DNA breaks that require recombination pathways to repair. Thus, these breaks
block essential nucleic acid functions, induce the SOS response, generate muta-
tions, and trigger processes that ultimately impair cell survival [2, 3, 12, 26, 57, 58].
Compounds that increase levels of gyrase or topoisomerase IV cleavage com-
plexes are referred to as “poisons” [59], because they are said to poison these pro-
teins, converting them to cellular toxins that have the potential to fragment the
bacterial chromosome [2, 9, 26, 57, 58]. The term “poison” distinguishes these
compounds from “catalytic inhibitors,” which act primarily by robbing the cell of
the catalytic functions of these enzymes [34].

16.4 Fluoroquinolone Mechanism

Fluoroquinolones are potent gyrase/topoisomerase IV poisons [9, 11, 12, 23, 60–
62]. These drugs interact with both the protein and DNA within a cleavage complex
and intercalate into the DNA backbone at the cleaved scissile bonds [23, 62].
Consequently, two fluoroquinolone molecules are required to stabilize double-­
stranded breaks induced by the bacterial type II enzymes (Fig. 16.4). The interca-
lated fluoroquinolones likely produce some distortions within the enzyme active
site; however, these drugs act primarily as “molecular doorstops” that form a physi-
cal barrier to DNA ligation [23]. Thus, the presence of fluoroquinolones inhibits the
rate of gyrase- and topoisomerase IV-mediated DNA ligation. Furthermore, drugs
that induce higher levels of enzyme-mediated DNA strand breaks appear to form
more stable interactions within the cleavage complex and allow these complexes to
persist for longer periods of time [63, 64].
In addition to generating DNA strand breaks in the cell, fluoroquinolones also
inhibit the overall catalytic strand-passage activities of gyrase and topoisomerase IV
[3, 11, 12]. As a result, there is debate as to whether the inhibition of strand passage
contributes to drug efficacy in cells. Although this issue has yet to be definitively
decided, a recent study suggests that the deleterious actions of drugs result primarily
from the enhancement of DNA cleavage [65]. In this study, the effects of ciprofloxa-
cin on three different fluoroquinolone-resistant mutations of Escherichia coli
­topoisomerase IV that are associated with clinical resistance were examined in vitro.
With all three enzymes, ciprofloxacin displayed virtually no ability to enhance DNA
514 E. G. Gibson et al.

Fig. 16.4 Crystal structure of a topoisomerase IV-DNA cleavage complex formed with A. bau-
mannii enzyme in the presence of moxifloxacin. This structure is a top view of the cleavage com-
plex with two fluoroquinolone molecules intercalating four base pairs apart at the cleaved scissile
bonds. The presence of the intercalated fluoroquinolones likely produces distortions within the
enzyme active site; however, these drugs act primarily as “molecular doorstops” that form a physi-
cal barrier to ligation. The catalytic core of the enzyme (blue and green for the A and B subunit,
respectively), moxifloxacin (red), and DNA (yellow) are shown. (Adapted from Aldred et al. [11])

cleavage but showed wild-type ability to inhibit DNA relaxation catalyzed by the
type II enzymes. Therefore, it appears that the ability to induce DNA cleavage is the
primary factor that determines quinolone-induced cytotoxicity.

16.5 Fluoroquinolone Resistance

The World Health Organization (WHO) ranks fluoroquinolones as one of the five
“highest priority” and “critically important” classes of antimicrobials [66]. However,
due to their widespread use and overuse, resistance has been on the rise since the
1990s [1, 2, 11]. As an extreme example, the CDC has classified Neisseria gonor-
rhoeae, the causative agent of gonorrhea, as one of its top three “urgent level” drug-­
resistant threats to the United States [67], primarily due to fluoroquinolone
resistance. Along with the WHO, it has issued dire warnings that gonorrhea is on the
precipice of joining HIV/AIDS and herpes as the third “incurable” sexually trans-
mitted disease [68].
Fluoroquinolones were used routinely to treat gonorrhea starting in 1993 and
were used in more than 40% of the cases by the year 2003 [69–71]. However, the
use of fluoroquinolones as frontline therapy against this disease was discontinued in
2006 due to the high incidence of resistance; 22.4% of cases reported in the United
16 Bacterial Type II Topoisomerases and Target-Mediated... 515

States in 2015 were resistant to fluoroquinolones (this value rose to 32.1% among
men who have sex with men) [69, 72, 73]. In parts of Asia, fluoroquinolone resis-
tance exceeds 80% [73]. Other infectious bacteria that have raised concerns due to
their high level of fluoroquinolone resistance include Campylobacter spp.,
Salmonella spp., and E. coli [74].
Thus far, three mechanisms of fluoroquinolone resistance have been described
[11, 58]. The first is “target-mediated resistance,” which results from specific muta-
tions in gyrase or topoisomerase IV [75–77]. The second is “plasmid-mediated
resistance,” which is caused by the presence of extrachromosomal DNA fragments
that encode three different classes of proteins [11, 78, 79]. Some plasmids encode
acetylases, which modify and inactivate quinolones and other drugs. Others encode
Qnr proteins, which block type II topoisomerases from binding to their DNA sub-
strates or to fluoroquinolones. Still others encode efflux pumps, which decrease the
fluoroquinolone concentration in cells. The third mechanism of fluoroquinolone
resistance is “chromosome mediated,” in which the expression of efflux pumps is
elevated or the expression of porins, which play a role in fluoroquinolone uptake, is
downregulated [11, 26, 78, 80].
Although the latter two mechanisms contribute significantly to fluoroquinolone
resistance, the target-mediated mechanism is generally the form most often associ-
ated with clinical resistance [11, 81, 82]. Because target-mediated resistance repre-
sents the most common and clinically relevant form of resistance, the remainder of
this chapter will focus on this mechanism.
Initial quinolone resistance is almost always associated with specific mutations
in gyrase, topoisomerase IV, or both. For example, in a recent clinical study on drug
resistance [83], 97% of 60 quinolone-resistant isolates of E. coli carried mutations
in gyrase, and 90% of these isolates also carried mutations in topoisomerase IV.
In general, the most commonly observed (up to ~90%) fluoroquinolone resis-
tance mutation is in a highly conserved serine residue that was first described as
Ser83 in the A subunit of E. coli gyrase [84–88]. This residue resides in helix-IV of
GyrA. The majority of other resistance mutations usually map to a conserved glu-
tamic/aspartic acid residue that is four amino acids downstream from the serine and
also resides in helix-IV. Mutations at these positions often provide a tenfold or
higher reduction in susceptibility to clinically relevant fluoroquinolones.
Corresponding mutations in E. coli topoisomerase IV also result in fluoroquinolone
resistance in vitro [85–88].
The prevalence of resistance mutations at the serine residue may reflect the fact
that this residue is highly conserved but nonessential. To this point, the common
mutations at this residue display no known phenotype, in cells or in vitro, with the
exception of fluoroquinolone resistance. It is not clear why this residue is conserved;
however, the presence of the serine appears to provide protection against nybomy-
cin, a naturally occurring antibiotic [89]. Thus, it has been proposed there has been
natural selection to maintain the serine in the bacterial genome. It is notable that
mutations at the glutamic/aspartic residue often decrease the overall catalytic
­activity of gyrase and topoisomerase IV [77, 90]. This may explain why a higher
proportion of resistance mutations are observed at the serine residue.
516 E. G. Gibson et al.

To determine the contributions of gyrase and topoisomerase IV to fluoroquino-

lone resistance in cells, E. coli strains carrying these mutations in gyrase, topoisom-
erase IV, or both were analyzed for drug efficacy. Strains carrying mutant gyrase
were ~10-fold less susceptible to fluoroquinolones. Although strains carrying
mutant topoisomerase IV displayed little, if any, resistance, those carrying muta-
tions in both enzymes had ~100-fold decrease in susceptibility [2, 3, 9, 58]. This
pattern of resistance strongly suggests that gyrase is the primary toxic target for
fluoroquinolones in E. coli (Gram-negative), and topoisomerase IV is a secondary
target for the drugs.
Since that initial set of experiments, the primary cellular target for fluoroquino-
lones in all other species has been identified by mutagenesis studies [11, 91–96].
The enzyme in which the first resistance mutations appear is believed to be the pri-
mary toxic target. Surprisingly, when these studies were carried out in Streptococcus
pneumoniae, a Gram-positive species, the first mutations appeared in topoisomerase
IV [94]. Thus, it became dogma in the field that gyrase was the primary target for
fluoroquinolones in Gram-negative species and topoisomerase IV was the primary
target in Gram-positive species. While this axiom generally holds true, subsequent
studies have found that there are often exceptions and that the target has to be deter-
mined on a species-by-species and drug-by-drug basis [21, 63, 65, 90, 97, 98].

16.6  ole of the Water-Metal Ion Bridge in Mediating

R
Fluoroquinolone Resistance and Gyrase/Topoisomerase
IV Interactions

Although the association of the serine and glutamic/aspartic residues with fluoro-
quinolone resistance was established in the late 1980s [99–102], the mechanism by
which they lead to resistance was described only recently. Ultimately, the mechanis-
tic basis for fluoroquinolone action and resistance turned out to be inextricably
linked [11, 63, 65, 103]. Thus, these two important aspects of fluoroquinolone-­
enzyme interaction will be discussed together.
The initial insight into the roles of the serine and glutamic/aspartic acid residues
of fluoroquinolone actions and resistance came from structural studies of cleavage
complexes formed with topoisomerase IV or gyrase in the presence of fluoroquino-
lones [23, 60–62]. Although these studies all localized fluoroquinolones in the same
binding pocket, which was proximal to the conserved amino acid residues, there
was disagreement regarding drug orientation within the pocket. Furthermore, none
of the studies found that the bound fluoroquinolone was close enough to either
amino acid to form a direct interaction.
However, one of the structures (which examined the cleavage complex of
Acinetobacter baumannii topoisomerase IV formed in the presence of m ­ oxifloxacin)
provided a potential mechanism by which mutations at the serine or glutamic/aspar-
tic residue could lead to fluoroquinolone resistance [23]. It had long been known
16 Bacterial Type II Topoisomerases and Target-Mediated... 517

Glu/Asp

O
O

HO Ser

Mg+2

O O
F
5 4
6 3 O
7
8 1 2
R1 N
R2

Fig. 16.5 A water-metal ion bridge mediates critical interactions between fluoroquinolones and
bacterial type II topoisomerases. A generic fluoroquinolone structure is depicted in black, water
molecules are in blue, Mg2+ is in orange, and the coordinating serine and glutamic/aspartic acid
residues are in green. Blue dashed lines indicate the interaction between the divalent metal ion,
four water molecules, and the C3/C4 keto acid of the fluoroquinolone. The green dashed lines
represent hydrogen bonds between the serine and glutamic/aspartic acid side-chain hydroxyl
groups and the water molecules

that the C3/C4 keto acid of fluoroquinolones chelates divalent metal ions, but the
physiological role of these bound metal ions, if any, was unknown. The structure of
A. baumannii topoisomerase IV was the first to capture this fluoroquinolone-­metal
ion interaction within a cleavage complex. In this structure, the C3/C4 keto acid of
moxifloxacin chelated a non-catalytic magnesium ion that appeared to be coordi-
nated to four water molecules. Two of these water molecules were in sufficiently
close proximity to Ser84 and Glu88 (equivalent to E. coli GyrA Ser83 and Glu87)
to form hydrogen bonds. Thus, the authors suggested that this water-metal ion coor-
dination might play a role in mediating interactions between fluoroquinolones and
bacterial type II topoisomerases. A subsequent study that determined the structures
of cleavage complexes formed with M. tuberculosis gyrase in the presence of moxi-
floxacin, 8-methyl-moxifloxacin, ciprofloxacin, levofloxacin, or gatifloxacin also
observed the chelated metal ion, the associated water molecules, and the protein
contacts [64]. A generalized diagram of the proposed water-metal ion “bridge” that
facilitates fluoroquinolone interactions with the conserved serine and glutamic/
aspartic residues is shown in Fig. 16.5 [22, 65, 90].
The initial functional evidence for the existence and role for the water-metal ion
bridge in mediating fluoroquinolone activity and resistance came from biochemical
studies on B. anthracis topoisomerase IV [90]. These studies utilized wild-type and
drug-resistant enzymes that carried mutations in the serine (Ser81) and/or glutamic
acid (Glu85) residues. The authors demonstrated that (1) the ability of
­fluoroquinolones to poison topoisomerase IV relied on the presence of a non-cata-
lytic divalent metal ion; (2) mutations in either the serine or glutamic acid restricted
518 E. G. Gibson et al.

the metal ions that could be used to support drug activity; and (3) mutations in either
amino acid decreased the affinity of the metal ion. Later studies extended these
conclusions to topoisomerase IV from E. coli and gyrase from B. anthracis and M.
tuberculosis [11, 63, 65]. Thus, it appears that the water-metal ion bridge is used to
mediate fluoroquinolone-enzyme interactions in a variety of bacterial species.
Furthermore, the loss of one or both of the amino acids that anchor the bridge is
sufficient to disrupt these interactions and cause drug resistance [22, 63, 65].
Despite the importance and apparent “universality” of the water-metal ion bridge,
it seems to be used differently by enzymes from different bacterial species.
Whereas the bridge is critical for the binding of clinically relevant fluoroquinolones
to B. anthracis gyrase and topoisomerase IV and M. tuberculosis gyrase, it is used
primarily to align fluoroquinolones in the active site of E. coli topoisomerase IV
[11, 63, 65, 90].
The divalent metal ion of the water-metal ion bridge interacts with fluoroquino-
lones through the C3/C4 keto acid of the drug scaffold [63, 65, 90, 97]. This may
explain why clinically relevant fluoroquinolones can accommodate such a wide
variety of substituents at the N1, C7, and C8 positions. Whereas substituents at the
latter positions are unlikely to form critical gyrase or topoisomerase IV interactions,
they may contribute minor or species-specific interactions. Furthermore, they may
influence the pharmacokinetics of the drugs.
Finally, the water-metal ion bridge appears to be the feature of drug-enzyme
interactions that allows discrimination between the bacterial and human type II
topoisomerases. Indeed, the amino acids in human topoisomerase IIα that corre-
spond to the serine and acidic residues of the bacterial helix-IV are methionine resi-
dues. This likely explains why clinically relevant fluoroquinolones display such poor
activity against the human type II enzymes. If these methionine residues in topoi-
somerase IIα are converted to serine and glutamic acid residues, the activity of cip-
rofloxacin and moxifloxacin against the human enzyme rises four- to fivefold [90].

16.7 Overcoming Target-Mediated Fluoroquinolone

Resistance: Modified Fluoroquinolones
and Fluoroquinolone-Like Compounds

16.7.1 C7 Substituents

The influence of the C7 substituent on fluoroquinolone resistance is highlighted by

recent studies on quinazolinediones. These compounds are similar in structure to
fluoroquinolones but display a strong ability to overcome resistance caused by
mutations in the amino acid residues that anchor the water-metal ion bridge [87]
(Fig. 16.6). The quinazolinedione scaffold differs from that of fluoroquinolones
only at the 2 and 3 ring positions, where the hydrogen at C2 has been replaced with
a ketone and the carboxylic acid at C3 has been replaced with an N3 amino group.
16 Bacterial Type II Topoisomerases and Target-Mediated... 519

Fig. 16.6 Fluoroquinolone-like compounds with unique properties. The quinazolinedione scaf-
fold is similar to the fluoroquinolone core; however, the loss of the C3/C4 keto acid disrupts the
ability to chelate the divalent metal ion used in the water-metal ion bridge. 8-Methyl-2,4-dione, a
quinazolinedione, and 8-methyl-moxifloxacin, a fluoroquinolone, overcome resistance by mediat-
ing interactions with bacterial type II enzymes through their C7 and C8 substituents, respectively.
CP 115,955, a fluoroquinolone, displays high activity against human (through the C7 substituent)
and bacterial (through the water-metal ion bridge) type II topoisomerases

These substitutions disrupt the C3/C4 keto acid required to chelate the divalent
metal ion used in the water-metal ion bridge. Thus, it is not surprising that interac-
tions between quinazolinediones and bacterial type II topoisomerases are indepen-
dent of bridge function and are unaffected by resistance mutations in the serine or
acidic amino acid residues that serve as bridge anchors. Consequently, it was
believed that quinazolinediones represented “fluoroquinolones” with a scaffold that
was impervious to classic resistance mutations [87].
However, three lines of evidence lead to the conclusion that the ability of quin-
azolinediones to overcome resistance results from the substituent at C7 rather than
the C3/C4 portion [22, 75]. First, the quinazolinediones reported in the literature
most often contained a 3′-(aminomethyl)-1-pyrrolidinyl [3′-(AM)P] (or related)
substituent at the C7 position. The 3′-(AM)P moiety is not represented in any clini-
cally relevant fluoroquinolone, opening the possibility that it has properties not pre-
520 E. G. Gibson et al.

viously ascribed to fluoroquinolone C7 groups. Second, when the C7 piperazine

group of ciprofloxacin was replaced with the 3′-(AM)P moiety, the resulting fluoro-
quinolone overcame resistance caused by mutations in the bridge anchoring amino
acid residues in a purified enzyme system [11, 63, 65, 90, 104]. Third, when the C7
3′-(AM)P substituent of the quinazolinedione was replaced with the C7 group of
either ciprofloxacin (piperazine) or moxifloxacin (diazabicyclonone), the resulting
drugs displayed little activity against either wild-type or resistant bacterial type II
topoisomerases [22, 75, 104].
The evidence described above indicates that the quinazolinedione scaffold does
not interact with gyrase and topoisomerase IV through metal ion-independent con-
tacts. Rather, these drugs are essentially quinolone derivatives that lack their most
important interactions with bacterial type II topoisomerases (i.e., the water-metal
ion bridge). These findings ultimately led to a very important conclusion with impli-
cations for future design of fluoroquinolones that overcome resistance: it is possible
to design C7 substituents for fluoroquinolones that display strong, bridge-­
independent interactions with gyrase and topoisomerase IV.
The role of the C7 substituent in mediating enzyme interactions has long been
known for the interaction of fluoroquinolones with eukaryotic type II topoisomer-
ases [105, 106]. Indeed, fluoroquinolones such as CP-115,953 and CP-115,955 dis-
play high activity against type II topoisomerase and are more potent and efficacious
against human topoisomerase IIα and IIβ than the anticancer drug etoposide [103].
Both of these compounds rely on a 4′-hydroxyphenyl substituent at the C7 position
for their activity [105, 106].
On the basis of structural and modeling studies, it has been proposed that the C7
3′-(AM)P moiety allows quinazolinediones to poison gyrase and topoisomerase IV
by interacting with a conserved glutamic acid residue in the GyrB and ParE/GrlB
subunit (corresponding to E. coli GyrB-Glu466) [87, 104]. Indeed, it appears that
the primary amine of the 3′-(AM)P substituent can form both a salt bridge and a
hydrogen bond with this acidic residue. Unfortunately, this glutamic acid residue is
also conserved among eukaryotic type II topoisomerases, and the quinazolinediones
that include the C7 3′-(AM)P moiety display activity against human topoisomerase
IIα similar to that of etoposide [22]. Thus, fluoroquinolone substitutions at C7 that
overcome resistance should be approached with caution, as they have the potential
to crossover into the human system.

16.7.2 C8 Substituents

Recent studies strongly suggest that substituents at the C8 position can have dra-
matic effects on fluoroquinolone resistance [63]. At the present time, structure-­
activity relationship studies that examined the effects of C8 substituents on resistance
have been confined to relatively minor changes at this position: hydrogen, methyl,
or methoxy groups. However, major effects on resistance have been observed. In
general, compounds that include a methyl or methoxy group at C8 display higher
16 Bacterial Type II Topoisomerases and Target-Mediated... 521

activity against enzymes that carry mutations in the bridge-anchoring serine or glu-
tamic/aspartic acid. In some cases, dramatic differences in sensitivity have been
reported. For example, converting the C8 methoxy of moxifloxacin to a methyl
group results in a fluoroquinolone that poisons M. tuberculosis gyrase with twice
the potency and efficacy of moxifloxacin and completely overcomes clinically rel-
evant resistance mutations in a purified enzyme system [63, 64]. The fact that such
a minor alteration in fluoroquinolone structure can produce such a dramatic differ-
ence in the resistance profile of the drug suggests that the C8 position is a ripe target
for future drug discovery.
Despite the potential of C8 substituents for overcoming drug resistance, the basis
for the high activity of “8-methyl-moxifloxacin” against fluoroquinolone-resistant
M. tuberculosis gyrase is unknown. Although this compound induces gyrase-­
mediated DNA strand breaks that are more stable than observed with moxifloxacin,
structural studies indicate that 8-methyl-moxifloxacin occupies a space within the
cleavage complex that is identical to that of the parent drug [63, 64]. Furthermore,
no specific protein or DNA contacts were observed with the C8 substituent. Thus,
future chemical studies will need to be combined with strong efforts in mechanistic
enzymology and structure in order to fully exploit the C8 substituent as a means to
overcome fluoroquinolone resistance.

16.8 Overcoming Target-Mediated Fluoroquinolone

Resistance: Novel Compounds

Currently, fluoroquinolones are the only antibacterials in clinical use that target
gyrase or topoisomerase IV [9, 11, 23, 60–62]. However, recent drug discovery
efforts have resulted in new classes (two of which are in clinical trials) with clinical
potential. All of these compounds lack the keto acid that fluoroquinolones use in
conjunction with the water-metal ion bridge to interact with their bacterial targets.
Consequently, they all display activity against fluoroquinolone-resistant bacterial
strains.

16.8.1 Novel Bacterial Topoisomerase Inhibitors (NBTIs)

NBTIs (Fig. 16.7) are napthyridone/aminopiperidine-based compounds that were

first reported to have antibacterial activity in 1999 [107]. It was not until 2007 that
these compounds were found to have activity against bacterial type II topoisomer-
ases [108]. Early studies demonstrated that at least some of the NBTIs are potent
inhibitors of overall catalytic activity [109]. Later studies determined that some of
these compounds could also poison the enzymes [60].
522 E. G. Gibson et al.

Fig. 16.7 Novel gyrase/topoisomerase IV-targeted compounds. NBTIs, such as gepotidacin, act
as gyrase/topoisomerase IV poisons. However, in contrast to fluoroquinolones, they induce only
single-stranded DNA breaks. MGIs, such as GSK000, were derived from NBTIs in an effort to
optimize activity against M. tuberculosis. The founding member of the spiropyrimidinetrione class
of antibacterials is zoliflodacin (ETX0914/AZD0914)

Compared to fluoroquinolones, NBTIs are distinct in two major respects. First,

structural studies demonstrate that only a single NBTI molecule interacts with the
DNA in the active site of gyrase. It binds between the two scissile bonds and elon-
gates the DNA in the active site of the enzyme. This is in contrast to the two fluoro-
quinolones (one at each cut scissile bond) that interact with DNA in the cleavage
complex. Second, whereas fluoroquinolones stabilize double-stranded DNA breaks
generated by gyrase or topoisomerase IV, NBTIs that act as gyrase/topoisomerase
IV poisons induce only single-stranded DNA breaks. Little else is known about how
these compounds interact with gyrase or topoisomerase IV. NBTIs display high
activity against bacterial cells that harbor fluoroquinolone-resistant mutations in
gyrase and topoisomerase IV. However, no study examining purified fluoroquinolone-­
resistant mutant enzymes has been reported. One member of the NBTI family,
gepotidacin, is currently in phase II clinical trials for the treatment of uncomplicated
gonorrhea [110].

16.8.2 Mycobacterium tuberculosis Gyrase Inhibitors (MGIs)

MGIs (Fig. 16.7) were derived from NBTIs in an effort to optimize activity against
M. tuberculosis [111]. These compounds display high activity against wild-type and
fluoroquinolone-resistant strains. On the basis of mutagenesis studies, MGIs are
16 Bacterial Type II Topoisomerases and Target-Mediated... 523

believed to target gyrase, the only type II topoisomerase in M. tuberculosis. No cor-

responding in vitro study has been reported to date.

16.8.3 Spiropyrimidinetriones

Spiropyrimidinetriones (Fig. 16.7) are a novel class of gyrase/topoisomerase IV

poisons. Similar to fluoroquinolones, these enzymes induce enzyme-mediated
double-­stranded DNA breaks [112]. The founding member of this class, zolifloda-
cin (ETX0914/AZD0914), maintains activity against multidrug-resistant
Pseudomonas aeruginosa [113], which contains fluoroquinolone-resistant muta-
tions in the bridge-anchoring residues in both gyrase and topoisomerase IV. The
drug is currently in phase II clinical trials for the treatment of uncomplicated gonor-
rhea [114].

16.9 Conclusions

Fluoroquinolones are one of the most important and widely prescribed classes of
antibacterials in clinical use. However, their usefulness is being eroded by the rise
of drug resistance. Of the mechanisms that impair fluoroquinolone actions, those
that result from mutations in gyrase and topoisomerase IV are the most common
and detrimental. The mechanistic studies described above have led to a more com-
plete understanding of how fluoroquinolones interact with their enzyme targets and
how mutations alter these interactions. Furthermore, these studies have suggested
new strategies for overcoming resistance. Among these strategies are the design of
novel fluoroquinolones and the development of new drug classes that do not rely on
the water-metal ion bridge for their actions. Hopefully, these approaches will allow
gyrase and topoisomerase IV to remain important antibacterial targets in the decades
to come.
Major Points
• Although fluoroquinolones are the most efficacious and broad-spectrum oral
antibacterials in the clinic, their use is being eroded by resistance.
• Fluoroquinolone binding to gyrase and topoisomerase IV, the cellular targets of
these drugs, involves a water-metal ion bridge that is anchored by a keto acid on
the drug and a highly conserved serine and glutamic/aspartic acid residue in the
enzyme.
• Substitutions in the serine and glutamic/aspartic amino acid residues are respon-
sible for most of the target-mediated fluoroquinolone resistance.
• The development of compounds that interact with gyrase and topoisomerase IV
through mechanisms that are independent of the water-metal ion bridge may
provide an approach to bypassing existing target-mediated resistance.
524 E. G. Gibson et al.

Acknowledgments The authors of this article were supported by the US Veterans Administration
(Merit Review Award I01 Bx002198 to N.O.), the National Institutes of Health (grants AI87671 to
R.J.K., GM33944 and GM126363 to N.O., and GM007628 to E.G.G.), the National Science
Foundation (grant DGE-0909667 to R.E.A.), the Pharmaceutical Research and Manufacturers of
America Foundation (to E.G.G.), and the American Association of Pharmaceutical Scientists
Foundation (to E.G.G.).

References

1. Andriole VT. The quinolones: past, present, and future. Clin Infect Dis. 2005;41(Suppl.
2):S113–9.
2. Drlica K, Hiasa H, Kerns R, Malik M, Mustaev A, Zhao X. Quinolones: action and resistance
updated. Curr Top Med Chem. 2009;9(11):981–98.
3. Hooper DC. Mechanisms of action of antimicrobials: focus on fluoroquinolones. Clin Infect
Dis. 2001;32(Suppl. 1):S9–S15.
4. Linder JA, Huang ES, Steinman MA, Gonzales R, Stafford RS. Fluoroquinolone prescribing
in the United States: 1995 to 2002. Am J Med. 2005;118(3):259–68.
5. Hooper DC. Clinical applications of quinolones. Biochim Biophys Acta. 1998;1400:45–61.
6. Ressel G. Centers for Disease Control and Prevention. CDC updates interim guide-
lines for anthrax exposure management and antimicrobial therapy. Am Fam Physician.
2001;64(11):1901–2. 4
7. Jeon D. WHO treatment guidelines for drug-resistant tuberculosis, 2016 update: applicability
in South Korea. Tuberc Respir Dis (Seoul). 2017;80(4):336–43.
8. WHO. Global tuberculosis report 2016. 2016.
9. Anderson VE, Osheroff N. Type II topoisomerases as targets for quinolone antibacterials:
turning Dr. Jekyll into Mr. Hyde. Curr Pharm Des. 2001;7(5):337–53.
10. Champoux JJ. DNA topoisomerases: structure, function, and mechanism. Annu Rev
Biochem. 2001;70:369–413.
11. Aldred KJ, Kerns RJ, Osheroff N. Mechanism of quinolone action and resistance.
Biochemistry. 2014;53(10):1565–74.
12. Bush NG, Evans-Roberts K, Maxwell A. DNA topoisomerases. EcoSal Plus. 2015;6(2)
13. Levine C, Hiasa H, Marians KJ. DNA gyrase and topoisomerase IV: biochemical activities,
physiological roles during chromosome replication, and drug sensitivities. Biochim Biophys
Acta. 1998;1400:29–43.
14. Deweese JE, Burch AM, Burgin AB, Osheroff N. Use of divalent metal ions in the DNA
cleavage reaction of human type II topoisomerases. Biochemistry. 2009;48(9):1862–9.
15. Gentry AC. DNA topoisomerases: type II. In: Lennarz WJ, Daniel Lane M, editors. The
encyclopedia of biological chemistry. Waltham: Academic; 2013. p. 163–8.
16. Fraser CM, Norris SJ, Weinstock GM, White O, Sutton GG, Dodson R, et al. Complete genome
sequence of Treponema pallidum, the syphilis spirochete. Science. 1998;281(5375):375–88.
17. Tomb JF, White O, Kerlavage AR, Clayton RA, Sutton GG, Fleischmann RD, et al.
The complete genome sequence of the gastric pathogen Helicobacter pylori. Nature.
1997;388(6642):539–47.
18. Parkhill J, Wren BW, Mungall K, Ketley JM, Churcher C, Basham D, et al. The genome
sequence of the food-borne pathogen Campylobacter jejuni reveals hypervariable sequences.
Nature. 2000;403(6770):665–8.
19. Cole ST, Eiglmeier K, Parkhill J, James KD, Thomson NR, Wheeler PR, et al. Massive gene
decay in the leprosy bacillus. Nature. 2001;409(6823):1007–11.
20. Cole ST, Barrell BG. Analysis of the genome of Mycobacterium tuberculosis H37Rv.
Novartis Found Symp. 1998;217:160–72. discussion 72–7
16 Bacterial Type II Topoisomerases and Target-Mediated... 525

21. Ashley RE, Blower TR, Berger JM, Osheroff N. Recognition of DNA supercoil geometry by
Mycobacterium tuberculosis gyrase. Biochemistry. 2017;56:5440.
22. Aldred KJ, Schwanz HA, Li G, McPherson SA, Turnbough CL, Kerns RJ, et al. Overcoming
target-mediated quinolone resistance in topoisomerase IV by introducing metal-ion-­
independent drug–enzyme interactions. ACS Chem Biol. 2013;8(12):2660–8.
23. Wohlkonig A, Chan PF, Fosberry AP, Homes P, Huang J, Kranz M, et al. Structural basis of
quinolone inhibition of type IIA topoisomerases and target-mediated resistance. Nat Struct
Mol Biol. 2010;17(9):1152–3.
24. Lesher GY, Froelich EJ, Gruett MD, Bailey JH, Brundage RP. 1,8-Naphthyridine derivatives.
A new class of chemotherapeutic agents. J Med Pharm Chem. 1962;91:1063–5.
25. Emmerson AM, Jones AM. The quinolones: decades of development and use. J Antimicrob
Chemother. 2003;51(Suppl 1):13–20.
26. Mitscher LA. Bacterial topoisomerase inhibitors: quinolone and pyridone antibacterial
agents. Chem Rev. 2005;105(2):559–92.
27. Stein GE. The 4-quinolone antibiotics: past, present, and future. Pharmacotherapy.
1988;8(6):301–14.
28. Drusano G, Labro MT, Cars O, Mendes P, Shah P, Sorgel F, et al. Pharmacokinetics and
pharmacodynamics of fluoroquinolones. Clin Microbiol Infect. 1998;4(Suppl 2):S27–41.
29. Nitiss JL. Targeting DNA topoisomerase II in cancer chemotherapy. Nat Rev Cancer.
2009;9(5):338–50.
30. Deweese JE, Osheroff MA, Osheroff N. DNA topology and topoisomerases: teaching a
“knotty” subject. Biochem Mol Biol Educ. 2008;37(1):2–10.
31. Vos SM, Tretter EM, Schmidt BH, Berger JM. All tangled up: how cells direct, manage and
exploit topoisomerase function. Nat Rev Mol Cell Biol. 2011;12(12):827–41.
32. Wang JC. Cellular roles of DNA topoisomerases: a molecular perspective. Nat Rev Mol Cell
Biol. 2002;3(6):430–40.
33. Liu Z, Deibler RW, Chan HS, Zechiedrich L. The why and how of DNA unlinking. Nucleic
Acids Res. 2009;37(3):661–71.
34. Deweese JE, Osheroff N. The DNA cleavage reaction of topoisomerase II: wolf in sheep’s
clothing. Nucleic Acids Res. 2009;37(3):738–48.
35. Pommier Y, Leo E, Zhang H, Marchand C. DNA topoisomerases and their poisoning by anti-
cancer and antibacterial drugs. Chem Biol. 2010;17(5):421–33.
36. Gellert M, Mizuuchi K, O'Dea MH, Nash HA. DNA gyrase: an enzyme that introduces super-
helical turns into DNA. Proc Natl Acad Sci U S A. 1976;73:3872–6.
37. Bates AD, Berger JM, Maxwell A. The ancestral role of ATP hydrolysis in type II topoisom-
erases: prevention of DNA double-strand breaks. Nucleic Acids Res. 2011;39(15):6327–39.
38. Gentry AC, Osheroff N. DNA topoisomerases: type II. In:Encyclopedia of biological chem-
istry. Amsterdam: Elsevier Inc.; 2013. p. 163–8.
39. Corbett KD, Berger JM. Structure, molecular mechanisms, and evolutionary relationships in
DNA topoisomerases. Annu Rev Biophys Biomol Struct. 2004;33:95–118.
40. Hiasa H, Marians KJ. Topoisomerase IV can support oriC DNA replication in vitro. J Biol
Chem. 1994;269(23):16371–5.
41. Zechiedrich EL, Khodursky AB, Bachellier S, Schneider R, Chen D, Lilley DM, et al. Roles
of topoisomerases in maintaining steady-state DNA supercoiling in Escherichia coli. J Biol
Chem. 2000;275(11):8103–13.
42. Crisona NJ, Strick TR, Bensimon D, Croquette V, Cozzarelli NR. Preferential relaxation of
positively supercoiled DNA by E. coli topoisomerase IV in single-molecule and ensemble
measurements. Genes Dev. 2000;14(22):2881–92.
43. Wang X, Reyes-Lamothe R, Sherratt DJ. Modulation of Escherichia coli sister chromosome
cohesion by topoisomerase IV. Genes Dev. 2008;22(17):2426–33.
44. Joshi MC, Magnan D, Montminy TP, Lies M, Stepankiw N, Bates D. Regulation of sis-
ter chromosome cohesion by the replication fork tracking protein SeqA. PLoS Genet.
2013;9(8):e1003673.
526 E. G. Gibson et al.

45. Zawadzki P, Stracy M, Ginda K, Zawadzka K, Lesterlin C, Kapanidis AN, et al. The localiza-
tion and action of topoisomerase IV in Escherichia coli chromosome segregation is coordi-
nated by the SMC complex, MukBEF. Cell Rep. 2015;13(11):2587–96.
46. Pendleton M, Lindsey RH Jr, Felix CA, Grimwade D, Osheroff N. Topoisomerase II and
leukemia. Ann N Y Acad Sci. 2014;1310:98–110.
47. Liu LF, Wang JC. DNA-DNA gyrase complex: the wrapping of the DNA duplex outside the
enzyme. Cell. 1978;15(3):979–84.
48. Kampranis SC, Maxwell A. Conversion of DNA gyrase into a conventional type II topoisom-
erase. Proc Natl Acad Sci U S A. 1996;93(25):14416–21.
49. Kramlinger VM, Hiasa H. The “GyrA-box” is required for the ability of DNA gyrase to wrap
DNA and catalyze the supercoiling reaction. J Biol Chem. 2006;281(6):3738–42.
50. Ullsperger C, Cozzarelli NR. Contrasting enzymatic acitivites of topoisomerase IV and DNA
gyrase from Escherichia coli. J Biol Chem. 1996;271:31549–55.
51. Marians KJ. DNA gyrase-catalyzed decatenation of multiply linked DNA dimers. J Biol
Chem. 1987;262(21):10362–8.
52. Brown PO, Cozzarelli NR. A sign inversion mechanism for enzymatic supercoiling of
DNA. Science. 1979;206(4422):1081–3.
53. Drlica K, Snyder M. Superhelical Escherichia coli DNA: relaxation by coumermycin. J Mol
Biol. 1978;120(2):145–54.
54. Koster DA, Crut A, Shuman S, Bjornsti MA, Dekker NH. Cellular strategies for regulating
DNA supercoiling: a single-molecule perspective. Cell. 2010;142(4):519–30.
55. Pruss GJ, Manes SH, Drlica K. Escherichia coli DNA topoisomerase I mutants: increased
supercoiling is corrected by mutations near gyrase genes. Cell. 1982;31(1):35–42.
56. Menzel R, Gellert M. Regulation of the genes for E. coli DNA gyrase: homeostatic control of
DNA supercoiling. Cell. 1983;34(1):105–13.
57. Drlica K, Malik M, Kerns RJ, Zhao X. Quinolone-mediated bacterial death. Antimicrob
Agents Chemother. 2008;52(2):385–92.
58. Hooper DC. Mechanisms of fluoroquinolone resistance. Drug Resist Updat. 1999;2(1):38–55.
59. Kreuzer KN, Cozzarelli NR. Escherichia coli mutants thermosensitive for deoxyribonucleic
acid gyrase subunit A: effects on deoxyribonucleic acid replication, transcription, and bacte-
riophage growth. J Bacteriol. 1979;140(2):424–35.
60. Bax BD, Chan PF, Eggleston DS, Fosberry A, Gentry DR, Gorrec F, et al. Type IIA topoisom-
erase inhibition by a new class of antibacterial agents. Nature. 2010;466(7309):935–40.
61. Laponogov I, Pan XS, Veselkov DA, McAuley KE, Fisher LM, Sanderson MR. Structural
basis of gate-DNA breakage and resealing by type II topoisomerases. PLoS One.
2010;5(6):e11338.
62. Laponogov I, Sohi MK, Veselkov DA, Pan XS, Sawhney R, Thompson AW, et al. Structural
insight into the quinolone-DNA cleavage complex of type IIA topoisomerases. Nat Struct
Mol Biol. 2009;16(6):667–9.
63. Aldred KJ, Blower TR, Kerns RJ, Berger JM, Osheroff N. Fluoroquinolone interactions with
Mycobacterium tuberculosis gyrase: enhancing drug activity against wild-type and resistant
gyrase. Proc Natl Acad Sci U S A. 2016;113(7):E839–46.
64. Blower TR, Williamson BH, Kerns RJ, Berger JM. Crystal structure and stability of gyrase-­
fluoroquinolone cleaved complexes from Mycobacterium tuberculosis. Proc Natl Acad Sci U
S A. 2016;113(7):1706–13.
65. Aldred KJ, Breland EJ, Vlčková V, Strub MP, Neuman KC, Kerns RJ, et al. Role of the
water-metal ion bridge in mediating interactions between quinolones and Escherichia coli
topoisomerase IV. Biochemistry. 2014;53(34):5558–67.
66. Collignon PC, Conly JM, Andremont A, McEwen SA, Aidara-Kane A, World Health
Organization Advisory Group BMoISoAR, et al. World Health Organization ranking of anti-
microbials according to their importance in human medicine: a critical step for developing
risk management strategies to control antimicrobial resistance from food animal production.
Clin Infect Dis. 2016;63(8):1087–93.
16 Bacterial Type II Topoisomerases and Target-Mediated... 527

67. CDC. Antibiotic/antimicrobial resistance. 2017. Available from: https://www.cdc.gov/dru-

gresistance/biggest_threats.html.
68. Koebler J. World Health Organization warns gonorrhea could join HIV as ‘uncurable’. US
News and World Reports. 2012 June 6.
69. CDC. 2015 Sexually transmitted diseases surveillance. 2015. Available from: https://www.
cdc.gov/std/stats15/toc.htm.
70. Peterman TA, O'Connor K, Bradley HM, Torrone EA, Bernstein KT. Gonorrhea control,
United States, 1972-2015, a narrative review. Sex Transm Dis. 2016;43(12):725–30.
71. Morgan MK, Decker CF. Gonorrhea. Dis Mon. 2016;62(8):260–8.
72. Kirkcaldy RD, Harvey A, Papp JR, Del Rio C, Soge OO, Holmes KK, et al. Neisseria gon-
orrhoeae antimicrobial susceptibility surveillance – the Gonococcal Isolate Surveillance
Project, 27 sites, United States, 2014. MMWR Surveill Summ. 2016;65(7):1–19.
73. Chen PL, Lee HC, Yan JJ, Hsieh YH, Lee NY, Ko NY, et al. High prevalence of mutations in
quinolone-resistance-determining regions and mtrR loci in polyclonal Neisseria gonorrhoeae
isolates at a tertiary hospital in Southern Taiwan. J Formos Med Assoc. 2010;109(2):120–7.
74. WHO. Global priority list of antibioitic-resistant bacteria to guide research, discovery, and
development of new antibiotics. 2017.
75. Aldred KJ, McPherson SA, Wang P, Kerns RJ, Graves DE, Turnbough CL Jr, et al. Drug
interactions with Bacillus anthracis topoisomerase IV: biochemical basis for quinolone
action and resistance. Biochemistry. 2012;51(1):370–81.
76. Pan XS, Gould KA, Fisher LM. Probing the differential interactions of quinazolinedione PD
0305970 and quinolones with gyrase and topoisomerase IV. Antimicrob Agents Chemother.
2009;53(9):3822–31.
77. Hiasa H. The Glu-84 of the ParC subunit plays critical roles in both topoisomerase IV-quinolone
and topoisomerase IV-DNA interactions. Biochemistry. 2002;41(39):11779–85.
78. Robicsek A, Jacoby GA, Hooper DC. The worldwide emergence of plasmid-mediated quino-
lone resistance. Lancet Infect Dis. 2006;6(10):629–40.
79. Tran JH, Jacoby GA. Mechanism of plasmid-mediated quinolone resistance. Proc Natl Acad
Sci U S A. 2002;99(8):5638–42.
80. Tsai YK, Liou CH, Chang FY, Fung CP, Lin JC, Siu LK. Effects of different resistance mech-
anisms on susceptibility to different classes of antibiotics in Klebsiella pneumoniae strains:
a strategic system for the screening and activity testing of new antibiotics. J Antimicrob
Chemother. 2017;72:3302.
81. Munita JM, Arias CA. Mechanisms of antibiotic resistance. Microbiol Spectr. 2016;4(2)
82. Davies J, Davies D. Origins and evolution of antibiotic resistance. Microbiol Mol Biol Rev.
2010;74(3):417–33.
83. Morgan-Linnell SK, Becnel Boyd L, Steffen D, Zechiedrich L. Mechanisms account-
ing for fluoroquinolone resistance in Escherichia coli clinical isolates. Antimicrob Agents
Chemother. 2009;53(1):235–41.
84. Price LB, Vogler A, Pearson T, Busch JD, Schupp JM, Keim P. In vitro selection and char-
acterization of Bacillus anthracis mutants with high-level resistance to ciprofloxacin.
Antimicrob Agents Chemother. 2003;47(7):2362–5.
85. Malik M, Mustaev A, Schwanz HA, Luan G, Shah N, Oppegard LM, et al. Suppression
of gyrase-mediated resistance by C7 aryl fluoroquinolones. Nucleic Acids Res.
2016;44(7):3304–16.
86. Strahilevitz J, Robicsek A, Hooper DC. Role of the extended alpha4 domain of Staphylococcus
aureus gyrase a protein in determining low sensitivity to quinolones. Antimicrob Agents
Chemother. 2006;50(2):600–6.
87. Malik M, Marks KR, Mustaev A, Zhao X, Chavda K, Kerns RJ, et al. Fluoroquinolone and
quinazolinedione activities against wild-type and gyrase mutant strains of Mycobacterium
smegmatis. Antimicrob Agents Chemother. 2011;55(5):2335–43.
88. Pfeiffer ES, Hiasa H. Replacement of ParC alpha4 helix with that of GyrA increases the sta-
bility and cytotoxicity of topoisomerase IV-quinolone-DNA ternary complexes. Antimicrob
Agents Chemother. 2004;48(2):608–11.
528 E. G. Gibson et al.

89. Hiramatsu K, Igarashi M, Morimoto Y, Baba T, Umekita M, Akamatsu Y. Curing bacteria of

antibiotic resistance: reverse antibiotics, a novel class of antibiotics in nature. Int J Antimicrob
Agents. 2012;39(6):478–85.
90. Aldred KJ, McPherson SA, Turnbough CL Jr, Kerns RJ, Osheroff N. Topoisomerase
IV-quinolone interactions are mediated through a water-metal ion bridge: mechanistic basis
of quinolone resistance. Nucleic Acids Res. 2013;41(8):4628–39.
91. Sugino A, Peebles CL, Kreuzer KN, Cozzarelli NR. Mechanism of action of nalidixic acid:
purification of Escherichia coli nalA gene product and its relationship to DNA gyrase and a
novel nicking-closing enzyme. Proc Natl Acad Sci U S A. 1977;74(11):4767–71.
92. Gellert M, Mizuuchi K, O'Dea MH, Itoh T, Tomizawa J. Nalidixic acid resistance: a second
genetic character involved in DNA gyrase activity. Proc Natl Acad Sci U S A. 1977;74:4772–6.
93. Khodursky AB, Zechiedrich EL, Cozzarelli NR. Topoisomerase IV is a target of quinolones
in Escherichia coli. Proc Natl Acad Sci U S A. 1995;92:11801–5.
94. Pan X-S, Ambler J, Mehtar S, Fisher LM. Involvement of topoisomerase IV and DNA
gyrase as ciprofloxacin targets in Streptococcus pneumoniae. Antimicrob Agents Chemother.
1996;40(10):2321–6.
95. Fournier B, Zhao X, Lu T, Drlica K, Hooper DC. Selective targeting of topoisomerase IV and
DNA gyrase in Staphylococcus aureus: different patterns of quinolone-induced inhibition of
DNA synthesis. Antimicrob Agents Chemother. 2000;44(8):2160–5.
96. Pan X-S, Fisher LM. DNA gyrase and topoisomerase IV are dual targets of clinafloxacin
action in Streptococcus pneumoniae. Antimicrob Agents Chemother. 1998;42:2810–6.
97. Aldred KJ, Breland EJ, McPherson SA, Turnbough CL Jr, Kerns RJ, Osheroff N. Bacillus
anthracis GrlAV96A topoisomerase IV, a quinolone resistance mutation that does not affect
the water-metal ion bridge. Antimicrob Agents Chemother. 2014;58(12):7182–7.
98. Ashley RE, Lindsey RH Jr, McPherson SA, Turnbough CL Jr, Kerns RJ, Osheroff
N. Interactions between quinolones and Bacillus anthracis gyrase and the basis of drug resis-
tance. Biochemistry. 2017;56(32):4191–200.
99. Cullen ME, Wyke AW, Kuroda R, Fisher LM. Cloning and characterization of a DNA gyrase
A gene from Escherichia coli that confers clinical resistance to 4-quinolones. Antimicrob
Agents Chemother. 1989;33(6):886–94.
100. Sreedharan S, Oram M, Jensen B, Peterson LR, Fisher LM. DNA gyrase gyrA mutations
in ciprofloxacin-resistant strains of Staphylococcus aureus: close similarity with quinolone
resistance mutations in Escherichia coli. J Bacteriol. 1990;172(12):7260–2.
101. Hopewell R, Oram M, Briesewitz R, Fisher LM. DNA cloning and organization of the
Staphylococcus aureus gyrA and gyrB genes: close homology among gyrase proteins and
implications for 4-quinolone action and resistance. J Bacteriol. 1990;172(6):3481–4.
102. Oram M, Fisher LM. 4-Quinolone resistance mutations in the DNA gyrase of Escherichia
coli clinical isolates identified by using the polymerase chain reaction. Antimicrob Agents
Chemother. 1991;35(2):387–9.
103. Aldred KJ, Schwanz HA, Li G, Williamson BH, McPherson SA, Turnbough CL Jr, et al.
Activity of quinolone CP-115,955 against bacterial and human type II topoisomerases is
mediated by different interactions. Biochemistry. 2015;54(5):1278–86.
104. Drlica K, Mustaev A, Towle TR, Luan G, Kerns RJ, Berger JM. Bypassing fluoroquinolone
resistance with quinazolinediones: studies of drug-gyrase-DNA complexes having implica-
tions for drug design. ACS Chem Biol. 2014;9(12):2895–904.
105. Elsea SH, McGuirk PR, Gootz TD, Moynihan M, Osheroff N. Drug features that contribute
to the activity of quinolones against mammalian topoisomerase II and cultured cells: cor-
relation between enhancement of enzyme-mediated DNA cleavage in vitro and cytotoxic
potential. Antimicrob Agents Chemother. 1993;37(10):2179–86.
106. Elsea SH, Osheroff N, Nitiss JL. Cytotoxicity of quinolones toward eukaryotic cells.
Identification of topoisomerase II as the primary cellular target for the quinolone
CP-115,953 in yeast. J Biol Chem. 1992;267(19):13150–3.
16 Bacterial Type II Topoisomerases and Target-Mediated... 529

107. Coates WJ, et al. Preparation of piperidinylalkylquinolines as antibacterials. 1999; United

States Patent W09937635.
108. Gomez L, Hack MD, Wu J, Wiener JJ, Venkatesan H, Santillan A Jr, et al. Novel pyrazole
derivatives as potent inhibitors of type II topoisomerases. Part 1: synthesis and preliminary
SAR analysis. Bioorg Med Chem Lett. 2007;17(10):2723–7.
109. Black MT, Stachyra T, Platel D, Girard AM, Claudon M, Bruneau JM, et al. Mechanism of
action of the antibiotic NXL101, a novel nonfluoroquinolone inhibitor of bacterial type II
topoisomerases. Antimicrob Agents Chemother. 2008;52(9):3339–49.
110. Negash K, Andonian C, Felgate C, Chen C, Goljer I, Squillaci B, et al. The metabolism and
disposition of GSK2140944 in healthy human subjects. Xenobiotica. 2016;46(8):683–702.
111. Blanco D, Perez-Herran E, Cacho M, Ballell L, Castro J, Gonzalez Del Rio R, et al.
Mycobacterium tuberculosis gyrase inhibitors as a new class of antitubercular drugs.
Antimicrob Agents Chemother. 2015;59(4):1868–75.
112. Kern G, Palmer T, Ehmann DE, Shapiro AB, Andrews B, Basarab GS, et al. Inhibition
of Neisseria gonorrhoeae Type II Topoisomerases by the Novel Spiropyrimidinetrione
AZD0914. J Biol Chem. 2015;290(34):20984–94.
113. Huband MD, Bradford PA, Otterson LG, Basarab GS, Kutschke AC, Giacobbe RA, et al.
In vitro antibacterial activity of AZD0914, a new spiropyrimidinetrione DNA gyrase/topoi-
somerase inhibitor with potent activity against gram-positive, fastidious gram-negative, and
atypical bacteria. Antimicrob Agents Chemother. 2015;59(1):467–74.
114. Giacobbe RA, Huband MD, de Jonge BL, Bradford PA. Effect of susceptibility testing
conditions on the in vitro antibacterial activity of ETX0914. Diagn Microbiol Infect Dis.
2017;87(2):139–42.
 Part III
Finding New Antimicrobials
Chapter 17
Natural Products in Antibiotic Discovery

Fern R. McSorley, Jarrod W. Johnson, and Gerard D. Wright

17.1 History of Natural-Product Antibiotics

17.1.1 Natural Medicinal Therapies

Nature is rich in bioactive molecules that can be used as medicines. The earliest
records describing natural medicines are found on clay tablets dating from
2600 BCE in Mesopotamia; these records contain over 1000 plant derived-sub-
stances [1, 2]. The most well-known ancient medicinal record is the Egyptian Ebers
Papyrus, which dates from 1500 BCE and contains over 700 natural remedies, most
of plant origin [3]. Natural/herbal treatments are found throughout history and from
all over the world. Examples are the Chinese Materia Medica (Shennong Bencao
Jing) (1100 BCE – 659 CE) [3, 4] and the Indian Ayurvedic system (1000 BCE)
[3, 5]. These collections of ancient remedies were directed at a range of ailments
that included infections; some are still in use today. Such traditional medicines
generally consist of complex extracts and mixtures of agents whose bioactive
component(s) went unidentified for hundreds of years. Improvements came largely
from trial and error efforts that were hindered by confirmational bias and placebo
effects.
In the early 1800s, morphine became the first bioactive natural product isolated
from a medicinal plant [6]. This milestone led the Western medical field away from

Authors Fern R. McSorley and Jarrod W. Johnson contributed equally to this work.
F. R. McSorley · J. W. Johnson · G. D. Wright (*)
Department of Biochemistry and Biomedical Sciences, McMaster University, Michael
G. DeGroote Institute for Infectious Disease Research, McMaster University,
Hamilton, ON, Canada
e-mail: [email protected]

© Springer International Publishing AG, part of Springer Nature 2018 533

I. W. Fong et al. (eds.), Antimicrobial Resistance in the 21st Century, Emerging Infectious
Diseases of the 21st Century, https://doi.org/10.1007/978-3-319-78538-7_17
534 F. R. McSorley et al.

complex mixtures toward the pharmacology of pure compounds [7]. Old methods
based on impure mixtures were unreliable: the variability of growth conditions and
extractions from plant materials and microorganisms impacted the concentrations
of both beneficial and toxic bioactive compounds present in the mixtures. Isolating
the beneficial bioactive compound away from other material found in the natural
source, some of which could cause negative effects, allows in-depth analysis of
mode of action and efficacy. It also enables physicians to more accurately regulate
the dosage of the bioactive compound.
In 1928, just over 100 years after the isolation of morphine, Fleming serendipi-
tously discovered that the fungus Penicillium notatum secretes an agent that pre-
vents the growth of bacteria. A decade and a half later, penicillin became the first
natural-product antimicrobial to reach widespread clinical use [8]. Selman Waksman
coined the term antibiotics to refer to microbially produced compounds that are
“against life”; they either halt the growth of (bacteriostatic) or kill (bactericidal)
microbes. Among these are small peptides having antibiotic activity [9]; they typi-
cally disrupt bacterial membranes in a nonselective manner. Indeed, these peptides
are the first line of defense against bacterial infections. A larger suite of highly
selective antibiotics is produced by microbes. These compounds have been the main
source of our antibiotic drugs thus far [10]. The physiological roles of microbially
produced antibiotics are still debated; these molecules evolved either for signaling
functions or as chemical warfare agents to ward off neighboring microbes [11]. The
density and diversity of resistance elements in microbes are consistent with an
evolved detoxifying role to protect against the growth-impeding effects of antibiot-
ics. Regardless of the evolutionary basis for antibiotics, the introduction of penicil-
lin into the clinic in the early 1940s led to the “antibiotic era” of medicine. Natural
antibiotics have aided in the treatment and control of infections for the last 80 years
(Fig. 17.1). By controlling infections, antibiotics have revolutionized medicine,
allowing physicians to perform lifesaving organ transplants and invasive surgeries
and to treat cancer using disruptive immune system chemotherapy. The emergence
of antibiotic resistance now threatens these breakthroughs, our quality of life, and
our life expectancy.

Golden Age of Era of Antibiotic Era of Antibiotic

Antibiotic Discovery Medicinal Chemistry Resistance

1930 1940 1950 1960 1970 1980 1990 2000 2010

sulfonamides streptogramins oxazolidinones

β-lactams quinolones lipopeptides
chloramphenicol glycopeptides mutilins
tetracyclines macrolides fidaxomicin
aminoglycosides diarlyquinolones

Fig. 17.1 Timeline for the introduction of each class of antibiotic into clinical use. (Adapted from
ref. [12])
17 Natural Products in Antibiotic Discovery 535

17.1.2 The Development of Modern Antibiotics

The first antibacterial compounds to enter the clinic were synthetic molecules dis-
covered in the laboratory of Paul Ehrlich [13, 14]. After noticing that some microbes
stained differently than others when exposed to synthetic aniline and azo dyes,
Ehrlich postulated that a chemical compound could selectively target pathogenic
cells and not host cells. His “magic bullet” theory led him to screen hundreds of
organoarsenic derivatives for activity. One of these, arsphenamine (Salvarsan), was
successfully used to treat syphilis. Salvarsan, although difficult to administer, was
the most prescribed antimicrobial drug until it was replaced by penicillin [15].
Four classes of synthetic antibiotics1 remain successful as drugs in the clinic
today: sulfonamides, diaminopyrimidines, quinolones, and oxazolidinones
(Table 17.1, Fig. 17.2). The initial class of systemic synthetic antimicrobials was the
sulfonamide sulfa drugs [14, 17]. Prontosil, the first sulfa drug for human use, is a
prodrug; after administration, it is metabolized into the bioactive agent sulfanil-
amide. Sulfonamides work by inhibiting dihydropteroate synthase, a critical com-
ponent in folate synthesis. Humans acquire folate from their diet, while bacteria
must biosynthesize this essential nutrient. Consequently, sulfa drugs are selectively
active against microbes. Sulfonamides are generally co-administered with the
diaminopyrimidine trimethoprim. This synthetic antibiotic also targets folate bio-
synthesis, inhibiting dihydrofolate reductase. The synergistic combination of trim-
ethoprim and sulfamethoxazole (co-trimoxazole) is sold under a variety of trade
names (e.g., Septra, Bactrim).
The quinolone drugs target type II DNA topoisomerases and replication. These
agents are potent against both Gram-negative and Gram-positive bacteria. A widely
used example is ciprofloxacin (Fig. 17.2), a second-generation fluoroquinolone
that is orally available and used to treat urinary tract infections, sinusitis, and many
other infections. Several newer generations of quinolones have found a place in the
market, making the fluoroquinolones the most successful class of synthetic drugs
to date.
The oxazolidinones comprise the fourth class of synthetic antibiotic. These com-
pounds, which block the ribosomal peptidyl transferase center essential for protein
synthesis, are effective primarily against Gram-positive bacteria. Linezolid
(Fig. 17.2) represents the first generation of this class; it was approved by the US
Food and Drug Administration in 2000 and became the first novel chemical scaffold
to enter the clinic in several decades.
Fleming’s discovery that P. notatum secretes a bactericidal substance helped
launch the intensive mining of microbes as sources of antibiotics rather than exam-
ining synthetic chemical libraries [8]. Florey, Chain, and Heatley’s efforts to develop
a protocol for isolation of penicillin, followed by in vivo efficacy studies in animals
and clinical tests in humans, showed that penicillin, a natural product, was a viable

1
Here we are deviating from Waksman’s definition of antibiotics restricted narrowly to compounds
synthesized by microbes to include synthetic and semisynthetic human-made compounds as well.
Table 17.1 Timeline of discovery and introduction of antibiotic classes
Resistance
Antibiotic class; examples Discovery Introduction observed Mechanism of action Activity; target species
Organoarsenics; salvarsan 1909 1910 1924 Unknown Bactericidal; antisyphilitic
Sulfadrugsa; prontosil, sulfanilamide 1932 1936 1942 Inhibits dihydropteroate synthetase in folate synthesis Bacteriostatic; Gram-positive bacteria
β-Lactams; penicillins, 1928 1938 1945 Inhibits penicillin-­binding proteins in cell-wall Bactericidal; broad-spectrum
cephalosporins, carbapenems biosynthesis
Aminoglycosides; streptomycin, 1943 1946 1946 Binds 30S ribosomal subunit inhibiting protein synthesis Bactericidal; broad-spectrum
spectinomycin, kanamycin,
neomycin
Chloramphenicols; 1946 1948 1950 Binds 50S ribosomal subunit inhibiting protein synthesis Bacteriostatic; broad-spectrum
chloramphenicol
Macrolides; erythromycin, 1948 1951 1955 Binds peptidyl transferase center of 50S ribosomal Bacteriostatic; broad-spectrum
clarithromycin subunit inhibiting protein synthesis
Tetracyclines; chlortetracycline, 1944 1952 1950 Binds 30S ribosomal subunit inhibiting protein synthesis Bacteriostatic; broad-spectrum
doxycycline
Ansamycins; rifamycins rifampicin 1957 1958 1962 Binds RNA polymerase β-subunit inhibiting RNA synthesis Bactericidal; Gram-positive bacteria
Glycopeptides; vancomycin, 1953 1958 1960 Transpeptidase blockade inhibiting cell-wall biosynthesis Bactericidal; Gram-positive bacteria
teicoplanin
Quinolonesa; ciprofloxacin 1961 1968 1968 Binds DNA gyrase inhibiting DNA synthesis Bactericidal; broad-spectrum
Streptogramins; pristinamycins 1963 1998 1964 Binds 50S ribosomal subunit inhibiting protein synthesis Bactericidal, Gram-positive bacteria
Oxazolidinones; linezolid 1955 2000 2001 Binds peptidyl transferase center of 50S ribosomal Bacteriostatic; Gram-positive
subunit inhibiting protein synthesis bacteria
Lipopeptides; daptomycin 1986 2003 1987 Depolarization of cell membrane Bactericidal; Gram-positive bacteria
Mutulins; pleuromutilin, 1951 2007 1999 Binds peptidyl transferase center of 50S ribosomal Bacteriostatic; Gram-positive
retapamulin subunit inhibiting protein synthesis bacteria
Fidaxomicin (targeting Clostridium 1975 2011 1977 Inhibition of RNA polymerase Bactericidal; Gram-positive bacteria
difficile)
Diarylquinolinesa; bedaquiline 1997 2012 2006 Inhibition of F1FO-ATPase Narrow-­spectrum activity,
Mycobacterium tuberculosis
a
Synthetic antibiotic classes. Table assembled from references [11, 12, 16]
17 Natural Products in Antibiotic Discovery 537

O
O O NH2
S F CO2H O
OMe O
NH2 NH2 N N
N N N N O
N H 2N N OMe
HN F
OMe NHAc
H 2N
prontosil trimethoprim ciprofloxacin linezolid
(sulfonamide) (diaminopyrimidine) (fluoroquinolone) (oxazolidinone)

Fig. 17.2 Synthetic antibiotics used in the clinic

antibiotic drug candidate. The discovery of penicillin gave rise to the “golden age”
of antibiotic discovery (Fig. 17.1). For the next 20 years, extensive screening of
microbes, particularly soil-dwelling actinomycetes, was conducted to identify natu-
ral antibiotic compounds. In that time many natural-product scaffolds (aminoglyco-
sides, chloramphenicol, macrolides, tetracyclines, ansamycins, glycopeptides, and
streptogramins) were identified and often rapidly developed for clinical use
(Table 17.1). By the middle of the 1960s, this simple but effective screening plat-
form [18] appeared to have exhausted its sources, as new antibiotic scaffolds suit-
able for drug development became harder and harder to find.
The decline of success in isolating new chemical scaffolds from natural sources
suitable for drug development, the generally poor pharmacological and toxicologi-
cal properties of natural products as drugs, and the problematic emergence of
­resistance ushered in the next phase of antibiotic innovation. The focus shifted to
medicinal chemistry efforts as we entered the “medicinal chemistry era.” Synthetic
chemists prepared and derivatized core antibiotic scaffolds already used in the clinic
and screened them for improvements. Chemists were successful in creating so-­
called “generations” of enhanced synthetic variants that improved pharmacological
properties, expanded antibiotic spectra, and evaded resistance. Although many
improved drugs and new generations of known drugs emerged from these efforts, no
truly novel chemical scaffold entered the clinic from the 1960s to 2000s.
The lack of innovation in antibiotic discovery over the past two decades and the
general failure of in vitro target-based drug discovery methods have prompted a
renewed interest in natural products [16, 19]. This return to the natural-product
compounds that previously dominated antibiotic drug discovery reflects the historic
success of these drugs, a growing understanding of the physiochemical properties of
small molecules for efficacy against bacterial targets, and new thinking resulting
from advances in bacterial genomics, synthetic biology, and the properties of anti-
microbial targets.

17.2  ajor Classes of Natural-Product Antibiotics and Their

M
Modes of Action

Microbial natural products are the source of most of our antibiotic scaffolds in cur-
rent clinical use. Brief descriptions of the most prominent, clinically used drugs and
their modes of action are outlined below.
538 F. R. McSorley et al.

β-Lactams
Penicillin falls into the β-lactam category of natural products. Five classes of
β-lactams are important in the antibiotic field: penams, cephems, carbapenems, cla-
vams, and monobactams (Fig. 17.3). All β-lactams contain a strained 4-membered
β-lactam ring system that in the majority of clinically relevant compounds is fused
to a 5- or 6-membered ring system.
Penams, cephems, and carbapenems covalently bind to penicillin-binding proteins
(PBPs), essential enzymes that process the D-Ala-D-Ala termini of the pentapeptide
portion of the peptidoglycan intermediates in cell-wall metabolism. The electro-
philic β-lactam antibiotics mimic the D-Ala-D-Ala substrate (Fig. 17.4) and cova-
lently bind to the PBP [20], preventing the enzyme from facilitating transpeptidation
in the final step of peptidoglycan synthesis. This results in a complex series of
molecular events that include inhibition of cell division and eventually cell
rupture.

HO R
A H H R' H H H OMe H H H R' R' H
RHN S R S RHN S RHN S RHN R'' R O
R'
X X R''
N R'' N N N N N N
O O O X O X O O X O
CO2H CO2H CO2H CO2H CO2H CO2H
penam penem cephem cephamycin carbapenem monobactam clavam

H H H penicillin H H
B N S deacylase H 2N S acylation
6 5 2 semisynthetic
7
O N N 3
penicillins
O O
CO2H CO2H
penicillin G 6-APA
Me
O NH2 OMe
N H H H H H H O O Ph
N N H H H H H H
S S N O N
S N N S
O N O N H
O R O MeO O N N O N
O Et O
CO2H CO2H
CO2H CO2H
oxacillin ampicillin: R = H methicillin piperacillin
amoxicillin: R = OH

C
H H H enzymatic H H
HO2C N S H 2N S acylation
or 7 6 2 semisynthetic
8 3
NH2 O N OAc chemical N 4 3' OAc cephalosporins
O O
cleavage
CO2H CO2H
cephalosporin C 7-ACA
CO2H Me
N
O OMe OEt
N N N
H H H H H H H H H H H H
S N S N N S N N S N N S
H2N H2N H 2N N
O N OAc S O N N S O N N S N O N
S S
O O O Me O
CO2H CO2H CO2H CO2H
cefalothin ceftazidime cefepime ceftaroline

CO2H
D E
HO HO HO Me O
H H H H H H O
H N
NH2 N H NMe2 H H Me
5 1 N
S S S N H
N N N NH H 2N
O O NH O
S O N
CO2H CO2H CO2H O SO3H
thienamycin imipenem meropenem aztreonam

Fig. 17.3 β-Lactam antibiotics. (A) Major structural classes of β-lactam antibiotics. (B) The deac-
ylation of penicillin G to generate 6-APA opened the door for the synthesis of new semisynthetic
penicillins having improved antibiotic potency and spectrum of activity. (C) Similarly, 7-ACA has
been used as a key intermediate for the preparation of countless semisynthetic cephalosporins hav-
ing potent, broad-spectrum activity against clinically important Gram-negative bacteria. Ceftaroline
has activity against methicillin-resistant Staphylococcus aureus (MRSA). (D) Natural and syn-
thetic carbapenem antibiotics. (E) Aztreonam, a representative monobactam
17 Natural Products in Antibiotic Discovery 539

A Me
H Me Me H
CO2 CO2
O O
N H N
H S
H
penicillin N N D-Ala-D-Ala
H H
R H R Me
O O

Mechanism of action: Resistance:

B R H R H
N HH penicillin-binding H H H Serine- or metallo- N HH
S protein (PBP) R N S b-lactamase S
O O
O N Ser OH O N hydrolytic HO N
O H O O H
CO2 CO2 deactivation CO2
Ser
acylated enzyme penicillin penicilloic acid
blocks the PBP active site;
prevents cell-wall cross-linking

C O
H HO O HO O H
O OH S S H 2N S N
N
N N N
N N N O B
O N CO2H
O O HO O
CO2H CO2H CO2H O OSO3Na
clavulanic acid sulbactam tazobactam avibactam vaborbactam

Fig. 17.4 β-Lactams mimic the D-Ala-D-Ala terminus of the peptidoglycan peptide strands and
block the cross-linking step in cell-wall biosynthesis. (A) Structural analogy of penicillin with the
D-Ala-D-Ala portion of peptidoglycan, as proposed by Tipper and Strominger in 1965 [20]. (B)
Interactions of a penicillin with penicillin-binding proteins (PBPs) and β-lactamases. The acyla-
tion of PBP active-site serine prevents cross-linking of the bacterial cell wall and leads to cell lysis.
In contrast, β-Lactamases hydrolyze β-lactams rapidly and confer high-level resistance. (C)
β-Lactamase inhibitors that are used in combination with β-lactams to overcome resistance due to
β-lactamases

Thousands of penicillins and cephalosporins have been synthesized through the

acylation of 6-aminopenicillanic acid (6-APA) and 7-aminocephalosporanic acid
(7-ACA) in order to improve antibiotic potency, the antibiotic spectrum against both
Gram-positive and Gram-negative bacteria, and their stability against β-lactamases
(Fig. 17.3b) [21]. β-Lactamases are resistance enzymes that rapidly cleave the
β-lactam ring and render the antibiotic inactive. Although clavulanic acid and penam
sulfones were found to be poor antibiotics, they are potent inhibitors of β-lactamases.
Thus, they are combined with β-lactam antibiotics to successfully overcome resis-
tance due to β-lactamases. Augmentin (amoxicillin + clavulanic acid), Unasyn
(ampicillin + sulbactam), and Zosyn (piperacillin + tazobactam) are among the
most successful combinations [22]. Two other β-lactamase inhibitors, avibactam
and vaborbactam (RPX 7009), have been developed and approved recently for use
in combinations with β-lactams [23, 24].
The discovery of several monobactams produced by soil-derived Gram-negative
bacteria inspired the development of antibiotics based on a more simplified
4-­membered lactam ring [25]. Before this discovery chemists had not considered
using an N-sulfonic acid substituent to stabilize the β-lactam system. This modifica-
tion eventually led to the successful antibiotic drug aztreonam (Fig. 17.3e) [26].
This example illustrates the importance of natural-product discovery in guiding the
synthesis of new scaffolds.
540 F. R. McSorley et al.

Aminoglycosides
In 1944, Waksman’s laboratory discovered streptomycin, the first aminoglycoside
antibiotic, as a product of a strain of Streptomyces griseus [27]. Over the next two
decades, several members of the class were discovered that included kanamycin in
1956 from Streptomyces kanamyceticus and gentamicin C in 1963 from
Micromonospora purpurea [28, 29]. These were the first antibiotics to show effi-
cacy in the treatment of tuberculosis and in infections caused by Gram-negative
bacteria. Aminoglycosides consist of a core aminocyclitol ring modified to varying
extents by saccharides. Multiple amino groups provide a positive charge at physio-
logical pH and impart high water solubility. Currently, three aminoglycosides are
commonly used in the clinic: the natural products, gentamicin C and tobramycin,
and the semisynthetic agent amikacin, each of which contains the 2-­deoxystreptamine
aminocyclitol core (Fig. 17.5a) [30]. Streptomycin, which is unique in this drug
class for having a streptamine core aminocyclitol ring, continues to find some clini-
cal use in the treatment of tuberculosis and in combination with penicillins for
enterococcal infections that are difficult to treat.
The amino and hydroxy groups of the aminoglycosides interact with the 16S
rRNA in the 30S ribosomal unit through a network of hydrogen bonds (Fig. 17.5b).
The binding of aminoglycosides to the 16S rRNA results in a conformational change
that impedes cognate codon–anticodon validation by the ribosome. The result is the
corruption of the genetic code and the synthesis of proteins with incorrect amino
acids. This corruption contributes to cell death.
Resistance to aminoglycosides can take many forms. Active efflux can be a sig-
nificant contributor to resistance in bacteria such as P. aeruginosa; however, the
main mechanisms of resistance are chemical modification of the drugs or the target
[30]. A large number of aminoglycoside N-acetyltransferases, O-phosphotransferases,
and O-adenylyltransferases are present in both Gram-positive and Gram-negative
pathogens (Fig. 17.5c). Over the past decade, ribosomal methyltransferases that
modify the 16S rRNA (e.g., G-1405 and A-1408) and confer high-level aminogly-
coside resistance in Gram-negative pathogens have also emerged as significant
clinical challenges (Fig. 17.5b).

Macrolides
The term macrolide is a portmanteau that combines macrolactone, a lactone ring
containing eight or more atoms, and polyketide. Erythromycin is a first-generation
macrolide that was isolated from a soil actinomycete Saccharopolyspora erythraea.
Erythromycin contains a 14-membered macrolactone framework and a 2-amino
sugar (Fig. 17.6). The presence of a ketone at position 9 of the macrolactone can
result in formation of a hemiketal with the hydroxyl group at position 6 under acidic
conditions (e.g., exposure to gastric acids), thereby decreasing the levels of bio-
available drug [31]. Semisynthetic conversion of erythromycin to clarithromycin or
azithromycin removes this possibility and results in improved efficacy and bioavail-
ability. Macrolide antibiotics are most effective against Gram-positive bacteria, but
they also have efficacy against some common Gram-negative, upper respiratory
tract pathogens.
 17

HN
A NH2 OH OH OH
HN OH R
HO O R OH HO O H2N R Me O H2N
Me O O OH O O O
HO NH H2N H2N MeHN
HO OH NH2 HO OH NH2 HO OH NH2
OHC OH NH 2 O O O O
O O
HN H2N NH2 H2N NH2 H2N NH2
HO O
HO
O streptomycin kanamycin A: R = OH tobramycin: R = OH gentamicin C2: R = Me
HO NHMe
HO kanamycin B: R = NH2 dibekacin: R = H gentamicin B: R = H

NH2 OH OH
NH2 OH
HO O O O
HO HO OH Me H2N
HO O H2N O MeHN H
H2N NH2 OH OH O N
O HO NH2 HO
OH O O O
O O NH HO O NH2 HO HN NH2 OH
HN
HO OH

H2N O H2N O
HO OH butirosin amikacin plazomicin
Natural Products in Antibiotic Discovery

C-1496 U-1495 G-1494

B NH2
C AAC(6')
A-1493 AAC(4',4")
HO O 6' NH2
HO 4'
AAC(3)
H2N NH2 G-1405 * HO O AAC(1)
O A-1492 HO
NH2 3'
HO O O H2N 4 NH2 2
OH O 1
* APH(3') HO NH2
NH2 5
HO O
O OH AAC(2') OH
U-1406 1''
H2N O A-
G-1491 O
1408 HO 2'' 3'' OH
OH C-1409 APH(2")
H2N
neomycin B neomycin B ANT(2")
bound to E. coli AAC(3")
ribosome (PDB: 4V9C) U-1490 kanamycin B

Fig. 17.5 Aminoglycosides. (A) Structures of natural and semisynthetic aminoglycosides. (B) Neomycin B and its binding mode in the H44 region of the E.
coli ribosome (PDB: 4V9C). Asterisks at G-1405 and A-1408 indicate sites of methylation by methyltransferases that confer aminoglycoside resistance (e.g.,
ArmA and NpmA, respectively). (C) Major target sites of aminoglycoside-modifying enzymes (AMEs), including O-phosphotransferases (APHs),
541

N-acetyltransferases (AACs), and O-nucleotidyltransferases (ANTs)

 542

O Me
A Me Me Me N Me
N N
O
9 N Me
1 Me
HO OR HO OMe O
N
Me OH 6 Me NMe2 Me OH Me NMe2 OMe
Me HO Me HO O Me
Et O O O Me Et O O O Me Me NMe2
Me HO
NH2 O O O Me
O O OR2 O O OMe
Me
Me Me Me Me O O
O OH O OH F Me
Me Me
erythromycin A: R1 = H, R2 = H azithromycin solithromycin
erythromycin C: R1 = H, R2 = Me
clarithromycin: R1 = Me, R2 = Me

R HO NMe2 R OH NMe2 NMe2 NMe2

B H H H H H H
OH OH OH
O
H
NH2 NH2 t-BuN NH2
N
OH OH H OH
OH O HO O O OH O HO O O OH O HO O O

tetracycline: R = H oxytetracycline: R = OH tigecycline

chlortetracycline: R = Cl doxycycline: R = H

C
Me Me Me Me Me Me Me Me Me Me Me Me
AcO AcO AcO AcO

MeO OH OH O MeO OH OH O MeO OH OH O MeO OH OH O

Me Me Me Me Me
OH O Me OH OH Me OH OH Me OH OH
Me NH Me NH Me NH Me NH

O O O O O O N O O N
O O HO N N
O O R O O
Me Me Me NMe Me
Me
rifamycin S rifamycin SV: R = H rifampin rifaximin
rifamycin B: R = CH2CO2H

Fig. 17.6 Selected examples of macrolides (A), tetracyclines (B), and ansamycin antibiotics (C)
F. R. McSorley et al.
17 Natural Products in Antibiotic Discovery 543

Macrolides interact with the peptidyl transferase center of the 50S ribosomal
subunit, blocking peptide-chain elongation. Specifically, the hydrophobic surface of
the macrolide binds to the sidewall of the exit tunnel and causes premature release
of short peptidyl-tRNAs [31].
The main causes of clinical macrolide resistance are the 23S rRNA methyltrans-
ferases (Erm) [32]. However, a number of macrolide kinases, which modify the
desosamine sugar, are increasing in prevalence and diversity. Ring-opening ester-
ases and desosamine glycosyltransferases are also known, but they are less fre-
quently encountered. Active macrolide efflux is common in many Gram-positive
pathogens.

Tetracyclines
Chlortetracycline (Aureomycin) and oxytetracycline (Terramycin), discovered in
1948 and 1950, respectively, were the first members of the tetracycline class
(Fig. 17.6b). Tetracyclines contain a tetracyclic polyketide core of four fused
6-membered rings; they were the first broad-spectrum antibiotics (i.e., effective
against both Gram-positive and Gram-negative pathogens) to enter clinical use.
After it was found that the C6-hydroxy group could be reductively removed to form
a more stable 6-deoxytetracline, further modification ensued. Multiple generations
of tetracyclines have now emerged that include doxycycline, minocycline, and tige-
cycline [33]. Tigecycline (approved for use in the USA in 2005) has broad-­spectrum
activity for both Gram-positive and Gram-negative bacteria; it is also effective
against methicillin-resistant Staphylococcus aureus (MRSA).
Before the emergence of resistance limited their use, the tetracyclines were used
for decades to treat infections of the respiratory tract, middle ear, and urinary tract.
Like aminoglycosides, tetracyclines bind to the 30S ribosomal subunit. However,
they do not cause the production of aberrant proteins. Instead they bind to the
aminoacyl-­tRNA binding site of the ribosome, thereby competitively preventing
translation of mRNA.
Tetracycline resistance in Gram-negative bacteria is most often the result of
active efflux. In Gram-positive bacteria, the expression of ribosomal protection pro-
teins lowers the affinity of tetracycline for the bacterial ribosome. Enzyme-mediated
inactivation has been reported through the action of TetX, a flavin-dependent mono-
oxygenase that hydroxylates the antibiotic, thereby precipitating a nonenzymatic
breakdown of the compounds [34].

Ansamycins
As with the macrolides, ansamycins are polyketide macrocycles; however, ansa-
mycins cyclize to form a macrolactam instead of a macrolactone (Fig. 17.6c). The
ansa-bridged macrolactam is formed with an intramolecular amine nucleophile
derived from the biosynthetic starter unit, 3-amino-5-hydroxybenzoyl-CoA. The
natural ansamycin, rifamycin, is an 18-membered macrolactam that is converted
through semisynthesis to the commonly used rifampin. Addition of the pipera-
zinyl hydrazide to the rifamycin naphthyl core in rifampin increases its oral bio-
availability and broadens antimicrobial activity. Rifampin is a WHO essential
544 F. R. McSorley et al.

medicine designated for use in combination therapy for the treatment of

tuberculosis.
Rifamycins are inhibitors of transcription [35]. In particular they bind to the
mRNA exit site of the β-subunit of RNA polymerase. Resistance mutations in the
rifamycin-binding site are common. As a result, rifamycins are most effective when
in combination with other antimicrobials [35]. In the treatment of M. tuberculosis,
which is the main use of the rifamycins, resistance is most often the result of point
mutations in the target RNA polymerase [36]. In contrast, many environmental
microbes express a wide variety of rifamycin-inactivating enzymes (e.g., kinases,
ADP-ribosyltransferases, monooxygenases, and glycosyltransferases).

Glycopeptides
As their name implies, glycopeptides are peptides that are decorated with sugar
moieties. Vancomycin and teicoplanin (Fig. 17.7) exemplify this class; both have
been developed as Gram-positive-directed antibiotics. Vancomycin, which was dis-
covered in the early 1950s as a product of the actinomycete Amycolatopsis orienta-
lis, was used only sporadically for several decades, largely due to difficulties in
obtaining pure compound. However, use increased in the 1980s following the wide-
spread emergence of MRSA in hospitals [38]. Emergence of resistance in entero-
cocci (VRE) and then intermediate resistance in S. aureus (VISA) spurred the
development of second-generation glycopeptides such as telavancin, dalbavancin,
and oritavancin that are less susceptible to resistance [39].
Vancomycin and teicoplanin are highly cross-linked pentapeptides that have a
high affinity for D-Ala-D-Ala termini of uncrosslinked peptidoglycan chains.
Vancomycin forms five hydrogen bonds with the D-Ala-D-Ala terminus of lipid II
and prevents the formation of interpeptidyl cross-links by PBPs. That reduces the
integrity of the cell wall and leads to cell death. Although glycopeptides and
β-lactams both inhibit cell-wall biosynthesis, the glycopeptides sequester the sub-
strate of transpeptidation rather than directly interacting with the PBP catalyst.
Resistance to glycopeptide antibiotics can take two forms. In Enterococci, repro-
gramming of cell-wall biosynthesis to terminate in either D-Ala-D-Lac or D-Ala-D-­
Ser reduces affinity of the antibiotic. In Staphylococcus aureus, acquisition of the

HO HO OH Cl
NH2
H
Me O N OH
Me OH
O NH
OH HO O O
vancomycin O O OH HO
OH HO O Me O
Cl Me OH
O O O O O
HO NH2 O OH
O O
HO Cl OH AcHN Cl Me OH
O O O Cl
O H H O O O O O
6 5 3 1 H H Me
N N NHMe N N O
O N 4 N 2 N O N N O Cl OH
H H N
H H H H O O O
NH O O O O H H
O HN NH2 N N NHMe
7 O N N
HO2C N
NH2 HO2C O H H H
OH O O HO HN O O
Me O O
OH HO
HO X HO HO HO2C NH2
N O O OH
H HO
peptide stem O Me OH OH
HO HO
OH teicoplanin A2-2 oritavancin
oflipidII -D-Ala-D-Ala: X = NH
-D-Ala-D-Lac: X = O

Fig. 17.7 Glycopeptide antibiotics and the interaction between vancomycin and the D-Ala-D-Ala
portion of lipid II. In vancomycin-resistant enterococci (VRE), the peptide stem contains a D-Ala-­
D-Lac terminal region, and the affinity for the glycopeptide is decreased 1000-fold [37]
17 Natural Products in Antibiotic Discovery 545

genes encoding D-Ala-D-Lactate biosynthesis is known but exceedingly rare.

Instead, in this organism glycopeptide resistance is primarily the result of increased
production of cell-wall polymers that bind the antibiotic.

Streptogramins
Dalfopristin and quinupristin (Fig. 17.8a) are semisynthetic derivatives of virginia-
mycins/pristinamycins that belong to the A-type and B-type streptogramin families.
Type B streptogramins are cyclic hexapeptides, while type A streptogramins are
cyclic polyketide–peptide hybrids. This illustrates that a pair of molecules, which
have only bacteriostatic activity and are not effective treatments alone, can be com-
bined to work synergistically and form a potent bactericidal drug. The molecular
mechanism of synergy is based on the affinity of the compounds for different but
adjacent regions of the bacterial ribosome [40]. Dalfopristin binds to the peptidyl
transferase center where it reduces the affinity of aminoacyl-tRNAs for the amino-
acyl site, which lowers subsequent peptide bond formation and chain elongation in
the peptidyl site. In contrast, quinupristin binds in a similar manner to erythromycin
at the proximal end of the tunnel, thereby accelerating the release of small
oligopeptidyl-­tRNAs [41]. When administered together, the two agents form a com-
bination drug known as Synercid. It is used to treat staphylococcal infections [42].
Both type A and B streptogramins are susceptible to efflux-mediated resistance;
indeed, the efflux protein Lsa intrinsic to Enterococcus faecalis confers resistance
to type B streptogramins, limiting Synercid use [40]. A group of O-acetyltransferases
confer high-level resistance to type A streptogramins, while Vgb is a ring-opening
C–O lyase that provides resistance to type B antibiotics.

Lipopeptides
As their name suggests, these compounds are peptides that contain a lipid moiety.
Both linear and cyclic, macrolactone and macrolactam lipopeptides exist. Due to the
large variations in structure, these molecules have few well-characterized cellular
targets. Daptomycin (Fig. 17.8b), initially discovered in the 1980s and discarded at
Phase II clinical trials by Eli Lilly due to toxicity, was revisited with a new dosing
regimen; it was approved for clinical use in 2003 [43]. Daptomycin has pleiotropic
effects on the membrane of Gram-positive bacteria that result in depolarization and
physical alteration of the cell membrane that leads to cell death [44]. Daptomycin is
effective against most Gram-positive pathogens, including drug-resistant forms
such as VRE and MRSA.
Colistin (polymyxin E) is a lipopeptide of the polymyxin class. It is one of the
few antibiotics in clinical use that was derived from a non-actinomycete bacterium,
Paenibacillus polymyxa. Discovered in 1949, colistin has been used sparingly for
the treatment of serious infections caused by Gram-negative bacteria due to toxic-
ity issues [45]. As a result of the rise of carbapenem-resistant Gram-negative
pathogens, clinicians have been left with few therapeutic options other than colis-
tin. Consequently, its use has increased significantly. The mode of action of poly-
myxins involves disruption of the outer and inner membranes of Gram-negative
bacteria [46].
 546

O O
A N
O O S
H H
OH N OH N
N N N N
H OH H OH
Me O O Me O O
O O O O O O O O O O O O
N O N N O N
O O Me R O O Me NMe2
N N HN H N N HN H
O N O O N
N O S N
O O

virginiamycin M1 virginiamycin S1: R = H Et2N dalfopristin quinupristin

(= pristinamycin IIA) pristinamycin IA: R = NMe 2

H2N H2N
B C NH2
NH
O O
H Me HN
N NH
NH N N HO O O
H H O
CONH2 O NH O HN O O
O O HO2C HN
H OH O NH
N Me HO2C O O O O
N N H H H
H H N N N
NH O O O HO NH N N N
CO2H O H H H
H O O O
O N
NH2 N N O
H H NH2 NH2 NH2
O
O Me
colistin
daptomycin CO2H (= polymyxin E)

Fig. 17.8 Streptogramins (A), daptomycin (B), and colistin (C)

F. R. McSorley et al.
17 Natural Products in Antibiotic Discovery 547

Resistance to colistin occurs through chemical modification of the lipopolysac-

charide component of the Gram-negative outer membrane. Expression of intrinsic
aminoarabinose transferases and phosphoethanolamine transferases, which modify
lipid A impact the physical properties of the outer membrane, confers colistin resis-
tance [47]. Mobilization of the mcr-1 phosphoethanolamine transferase gene in
Gram-negative pathogens is a growing concern [48]. In contrast, daptomycin resis-
tance remains rare and has not been mobilized. However, resistance mutations in
cell membrane and cell-wall structure and function can be selected in vitro and
during long-term therapy [49].
Pleuromutilins
In 2007 the first pleuromutilin compound approved for clinical use was the topical
antibiotic retapamulin [10]. Although retapamulin represents a new chemical scaf-
fold for clinical use, it is actually a semisynthetic version of the original pleuromu-
tilin, which was isolated from the fungus Pleurotus mutilis (renamed Clitopilus
scyphoides in 1951) [50]. Pleuromutilins contain a fused 8-6-5 tricyclic diterpene
architecture and an acyclic tail. It is among the few isoprenoid antibiotics to find
clinical use (Fig. 17.9). These antibiotics bind to the peptidyl transfer center of the
50S ribosomal subunit, thereby blocking protein synthesis. 23S rRNA methyltrans-
ferases can confer resistance to this class of antibiotic. Retapamulin is a topical
agent used for treatment of skin infections caused by Gram-positive bacteria; sev-
eral other pleuromutilin derivatives are currently in various stages of clinical assess-
ment for systemic use.
Although the isoprenoid class contains the most abundant natural products (over
50,000 known structures), very few are known to exhibit specific antibiotic activity.
Other examples include platensimycin, platencin, and fusidic acid.
Chloramphenicol
Chloramphenicol, discovered in 1947 in extracts of Streptomyces venezuelae, dis-
plays broad-spectrum bacteriostatic activity [51]. This small molecule contains a
dichloroacetamide moiety and an aromatic nitro group (Fig. 17.9). The dichloro-
acetyl moiety is important for activity, as it impedes tRNA from binding to the
peptidyl transferase in the 50S ribosomal subunit, thereby preventing elongation.
Due to the low-cost production of chloramphenicol, this agent is frequently used
in developing countries, even though it has been withdrawn from common use in
many areas due to resistance and safety concerns, the latter resulting from

Cl Et
O
Me OMe
HO O O
HO O
Cl OH
Me Me
OH OH O OH HO O
S O OH
MeN H H
OH RO O O
Me Me HO Me Me Me Me Me
HN CHCl2 Me Me P
O2N Me Me HO O Me
O O HO
O O Me O O
chloramphenicol mutilin: R = H retapamulin fosfomycin O fidaxomicin
pleuromutilin: R = COCH2OH HO Me

Fig. 17.9 Chloramphenicol, pleuromutilins, fosfomycin, and fidaxomicin

548 F. R. McSorley et al.

low-frequency association with aplastic anemia. Resistance is primarily the result

of O-acetyltransferases that modify the antibiotic.

Fosfomycin
The phosphonate fosfomycin is a small (138 Da) antibiotic produced by Streptomyces
fradiae. It has been in clinical use since the early 1970s for the treatment of urinary
tract infections. The rapid emergence of resistance has limited its use, but its high
water solubility and low toxicity enables a single dose (3 g of drug/day) or very
short course therapy. Fosfomycin targets bacterial cell-wall biosynthesis by inhibi-
tion of MurA, an enzyme involved in the first step in the biosynthesis of
N-acetylmuramic acid [52].
Fosfomycin resistance is readily selected during therapy in the form of mutants
in the glycerol-3-phosphate transporter, which is needed for fosfomycin entry into
the cell but not essential for bacterial cell growth [52]. A series of enzymes, includ-
ing glutathione transferases and epoxide hydrolases, are known to inactivate the
antibiotic via opening the essential epoxide ring.

Fidaxomicin
Fidaxomicin is the first member of the newest class of natural products to enter the
market (2011). It consists of an 18-membered macrolactone polyketide that was
discovered independently in Italy (lipiarrmycin), Japan (clostomicin), and the USA
(tiacumicin) in the early 1970s [53]. The macrolactone is decorated with two acyl-
ated rhamnoses (Fig. 17.9). Fidaxomicin inhibits RNA polymerase by binding to a
site distinct from the rifamycin binding site. There, fidaxomicin blocks the conver-
sion of bound promoter DNA to the open single-strand complex that forms the
transcription bubble. It has been approved for clinical use for treatment of
Clostridium difficile-associated diarrhea.

17.3 Natural Products: A Privileged Source of Antibiotics

The vast majority of antibacterials used clinically are natural products, semisyn-
thetic derivatives, or analogues thereof. As mentioned above, the only synthetic
classes that are not derived from natural scaffolds are the sulfonamides, diaminopy-
rimidines, oxazolidinones, and quinolones. Natural products have always been a
major source of human medicines and continue to be especially important as leads
for antimicrobials – 74 of the 98 small molecules approved as antibacterials from
1981 to 2006 are natural products, semisynthetic derivatives, or natural-product
mimics [10].
In the 1980s the arrival of combinatorial chemistry allowed the rapid synthesis of
large numbers of synthetic compounds. That transformed the pharmaceutical indus-
try, and companies began to favor screening vast libraries of synthetic compounds
over natural-product extracts. With major advances in high-throughput screening
(HTS) technologies through the 1990s and 2000s, companies obtained the ability to
17 Natural Products in Antibiotic Discovery 549

quickly screen libraries of millions of compounds. This strategy has been enor-
mously successful in identifying lead compounds having targets in human cells, but
it has not been successful for identifying new antibacterials. Extensive high-­
throughput screening campaigns at GlaxoSmithKline (GSK) [54] and AstraZeneca
[55] both failed to identify any candidate structure worthy of further clinical devel-
opment. The lack of success is due to a combination of several factors, with retro-
spective analyses of both campaigns pointing to a lack of chemical diversity in the
compound libraries.
The chemical libraries of pharmaceutical companies have largely been con-
structed with guidance from Lipinski’s Rules [56], which aim to improve the likeli-
hood of oral bioavailability by keeping molecular weights (MW) under 500,
measures of hydrophobicity (logP or logD) less than 5, and the number of hydrogen-­
bond donors and acceptors in the molecule less than 10. However, antibiotics have
long been known to occupy a different “chemical space” than other drugs, and often
they exhibit multiple violations of Lipinski’s guidelines. An analysis of physio-
chemical properties by O’Shea and Moser in 2008 [57] compared a reference set of
human drugs against compounds active against Gram-positive and Gram-negative
bacteria. Average molecular weights were 338, 813, and 414, and average clogD7.4
values were 1.6, −0.2, and −2.8, respectively. Anti-Gram-positive compounds are
more polar than reference drugs and can be much larger, especially if their targets
are on the cell exterior (e.g., glycopeptides, lipopeptides). Compounds active against
Gram-negative bacteria, which must cross the outer membrane, are much more
polar and have a strict molecular weight cutoff at 600 Da, likely due to the limita-
tions of transport through porins. Overington pointed out, however, that the ­bacterial
target should be taken into account, since the physiochemical properties of antibiot-
ics targeting the ribosome fall further outside Lipinski’s rules than antibacterials
that have bacterial protein targets [58].
An analysis of 23 HTS campaigns at AstraZeneca, reported by Brown et al. [59],
showed that active antibacterial project compounds were significantly more polar
than the screening collection average. Improving biochemical potency through
chemical modification of active leads often came with an increase in hydrophobicity
and an increased probability of problematic plasma protein binding or cytotoxicity.
In cases in which biochemical potency was maintained by increasing polarity,
whole-cell activity remained elusive; designing polar compounds was not sufficient
for antibacterial activity. Overall, the study highlights the complexities of bacterial
cell penetration and efflux systems, especially in Gram-negative bacteria. The
authors note that one possibility for improving the antibiotic chemical space of
screening libraries would be to return to natural-product screening.
While there is little overlap in the chemical space of compounds in screening
libraries with that of antibacterials, there is far more overlap in the physiochemical
properties between antibiotics and natural products [60, 61]. In addition to hydro-
philicity (i.e., log P), other properties, such as the number of rotatable bonds
(molecular flexibility), polar surface area, H-bond donors and acceptors, molecular
complexity, and 3-dimensionality [62, 63], are better represented in natural-product
chemical space.
550 F. R. McSorley et al.

17.4 Traditional Natural-Product Discovery

Selman Waksman is credited for developing a procedure in which microbial exu-

dates are screened for cell growth inhibition on the surface of solid agar medium
plates. This method measures “zones of inhibition” around paper disks to which
natural-product samples are applied [16]. The so-called Waksman Platform is much
faster and more efficient than systematic testing for antibiotic efficacy in animal
disease models, as performed by Ehrlich. When Waksman used the method for
high-throughput analysis of soil microbe products, he discovered candicidin, the
first polyene antifungal agent; streptomycin and neomycin, the first aminoglyco-
sides; and many other agents that include streptothricin and actinomycin. Many
clinically used antibiotics were subsequently found using this method: chlortetra-
cycline (Lederle), chloramphenicol (Parke-Davis), erythromycin (Abbott and
Lilly), and tetracycline (Pfizer) [64]. After successfully mining soil-derived bacte-
ria, specifically streptomycetes, the returns dwindled as known compounds repeat-
edly surfaced in the screens [65]. Consequently, the natural-product screening
programs of drug companies slowly shut down, and the focus switched to synthetic
chemistry.
Traditionally, antibiotic discovery using the Waksman platform begins with a
source of environmental microbes. These have been primarily obtained from soil
samples collected by the employees, family, and associates of drug companies
across the globe. The microbes in these samples generally focused on the
­actinomycetes, spore-forming bacteria that over the decades led to collections
containing tens to hundreds of thousands of strains. The producer strains are typi-
cally grown in a variety of defined media, since the contents of the medium can
significantly impact the production of a given compound. Following fermenta-
tion, organic solvent extracts or conditioned media samples are prepared and used
for screening against a set of pathogenic bacteria. If a natural-product extract
elicits antibiotic properties, then activity-guided purification is conducted to iso-
late the bioactive molecule, and the chemical structure of the active molecule is
elucidated. If the chemical hit is promising, then semisynthetic or total synthetic
variations of the lead compound are produced and tested. From hundreds of ana-
logs, a therapeutic candidate may emerge. Large-scale production of the opti-
mized lead compound is undertaken, and extensive safety tests are carried out
before the candidate enters clinical trials. Three phases of appropriate clinical
trials are performed, and if the candidate agent passes, it would proceed to the
regulatory approval step. The discovery and development pipeline of an antibiotic
can take upward of 10 years and cost hundreds of millions of dollars. Bérdy esti-
mated that ~28,000 antimicrobial natural products from microbial sources have
been reported using this approach. It is for this reason, along with the drought in
discovery of antibiotics using other chemical matter, that many pharmaceutical
companies have withdrawn investments and shut down antibiotic discovery pro-
grams [66].
17 Natural Products in Antibiotic Discovery 551

17.5 The Future of Natural-Product Discovery

The time is right for a reevaluation of natural products in antibiotic drug discovery.
Their historical success as drugs, the comparative shortcomings of screens of syn-
thetic compound libraries, and the serious need for innovation in securing new anti-
biotics demand a fresh look at this source of bioactive chemistry. The rediscovery of
well-known chemical scaffolds, which prompted a move away from natural prod-
ucts, can be mitigated in several ways. First, previously unsuccessful scaffolds can
be reevaluated; second, successful antibiotic drugs can be reinvigorated by combin-
ing them with inhibitors of resistance and other antibiotic adjuvants; third, new
scaffolds can be sourced from previously neglected genera or through mining of
microbial genomes and metagenomes; and finally, “new-to-nature” compounds can
be generated through synthetic biology strategies.

17.5.1 Revisiting Discarded Scaffolds

The three most recent natural-product antibiotics to enter the clinic, daptomycin,
fidaxomicin, and the pleuromutilins, were all discovered and discarded decades
before their successful clinical launch. In the case of daptomycin, off-target human
toxicity was deemed a sufficient concern by Eli Lilly to halt clinical development.
A decade later, with more careful drug dosing to avoid undesired effects, daptomy-
cin was championed by Cubist Pharmaceuticals, which successfully brought the
compound to market [43]. Fidaxomicin was discarded in the 1970s due to poor solu-
bility and narrow spectrum, properties that are advantages in its new incarnation as
an orally dosed drug to combat C. difficile [53]. These examples offer hope, perhaps
even certainty, that new antibiotic drugs can be sourced from known compounds.
The estimate that ~28,000 natural-product antibiotics have been reported, while
fewer than 500 have entered into clinical use, is encouraging that we can revisit
these compounds as antimicrobial sources.
There are challenges to this route, however. A practical consideration is that there
is no ready way to obtain these compounds for reevaluation. Most were reported in
the scientific or patent literature decades ago, but some remain in the yellowing lab
books in the vaults of pharma. Unless the compounds progressed in the development
process, the bacterial strains that produce them may not be available in public cul-
ture collections. The fate of the extensive libraries of producing organisms held by
many companies active during the 1950s–1980s is not widely known. Some have
been captured by new entities. For example, the historical Merck collections are now
foundational resources of Fundación MEDINA and the Natural Products Discovery
Institute. Most strain libraries, however, are not easy to access. This means that
interesting chemical scaffolds may be lost until rediscovered by traditional screens.
Another challenge is securing intellectual property on known natural compounds
and their activities [67]. Nevertheless, a deep reservoir of knowledge and chemistry
exists that can be tapped for twenty-first-century antibiotic drug discovery.
552 F. R. McSorley et al.

17.5.2  atural-Product Adjuvants, Resistance Inhibitors,

N
and Combination Therapies

All known antibiotic-producing microbes have multiple biosynthetic programs that

generate additional natural products. In most cases, these appear to be unrelated to
production of the antibiotic of interest, but there are examples in which the addi-
tional products are co-expressed to achieve improved antibiotic efficacy. Indeed,
coproduction to achieve synergy between molecules may be commonplace in pro-
ducing microbes [68]. This can include coproduction of nonantibiotic adjuvants that
enhance antibiotic activity or inhibit resistance. It can also include co-expression of
two (or more) antibiotic compounds that act synergistically. Examples of the latter
are the streptogramin antibiotics (described above in Sect. 17.2). Streptogramin pro-
ducers, such as Streptomyces pristinaespiralis, produce type A and B compounds in
a ratio of ~7:3. Binding of the type A streptogramin to the peptidyl transferase cen-
ter of the bacterial ribosome enhances binding of the type B antibiotic to the region
of the peptide exit tunnel by ~100-fold, thereby accounting for the observed synergy
(reviewed in Ref. 40).
Antibiotic adjuvants have little or no antimicrobial activity themselves, but they
enhance antibiotic activity by facilitating transport or blocking resistance [69, 70].
The discovery of clavulanic acid, a potent inactivator of β-lactamases produced by
the cephamycin C-producer Streptomyces clavuligerus, demonstrated that antibiotic
producers can “protect their investments” by co-expressing inhibitors of resistance.
Several other cephamycin producers also express clavulanic acid, suggesting that
the strategy of producing both antibiotics and inhibitors of resistance may be com-
mon. We have prepared a cell-based platform for the screening of resistance inhibi-
tors that can also be used in the rapid identification (and dereplication) of known
antibiotic scaffolds [71]. Using this platform, we identified an inactivator of metallo-­
β-­lactamases, including NDM-1 produced by a strain of Aspergillus versicolor [72].
This strain also has the biosynthetic machinery to produce a β-lactam antibiotic
(unpublished observation). There is little doubt that many other antibiotic–adjuvant
pairs exist in nature. Indeed, plant-derived natural products also show efficacy as
adjuvants [73].
Screening for lethal synergy is another strategy for extending the life of antibiot-
ics. For example, the combination of bacteriostatic inhibitors of gene expression,
such as tetracycline, rifampicin, and chloramphenicol, with the bacteriostatic com-
pound bicyclomycin (Fig. 17.10), an inhibitor of the Rho transcription terminator,
caused rapid killing of Gram-negative bacteria [74]. Screens for antimicrobial syn-
ergy often use growth inhibition assays and select for increased bacteriostatic activ-
ity; however, time–kill assays can be employed to screen for lethal synergy
combinations, which cause rapid killing and may diminish the rate of resistance.
The significant challenge in bringing such combinations to market is the need to
match pharmacological and dosing properties for each component. This is not triv-
ial and often cited as a complex barrier to systematic exploration of such pairs.
Nevertheless, a combination strategy is routine in the treatment of cancer, HIV dis-
ease, and even bacterial infections such as tuberculosis.
17 Natural Products in Antibiotic Discovery 553

H
A B C N NH2
O
O
O OH CONH2 N
N
HO2C HN O H
H O
HO2C N O NH O O O O HN
N CO2H H H H H
H HO OH MeHN N N N N
NH2 CO2H N N N N O
Me OH H H H H
O O O O
aspergillomarasmine A (AMA) bicyclomycin OH OH

teixobactin

Fig. 17.10 (A) Aspergillomarasmine A (AMA) is an inhibitor of metallo-β-lactamases (e.g.,

NDM-1) and is in preclinical development as an adjuvant with β-lactam antibiotics [72]. (B)
Bicyclomycin, an inhibitor of Rho transcription terminator, exhibits lethal synergy with inhibitors
of gene expression [74]. (C) Teixobactin is a cyclic depsipeptide that binds to lipid II [75]

OH OH Me Me OH Cl
HO O O
Me CO2H Me
Me O Me O O Cl OH O OH O NH2
HO2C
O
OH O
mupirocin enacyloxin IIa
(= pseudomonic acid A) Me

Me H
OH O Me O
Cl O NH
Cl
HO OH
Me O
H N NH2 O
MeO N Me H2N
O O N O Me O
Me O N CO2H Cl
O Me O O H
myxopyronin armeniaspirole C pantocin A salinosporamide A

Fig. 17.11 Antibiotics produced by Gram-negative bacteria and marine bacteria

17.5.3 Mining New Sources of Microbial Natural Products

The bulk of the actinomycetes screened by the pharmaceutical industry for antibi-
otic activities originate from soil environments. These sources are easy to access
and offer a wide variety of conditions for enriching various genera. The question of
whether such sampling reflects a reasonable representative distribution of microbial
natural-product diversity is unresolved. The same common scaffolds can be readily
found in samples from around the world, supporting the axiom that “everything is
everywhere, the environment selects.” However, careful genomic sampling of sig-
nature natural-product biosynthetic elements, such as ketosynthase domains from
polyketide synthases and adenylation domains of non-ribosomal peptide synthases
from a variety of soil environments, suggests that indeed there are significant envi-
ronmental differences in natural-product potential: there is significant chemical
diversity still to be identified [76]. Indeed, marine actinomycetes are sources of
several new compounds, many having biological activity (Fig. 17.11).
The fact that most of the focus in antibiotic natural-product discovery has been
on the actinomycetes has prompted a search in other orders of bacteria. The Gram-­
negative betaproteobacteria, such as members of the genus Burkholderia, are prodi-
gious producers of antibiotics [77]. The gamma-proteobacteria, pseudomonads, and
the deltaproteobacteria, such as the Myxococci [78], also produce numerous natural-­
product antibiotics (Fig. 17.11). These have only just begun to be mined to discover
new chemical scaffolds.
554 F. R. McSorley et al.

These sources though, are limited to strains that we can readily grow in the labo-
ratory. The “great plate count anomaly” refers to the fact that we are generally lim-
ited to growing <5% of the detectable microbes in a soil sample. Strategies to mine
this “microbial dark matter” offer ways to access new microbial genetic and chemi-
cal diversity [79]. An example of this approach is the iChip, a simple 96-­compartment
device to capture microbes and grow them in situ, with access to nutrients and
growth factors of their natural environment [80]. Using this device and strategy, a
new antibiotic, teixobactin, which represents a new scaffold, was identified from a
previously uncultured bacterium [75]. Teixobactin, produced by a Gram-negative
bacterium, has a Gram-positive-only profile. Its mode of action involves binding to
lipid II, which is required for cell wall biosynthesis. Mining other difficult-to-grow
bacteria for new chemistry should be possible and therefore offers hope that addi-
tional antibiotic scaffolds can be identified.

17.5.4 Genome and Metagenome Mining

Advancements in microbiology and molecular biology techniques have enabled the

culture of microbes that previously were difficult to access. Advances in next-­
generation sequencing (NGS) are providing unequaled access to the genomic details
of these organisms. Coupled with automated in silico prediction algorithms, such as
antiSMASH, to identify biosynthetic gene clusters [81], this new genomic informa-
tion has revealed a previously unappreciated and remarkable quantity and genetic
diversity of natural products that can be (at least in principle) synthesized by
microbes. On average, actinomycetes encode in their genomes 20–40 natural-­
product biosynthetic gene clusters; fungi encode even more. This new reality offers
unprecedented opportunity to mine previously unknown or overlooked chemical
diversity. Many of these compounds are difficult to detect and/or are found in low
abundance. However, recent advances in mass spectrometry-based sampling and
automated compound analysis and identification using artificial intelligence analy-
sis (e.g., [82–85]) enable rapid triage for novelty that was inaccessible from tradi-
tional activity-guided purification and characterization methods. Such approaches
are yielding new antibiotic scaffolds, such as the telomycins (Fig. 17.12), that target
components of the bacterial membrane [86].
Often the expression of biosynthetic gene clusters in the laboratory is challeng-
ing, thereby preventing testing or purification of new compounds. Strategies to acti-
vate such “silent” clusters are being explored, although none is universal [88–91].
This includes the deletion or overexpression of regulatory genes, addition of chemi-
cal perturbants, physical stress (e.g., pH, temperature), and selection of mutants of
various genes, such as encoded ribosomal proteins and antibiotic resistance. Failing
such strategies, capture of entire clusters and mobilization to surrogate hosts can be
used [92, 93].
17 Natural Products in Antibiotic Discovery 555

OH
O
H
OH N N
NH2 O HN N OH
H H
N O O NH
HO2C N O HN O
H
O OH O
O O
H
O N NH
N
H HN NH HN NH
HO O HN
telomycin HN turbomycin A turbomycin B

Fig. 17.12 Telomycin and turbomycins A and B [86, 87]

While genome mining has greatly expanded our access to known and new anti-
biotic scaffolds, the majority of environmental microbes remain difficult to culture.
Here, metagenomic strategies in which total DNA is collected from a source (e.g., a
soil sample, animal, or plant microbiomes) are being employed. Such strategies are
yielding new antimicrobial compounds, such as the turbomycins [87], variants of
glycopeptides [94], and colicins [95] (Fig.17.12).

17.5.5 Increasing Diversity Through Synthetic Biology

The long-term future for obtaining antibiotic diversity may be through the genera-
tion of nonnatural or synthetic natural products. This oxymoron refers to the engi-
neering of biosynthetic gene clusters to produce novel compounds, not yet known
to nature, using synthetic biology concepts [96–98]. The modularity of biosynthetic
gene clusters lends itself well to systematic synthetic biology. Biosynthetic gene
clusters include a predictable parts list: genes encoding scaffold assembly, tailoring
enzymes, supply of components not easily scavenged from primary metabolism
(amino acids, sugars, etc.), self-resistance, regulation, and transport (Fig. 17.13). In
principle, these elements can be mixed and matched to generate new compounds
having novel activities. For example, we have used this approach to generate new-­
to-­nature glycopeptide antibiotics that evade certain forms of resistance in VRE [99]
(Fig. 17.13).
As the costs of DNA synthesis continue to drop, one can envision synthesis of
large numbers of biosynthetic gene cluster parts, the combinatorial generation of
libraries of scaffolds, tailoring enzymes, regulatory elements, etc., and their
expression in a suitable heterologous host. The result would be millions of previ-
ously untested combinations of biosynthetic genes (Fig. 17.13). With suitable
selection, such libraries could deliver hits and lead for new antibiotic drug
development.
 A Teicoplanin (Actinoplanes teichomyceticus)
556

Vancomycin (Actinoplanes orientalis)

Ristocetin (Amycolatopsis lurida)

A47934 (Streptomyces toyocaensis)

NRPS Bht biosynthesis Resistance O-Methylation N-Methylation Acylation

Hpg biosynthesis Crosslinking Regulation C-Methylation Sulfation Miscellaneous
Dpg biosynthesis Halogenation Export Sugar biosynthesis / Glycosyltransfer / Sugar O-Methylation
Core Scaffold Genes Other Tailoring Genes
gtfEvan agtfB
B Teicoplanin-type Biosynthetic cluster Tailoring Gene(s) Glc GlcNAc
Minimal scaffold for A47934 or e.g., methyl-, sulfo-, or
desulfo-A47934 glycosyltransferases gtfEvan + orf14pek
OH stfpek
Cl R5 2-O-Me-Glc
O O SO3 O
Cl
Cl O O
HO R7
O O
H H O Cl
N N O O
O N N N O H H
H H H N N O
HN O O NH2 O N N N
Vector combination for H H H
HO2C HN O O NH
Cl O heterologous expression in
O HO O R1
HO H Streptomyces coelicolor Cl 4 O
HO O O R O
R2 R8 O HO R6 R3 O
A47934: R2 = SO3H HO
R2 mtfApek
desulfo-A47934: R2 = H orf19ris orf22ris orf23ris
Me
Me mannose Me 2
R =H
C MIC (mg/mL) of New Desulfo-A47934 Derivatives or SO3H

desulfo- R3 = Me R5 = GlcNAc R5 = Glc R5 = 2-O-Me-Glc R6 = mannose R7 = SO3H R8 = Me

vancomycin A47934 derivative derivative derivative derivative derivative derivative derivative
E. faecalis (ATCC 29212) 4 0.5 0.5 2 4 2 2 1 1
E. faecalis VRE B (ATCC 51299) 64 1 0.5 4 4 4 2 1 1

Fig. 17.13 A synthetic biology approach for increasing chemical diversity in glycopeptide antibiotics. (A) Biosynthetic clusters for selected glycopeptides
with genes colored according to function. Significant portions of many clusters are comprised of tailoring genes responsible for decorating the glycopeptide
backbone (e.g., methyltransferases, sulfotransferases, and glycosyltransferases). (B) Novel glycopeptides have been generated by mixing biosynthetic genes.
In a recent study by Yim et al. [99], the biosynthetic clusters for A47934 and desulfo-A47934 were expressed in S. coelicolor together with a variety of tailoring
F. R. McSorley et al.

genes from the biosynthetic clusters of other glycopeptides. Several new A47934- and desulfo-A47934 derivatives were produced and are more potent than
vancomycin against E. faecalis and VRE B (C)
17 Natural Products in Antibiotic Discovery 557

17.6 Concluding Remarks

Natural products, in particular those generated by bacteria and fungi, are the source
of the majority of our successful antibiotic drugs. These agents have changed the
world of medicine. For the first time in human history, we have good control over
infection. With that control has come much of modern medicine. Our natural-­
product antibiotics have also helped us feed the world by changing the way we raise
and care for food animals. It is not hyperbole to suggest that natural-product antibi-
otics may be the most important scientific discovery of the twentieth century.
Unfortunately, the evolution of antibiotic resistance and its selection in once-­
susceptible pathogens gravely threatens these advances. We need new antibiotics
and alternatives to maintain our control over infectious disease. The advances in our
knowledge of how natural products are made by microbes, new and unparalleled
access to the genetic determinants of natural-product biosynthesis by NGS of
microbial genomes and metagenomes, and the ability to harness this information to
identify and exploit this information are growing exponentially.
The proven efficacy of natural products as antibiotics, plus the disappointing
results of the past two decades of focus on synthetic compounds, means that we
must pivot back to these ancient compounds for leads and inspiration. There is good
reason to believe that the era of resistance depicted in Fig. 17.1 will be followed by
an era of anti-infective innovation.

Major Points
• Microbial natural products are the source of most of our successful antibiotic drugs.
• These compounds are the result of evolutionary processes that select for optimal
penetration and retention in target bacterial cells.
• The chemical diversity and physiochemical properties of microbial natural prod-
ucts cannot yet be effectively matched in most synthetic libraries.
• The re-isolation of known natural-product scaffolds diminished enthusiasm for
the natural-product approach in antibiotic discovery.
• Efforts to identify new antibiotic chemical diversity through revisiting discarded
compounds, mining of bacterial genomes, isolation of hitherto rare or unsampled
microbes, and increasing chemical diversity using synthetic biology strategies
offer new routes to identifying antibiotic leads.
• The use of inhibitors of resistance or other adjuvants can also extend the clinical
effectiveness of existing antibiotic scaffolds.

References

1. Borchardt JK. The beginnings of drug therapy: ancient Mesopotamian medicine. Drug News
Perspect. 2002;15(3):187–92.
2. Scurlock J. Sourcebook for ancient Mesopotamian medicine: Society of Biblical Literature;
2014.
558 F. R. McSorley et al.

3. Cragg GM, Newman DJ. Natural products: a continuing source of novel drug leads. Biochim
Biophys Acta. 2013;1830(6):3670–95.
4. Huang KC. The pharmacology of chinese herbs. 2nd ed. Boca Raton: CRC Press; 1999.
5. Kapoor LD. CRC handbook of ayurvedic medicinal plants. Boca Raton: CRC Press; 1990.
6. Sertuerner F. Ueber das Morphium, eine neue salzfähige Grundlage, und die Mekonsäure, als
Hauptbestandtheile des Opiums. Ann Phys (Berl). 1817;55(1):56–89.
7. Newman DJ, Cragg GM. Chapter 1 Natural products as drugs and leads to drugs: the historical
perspective. Natural Product Chemistry for Drug Discovery: The Royal Society of Chemistry;
2009. p. 3–27.
8. Fleming A. On the antibacterial action of cultures of a Penicillium, with special reference to
their use in the isolation of B. influenzæ. Br J Exp Pathol. 1929;10(3):226–36.
9. Zasloff M. Antimicrobial peptides of multicellular organisms. Nature. 2002;415(6870):389–95.
10. Newman DJ, Cragg GM. Natural products as sources of new drugs from 1981 to 2014. J Nat
Prod. 2016;79(3):629–61.
11. Walsh CT, Wencewicz T. Antibiotics: challenges mechanisms opportunities. Washington, DC:
ASM Press; 2016.
12. Brown ED, Wright GD. Antibacterial drug discovery in the resistance era. Nature.
2016;529(7586):336–43.
13. Ehrlich P, Hata S. Die experimentelle Chemotherapie der Spirillosen:(Syphilis, Rückfallfieber,
Hühnerspirillose, Frambösie). Wiesbaden: Springer; 1910.
14. Aminov RI. A brief history of the antibiotic era: lessons learned and challenges for the future.
Front Microbiol. 2010;1:134.
15. Mahoney JF, Arnold RC, Harris A. Penicillin treatment of early syphilis-a preliminary report.
Am J Public Health Nations Health. 1943;33(12):1387–91.
16. Lewis K. Platforms for antibiotic discovery. Nat Rev Drug Discov. 2013;12(5):371–87.
17. Domagk G. Ein Beitrag zur Chemotherapie der bakteriellen Infektionen. Dtsch Med
Wochenschr. 1935;61(7):250–3.
18. Lewis K. Antibiotics: recover the lost art of drug discovery. Nature. 2012;485(7399):439–40.
19. Wright GD. Opportunities for natural products in 21st century antibiotic discovery. Nat Prod
Rep. 2017;34(7):694–701.
20. Tipper DJ, Strominger JL. Mechanism of action of penicillins: a proposal based on their struc-
tural similarity to acyl-D-alanyl-D-alanine. Proc Natl Acad Sci U S A. 1965;54(4):1133–41.
21. Rolinson GD. Forty years of beta-lactam research. J Antimicrob Chemother.
1998;41(6):589–603.
22. Drawz SM, Papp-Wallace KM, Bonomo RA. New β-lactamase inhibitors: a therapeutic renais-
sance in an MDR world. Antimicrob Agents Chemother. 2014;58(4):1835–46.
23. Wang DY, Abboud MI, Markoulides MS, Brem J, Schofield CJ. The road to avibactam: the
first clinically useful non-β-lactam working somewhat like a β-lactam. Future Med Chem.
2016;8(10):1063–84.
24. Castanheira M, Rhomberg PR, Flamm RK, Jones RN. Effect of the β-lactamase inhibi-
tor Vaborbactam combined with Meropenem against serine carbapenemase-producing
Enterbacteriaceae. Antimicrob Agents Chemother. 2016;60(9):5454–8.
25. Sykes RB, Cimarusti CM, Bonner DP, Bush K, Floyd DM, Georgopapadakou NH, Koster
WM, Liu WC, Parker WL, Principe PA, Rathnum ML, Slusarchyk WA, Trejo WH, Wells
JS. Monocyclic beta-lactam antibiotics produced by bacteria. Nature. 1981;291(5815):489–91.
26. Sykes RB, Bonner DP, Bush K, Georgopapadakou NH. Azthreonam (SQ26,776), a synthetic
monobactam specifically active against aerobic gram-negative bacteria. Antimicrob Agents
Chemother. 1982;21(1):85–92.
27. Waksman SA, Reilly HC, Johnston DB. Isolation of streptomycin-producing strains of
Streptomyces griseus. J Bacteriol. 1946;52(3):393–7.
28. Umezawa H. Kanamycin: its discovery. Ann N Y Acad Sci. 1958;76(2):20–6.
17 Natural Products in Antibiotic Discovery 559

29. Weinstein MJ, Luedemann GM, Oden EM, Wagman GH, Rosselet JA, Coniglio CT, Charney
W, Herzog HL, Black J. Gentamicin, a new antibiotic complex from micromonospora. J Med
Chem. 1963;6(4):463.
30. Garneau-Tsodikova S, Labby KJ. Mechanisms of resistance to aminoglycoside antibiotics:
overview and perspectives. Medchemcomm. 2016;7(1):11–27.
31. Katz L, Ashley GW. Translation and protein synthesis: macrolides. Chem Rev.
2005;105(2):499–528.
32. Gomes C, Martinez-Puchol S, Palma N, Horna G, Ruiz-Roldán L, Pons MJ, Ruiz J. Macrolide
resistance mechanisms in Enterobacteriaceae: focus on azithromycin. Crit Rev Microbiol.
2017;43(1):1–30.
33. Wright PM, Seiple IB, Myers AG. The evolving role of chemical synthesis in antibacterial
drug discovery. Angew Chem Int Ed Engl. 2014;53(34):8840–69.
34. Yang W, Moore IF, Koteva KP, Bareich DC, Hughes DW, Wright GD. TetX is a flavin-­
dependent monooxygenase conferring resistance to tetracycline antibiotics. J Biol Chem.
2004;279(50):52346–52.
35. Floss HG, Yu TW. Rifamycin-mode of action, resistance, and biosynthesis. Chem Rev.
2005;105(2):621–32.
36. Goldstein BP. Resistance to rifampicin: a review. J Antibiot (Tokyo). 2014;67(9):625–30.
37. Bugg TD, Wright GD, Dutka-Malen, S, Arthur, M, Courvalin, P, Walsh CT. Molecular basis
for vancomycin resistance in Enterococcus faecium BM4147: biosynthesis of a depsipeptide
peptidoglycan precursor by vancomycin resistance proteins VanH and VanA. Biochem. 1991;
30(43):10408–10415.
38. Kirst HA, Thompson DG, Nicas TI. Historical yearly usage of vancomycin. Antimicrob
Agents Chemother. 1998;42(5):1303–4.
39. Kahne D, Leimkuhler C, Lu W, Walsh C. Glycopeptide and lipoglycopeptide antibiotics.
Chem Rev. 2005;105(2):425–48.
40. Mukhtar TA, Wright GD. Streptogramins, oxazolidnones, and other inhibitors of bacgerial
protein synthesis. Chem Rev. 2005;105(2):529–42.
41. Wilson DN. The A-Z of bacterial translation inhibitors. Crit Rev Biochem Mol Biol.
2009;44(6):393–433.
42. Allington DR, Rivey MP. Quinupristin/dalfopristin: a therapeutic review. Clin Ther.
2001;23(1):24–44.
43. Eisenstein BI, Oleson FB Jr, Baltz RH. Daptomycin: from the mountain to the clinic, with
essential help from Francis Tally, MD. Clin Infect Dis. 2010;50(1):S10–5.
44. Taylor SD, Palmer M. The action mechanism of daptomycin. Bioorg Med Chem.
2016;24(24):6253–68.
45. Ordooei Javan A, Shokouhi S, Sahraei Z. A review on colisten nephrotoxicity. Eur J Clin
Pharmacol. 2015;71(7):801–10.
46. Velkov T, Thompson PE, Nation RL, Li J. Structure-activity relationships of polymyxin anti-
biotics. J Med Chem. 2010;53(5):1898–916.
47. Yu Z, Quin W, Lin J, Fang S, Qiu J. Antibacterial mechanisms of polymyxin and bacterial
resistance. Biomed Res Int. 2015;2015:679109.
48. Srinivas P, Rivard K. Polymyxin resistance in gram-negative pathogens. Curr Infect Dis Rep.
2017;19(11):38.
49. Miller WR, Bayer AS, Arias CA. Mechanism of action and resistance to daptomycin in
Staphylococcus aureus and Enterocci. Cold Spring Harb Perspect Med. 2016;6(11):a026997.
50. Novak R. Are pleuromutilin antibiotics finally fit for human use? Ann N Y Acad Sci.
2011;1241:71–81.
51. Dinos GP, Athanassopoulos CM, Missiri DA, Giannopoulou PC, Vlachogiannis IA,
Papadopoulos GE, Papaioannou D, Kalpaxis DL. Chloramphenicol derivatives as antibac-
terial and anticancer agents: historic problems and current solutions. Antibiotics (Basel).
2016;5(2):20.
560 F. R. McSorley et al.

52. Silver LL. Fosfomycin: mechanism and resistance. Cold Spring Harb Perpect Med.
2017;7(2):a025262.
53. Fidaxomicin. A novel agent for the treatment of Clostridium difficile infection. Can J Infect
Dis Med Microbiol. 2015;26(6):305–12.
54. Payne DJ, Gwynn MN, Holmes DJ, Pompliano DL. Drugs for bad bugs: confronting the chal-
lenges of antibacterial discovery. Nat Rev Drug Discov. 2007;6(1):29–40.
55. Tommasi R, Brown DG, Walkup GK, Manchester JI, Miller AA. ESKAPEing the labyrinth of
antibacterial discovery. Nat Rev Drug Discov. 2015;14(8):529–42.
56. Lipinski CA, Lombardo F, Dominy BW, Feeney PJ. Experimental and computational
approaches to estimate solubility and permeability in drug discovery and development set-
tings. Adv Drug Deliv Rec. 2001;46(1–3):3–26.
57. O'Shea R, Moser HE. Physicochemical properties of antibacterial compounds: implications
for drug discovery. J Med Chem. 2008;51(10):2871–8.
58. Mugumbate G, Overington JP. The relationship between target-class and the physiochemical
properties of antibacterial drugs. Bioorg Med Chem. 2015;23(16):5218–24.
59. Brown DG, Tl M-D, Gagnon MM, Tommasi R. Trends and exceptions of physical proper-
ties on antibacterial activity for Gram-positive and Gram-negative pathogens. J Med Chem.
2014;57(23):10144–61.
60. Ganesan A. The impacet of natural product upon modern drug discovery. Curr Opin Chem
Biol. 2008;12(3):306–17.
61. Harvey AL, Edrada-Ebel R, Quinn RJ. The re-emergene of natural products for drug discovery
in the genomics era. Nat Rev Drug Discov. 2015;14(2):111–29.
62. Lovering F, Bikker J, Humblet C. Escape from flatland: increasing saturation as an approach
to improving clinial success. J Med Chem. 2009;52(21):6752–6.
63. Lovering F. Escape from Flatland 2: complexity and promiscuity. Med Chem Commun.
2013;4(3):515–9.
64. Buggs CW. Ten years after streptomycin; past and current practice in antibiotic therapy. J Natl
Med Assoc. 1957;49(3):142–9.
65. Baltz RH. Antimicrobials from actinomycetes: back to the future. Microbe. 2007;2(3):125.
66. Katz L, Baltz RH. Natural product discovery: past, present, and future. J Ind Microbiol
Biotechnol. 2016;43(2–3):155–76.
67. Chang B-C, Wang S-J. The impact of patent eligibility on biotech patents: a flow chart for
determining patent eligibility and an immune therapy case study. Hum Vaccin Immunother.
2015;11(3):789–94.
68. Challis GL, Hopwood DA. Synergy and contingency as driving forces for the evolution of
multiple secondary metabolite production by Streptomyces species. Proc Natl Acad Sci U S A.
2003;100(2):14555–61.
69. Wright GD. Antibiotic adjuvants: rescuing antibiotics from resistance. Trends Microbiol.
2016;24(11):862–71.
70. Melander RJ, Melander C. The challenge of overcoming antibiotic resistance: an adjuvant
approach? ACS Infect Dis. 2017;3(8):559–63.
71. Cox G, Siron A, King AM, De Pascale G, Pawlowski AC, Koteva K, Wright GD. Cell Chem
Biol. 2017;24(1):98–109.
72. King AM, Reid-Yu SA, Wang W, King DT, De Pascale G, Strynadka NC, Walsh TR, Coombes
BK, Wright GD. Aspergillomarasmine A overcomes metallo-β-lacamse antibiotic resistance.
Nature. 2014;510(7506):503–6.
73. Stermitz FR, Lorenz P, Tawara JN, Zenewicz LA, Lewis K. Synergy in a medicinal plant:
antimicrobial action of berberine potentiated by 5′-methoxyhydnocarpin, a multidrug pump
inhibitor. Proc Natl Acad Sci U S A. 2000;97(4):1433–7.
74. Malik M, Li L, Zhao X, Kerns RJ, Berger JM, Drlica K. Lethal synergy involving bicyclomy-
cin: an approach for reviving old antibiotics. J Antimicrob Chemother. 2014;69(12):3227–35.
75. Ling LL, Schneider T, Peoples AJ, Spoering AL, Engels I, Conlon BP, Mueller A, Schäberle
TF, Hughes DE, Epstein S, Jones M, Lazarides L, Steadman VA, Cohen DR, Felix CR,
17 Natural Products in Antibiotic Discovery 561

Fetterman KA, Millet WP, Nitti AG, Zullo AM, Chen X, Lewis K. A new antibiotic kills
pathogens without detectable resistance. Nature. 2015;517(7535):455–9.
76. Charlop-Powers Z, Owen JG, Reddy BV, Ternei MA, Guimarães DO, de Frias UA, Pupo MT,
Seepe P, Feng Z, Brady SF. Global biogeographical sampling of bacterial secondary metabo-
lism. elife. 2015;4:e05048.
77. Masschelein J, Jenner M, Challis GL. Antibiotics from Gram-negative bacteria: a comprehen-
sive overview and selected biosynthetic highlights. Nat Prod Rep. 2017;34(7):712–83.
78. Hermann J, Fayad AA, Müller R. Natural products from myxobacteria: novel metabolites and
bioactivities. Nat Prod Rep. 2017;34(2):135–60.
79. Lok C. Mining the microbial dark matter. Nature. 2015;522(7556):270–3.
80. Nichols D, Cahoon N, Trakhtenber EM, Pham L, Mehta A, Belanger A, Kanigan T, Lewis K,
Epstein SS. Use of ichip for high-throughput in situ cultivation of “uncultivable” microbial
species. Appl Environ Microbiol. 2010;76(8):2445–50.
81. Blin K, Wolf T, Chevrette MG, Lu X, Schwalen CJ, Kautsar SA, Suarez Duran HG, de Los
Santos ELC, Kim H, Nave M, Dickschat JS, Mitchell DA, Shelest E, Breitling R, Takano E,
Sy L, Webe T, Medema MH. antiSMASH 4.0-improvements in chemistry predictions and
gene cluster boundary identification. Nucleic Acids Res. 2017;45(W1):W36–41. https://doi.
org/10.1093/nar/gkx319.
82. Ibrahim A, Yang L, Johnston C, Liu X, Ma B, Magarvey NA. Dereplicating nonribosomal
peptides using an informatic search algorithm for natural products (iSNAP) discovery. Proc
Natl Acad Sci U S A. 2012;109(47):19196–201.
83. Johnston CW, Skinnider MA, Wyatt MA, Li X, Ranieri MR, Yang L, Zecehl DL, Ma B,
Magarvey NA. An automated Gencomes-to-Natural Products platform (GNP) for the discov-
ery of modular natural products. Nat Commun. 2015;8:8421.
84. Medema MH, Osbourn A. Computational genomic identification and functional reconstitution
of plant natural product biosynthetic pathways. Nat Prod Rep. 2016;33(8):951–62.
85. Skinnifer MA, Merwin NJ, Johnston CW, Magarvey NA. PRISM 3: expanded predic-
tion of natural product chemical structures from microbial genomes. Nucleic Acids Res.
2017;45(W1):W49–54.
86. Johnston CW, Skinnider MA, Ca D, Rees PN, Chen GM, Walker CG, French S, Brown ED,
Bérdy J, Liu DY, Magarvey NA. Assembly and clustering of natural antibiotics guides target
identification. Nat Chem Biol. 2016;12(4):233–9.
87. Gillespie DE, Brady SF, Bettermann AD, Cianciotto NP, Liles MR, Rondon MR, Clardy J,
Goodman RM, Handelsman J. Isolation of antibiotics turbomycin A and B from a metage-
nomic library of soil microbial DNA. Appl Environ Microbiol. 2002;68(9):4301–6.
88. Tanaka Y, Tokuyama S, Ochi K. Activation of secondary metabolite-biosynthetic gene
clusters by generating rsmG mutations in Streptomyces griseus. J Antibiot (Tokyo).
2009;62(12):669–73.
89. Tanaka Y, Kasahara K, Hirose Y, Murakami K, Kugimiya R, Ochi K. Activation and products
of the cryptic secondary metabolite biosynthetic gene clusters by rifampin resistance (rpoB)
mutations in actinomycetes. J Bacteriol. 2012;195(10):2959–70.
90. Yoon V, Nodwell JR. Activating secondary metabolism with stress and chemicals. J Ind
Microbiol Biotechnol. 2014;41(2):415–24.
91. Onaka H. Novel antibiotic screening methods to awaken silent or cryptic secondary metabolic
pathways in actinomycetes. J Antibiot (Tokyo). 2017;70(8):865–70.
92. Yamanaka K, Reynolds KA, Kersten RD, Tyan KS, Gonzalez DJ, Nizet V, Dorrestein PC,
Moore BS. Direct cloning and refactoring of a silent lipopeptide biosynthetic gene cluster
yields the antibiotic taromycin A. Proc Natl Acad Sci U S A. 2014;111(5):1957–62.
93. Li Y, Li Z, Yamanaka K, Xy Y, Zhang W, Wlamakis H, Kolter R, Moore BS, Qian PY. Directed
natural product biosynthesis gene cluster capture and expression in the model bacterium
Bacilus subtilis. Sci Rep. 2015;5(9383).
94. Banik JJ, Brady SF. Cloning and characterization of new glycopeptide gene clusters found in
environmental DNA megalibrary. Proc Natl Acad Sci U S A. 2008;105(45):17273–7.
562 F. R. McSorley et al.

95. Cohen LJ, Han S, Huang YH, Brady SF. Identification of the colicin V bacteriocin gene cluster
by functional screening of a human microbiome metagenomic library. ACS Infec Dis; 2017.
4(1):27–32.
96. Wright GD. Perspective: synthetic biology revives antibiotics. Nature. 2014;509(7498):S13.
97. Thaker MN, Wright GD. Opportunities for synthetic biology in antibiotics:expanding glyco-
peptide diversity. ACS Synth Biol. 2015;4(3):195–206.
98. Braff D, Shis D, Collins JJ. Synthetic biology platform technologies for antimicrobial applica-
tions. Adv Drug Deliv Rev. 2016;105(Pt A):35–43.
99. Yim G, Wang W, Thaker MN, Tan S, Wriight GD. How to make a glycopeptide: asynthetic
biology approach to expand glycopeptide antibiotic chemcial diversity. ACS Infect. Dis.
2016;2(9):642–50.
Chapter 18
The New Versus Old Target Debate
for Drug Discovery

Alice L. Erwin

18.1 What Debate?

If I had been invited to write on this topic a couple of decades ago, I would not have
understood what there was to debate. By mid-2000, complete genome sequences
had been published for a dozen or so bacterial pathogens, providing a wealth of
potential targets. For those of us whose careers in antibiotic discovery started just
before the turn of the century, target evaluation seemed extremely simple. At that
time it seemed obvious to me and to my coworkers that new targets were better than
old. We argued that for drugs with new mechanisms there would be no pre-existing
resistance. A second argument was that while the empiric methods of the past had
identified only a small number of antibiotic classes and even fewer targets, new
technologies would allow us to cast our net much more widely and be much more
productive.
At that time, I saw the need for new antibiotics to replace drugs for which resis-
tance had become common. I had no idea of the limitations of existing antibiotics
other than resistance. Moreover, it did not occur to me to wonder whether inhibitors
of the new targets (mostly enzymes) would be as effective as existing antibiotics
(most of which target the machinery of macromolecular synthesis).
Today, I consider that one of the most important advantages of new targets is the
possibility of finding drugs that are not only new but in some way better than current
antibiotics. Features that might be considered desirable for new anti-infective drugs
include reduced likelihood of resistance emergence, improved activity against per-
sistent infections, or better safety, including less disruption of normal flora.
The sections below will present my view of the advantages and risks of old and
new targets. I will illustrate my discussion with examples of antibacterial drugs that

A. L. Erwin (*)
Erwin Consulting, Seattle, WA, USA

© Springer International Publishing AG, part of Springer Nature 2018 563

I. W. Fong et al. (eds.), Antimicrobial Resistance in the 21st Century, Emerging Infectious
Diseases of the 21st Century, https://doi.org/10.1007/978-3-319-78538-7_18
564 A. L. Erwin

were recently approved or are currently in clinical trials, as well as some interesting
new compounds with efficacy in animals. These can be considered in three groups:
• Inhibitors of three good old targets, including both new members of old chemical
classes and first-in-class antibiotics with similar mechanisms to old drugs.
Contrary to the idea that old targets were not amenable to modern methods,
medicinal chemistry was guided by structural and biochemical data in addition
to antibacterial assays for many of these.
• New antibiotics with other, well-defined targets, for which knowledge of the
target was used in evaluating analogs. This list is short, as target-directed pro-
grams have not yet been very successful.
• Antibacterial drugs with new, often complex mechanisms. Some of these were
discovered empirically, with the mechanism of action determined later (if at all).
Others are not antibiotics per se but increase the susceptibility of bacteria to host
defenses or to other drugs. Some were designed to address aspects of infection
not well handled by existing antibiotics.
This review is not intended to provide a complete list of antibacterial drugs in the
clinical development pipeline. The reader is referred to recent reviews [12], to the
NIH database https://clinicaltrials.gov/, and to the pipeline web pages maintained
by the Pew Charitable Trusts Antibiotic Resistance Project (http://www.pewtrusts
.org/en/multimedia/data-visualizations/2014/antibiotics-currently-in-clinical-
development; http://www.pewtrusts.org/en/multimedia/data-visualizations/2017/
nontraditional-products-for-bacterial-infections-in-clinical-development).

18.2  e Have New Antibiotics in the Pipeline with the Good

W
Old Targets

The vast majority of antibiotics inhibit synthesis of macromolecules. Three mecha-

nisms dominate, with several chemical classes of inhibitors known for each. New
drugs with the same or similar mechanisms include both new scaffolds and new
members of old chemical classes.

18.2.1  ell Wall: Inhibitors of Transpeptidases

C
and β-Lactamases

Starting with the discovery of benzylpenicillin in 1928, inhibition of the transpepti-

dases (penicillin-binding proteins, PBPs) involved in peptidoglycan synthesis has
been one of the most important antibiotic mechanisms. Until recently, all chemical
classes of PBP inhibitors were β-lactams, and the first member of each class to be
discovered was a natural product. Additional penicillins, cephalosporins,
18 The New Versus Old Target Debate for Drug Discovery 565

carbapenems, and monobactams were often synthetic or semisynthetic. Derivatives

of the penicillin precursor 6β-amino-penicillanic acid (6-APA) allowed improve-
ments in ease of administration, broadening of bacterial spectrum, or restoration of
activity toward isolates that had become resistant. For example, benzylpenicillin
was administered by injection and was active only against Gram-positive bacteria.
Ampicillin, first used in 1961, can be taken orally and is active against some Gram-­
negative species, including Escherichia coli and Proteus mirabilis, as well as Gram-­
positive bacteria. By the late 1950s, many isolates of Staphylococcus aureus had
acquired penicillinase genes, making them resistant to the early penicillins but not
to penicillinase-stable semisynthetic penicillins or to the first-generation cephalo-
sporins. For Gram-negative bacteria, the spread of β-lactamase genes was countered
by co-administering β-lactamase inhibitors (BLIs) with β-lactam antibiotics. The
first such combination, clavulanic acid plus amoxicillin, was approved in 1984 as
Augmentin and is still widely used. Two other BLIs, tazobactam and sulbactam,
were also developed during the twentieth century. Like clavulanic acid, these are
themselves β-lactams and function by forming a stable acyl–enzyme complex, thus
poisoning the β-lactamase enzyme. These BLIs have little antibacterial activity of
their own and are always used in combination with other β-lactams, though the
activity of sulbactam for Acinetobacter baumannii has been studied recently [73].
New β-lactamase inhibitor scaffolds The continued evolution of β-lactamases in
response to the introduction of new β-lactam antibiotics spurred the search for novel
inhibitors of PBPs and/or β-lactamases. The diazabicyclooctane (DBO) class, dis-
covered at Roussel Uclaf, is the first non-β-lactam scaffold to be useful as inhibitors
of β-lactamases, with avibactam the first member to be developed [14, 103]. The
DBOs differ in mechanism from β-lactam BLI, in that covalent binding of avibac-
tam to the enzyme is described as slowly reversible [25]. The first DBO-β-lactam
combination to be approved was Avycaz (avibactam-ceftazidime, in 2015); other
avibactam-β-lactam combinations are currently in clinical trials. Relebactam (previ-
ously MK-7655) is currently in phase 3 studies, in combination with imipenem and
cilastatin (Merck). Two other DBOs, RG6080 (Meiji/Fedora) and zidebactam
(Wockhardt), are currently in phase 1.
Another new BLI class is the cyclic boronates, which are not acylase inhibitors
but transition-state analogs. The most advanced of the boronates is vaborbactam,
being developed by The Medicines Company. A meropenem/vaborbactam combi-
nation recently completed a phase 3 study.
Non-β-lactam PBP inhibitors It is remarkable that in decades of natural product
research, nearly all PBP inhibitors ever discovered were β-lactams. One exception
is the lactivicins, discovered in 1987, which have never been developed for clinical
use [103]. More recently, research by the Mobashery laboratory at Notre Dame
discovered synthetic PBP inhibitors in a specific search for compounds active
against PBP2a, the transpeptidase that mediates methicillin-resistance in
MRSA. Medicinal chemistry optimization of the oxadiazole scaffold identified lead
566 A. L. Erwin

molecules active against both methicillin-susceptible and methicillin-resistant S.

aureus, with efficacy in a mouse model of S. aureus infection [41].

18.2.2 Protein Synthesis: Bind 30S or 50S Ribosomal Subunits

Aminoglycosides, tetracyclines and macrolides The two largest classes of natu-

ral product protein synthesis inhibitors, the aminoglycosides and tetracyclines, have
a history that is generally similar to that of β-lactam antibiotics. Streptomycin and
chlorotetracycline, the first drugs of these classes, are unmodified natural products.
Subsequent members included both new natural products with similar chemical
structures and semi-synthetic derivatives thereof. Within a chemical class, com-
pounds differed in antibacterial spectrum, pharmacokinetics, and toxicity. As resis-
tance to early members of each class emerged, new antibiotics were developed
specifically to address those resistance mechanisms. Discovery of new tetracyclines
slowed after the 1960s until the discovery of tigecycline, a glycylcycline, at Lederle
in 1993. Plazomicin, derived from sisomicin at Achaogen, and omadacycline, a
semisynthetic derivative of minocycline discovered at Paratek, are both very much
in line with the previous history of aminoglycosides and tetracyclines, respectively.
Both are currently in phase 3 studies. The macrolide class has fewer members but a
similarly long history, beginning with erythromycin. Recent members include
nafithromycin (Wockhardt), now in phase 2, and solithromycin (Cempra), in phase
3 [29, 33, 106].
A new era of tetracycline research began with the development of chemical
methods for synthesizing tetracyclines and related compounds, at the Myers labora-
tory at Harvard. This technology was licensed to Tetraphase, who have several com-
pounds in the clinic. The most advanced of these, eravacycline, is currently in phase
3. More recent compounds are divergent in structure and have broader antibacterial
spectrum, with some active against Pseudomonas aeruginosa [20]. The Myers
group has also developed methods for synthesis of macrolides [84].
Pleuromutilins The pleuromutilin class of ribosome inhibitors was discovered in
1951, and a couple of semisynthetic derivatives have been developed for topical use.
Lefamulin (Nabriva), currently in phase 3, is the first pleuromutilin to be developed
as a systemic antibiotic [69].

Oxazolidinones In contrast to most ribosome inhibitors, oxazolidinones are not

natural products. The antibacterial activity of the scaffold was first described at
DuPont. Linezolid, the first oxazolidinone antibiotic, was discovered at Upjohn
through extensive medicinal chemistry. Since the turn of the century, new oxazolidi-
nones have been discovered through research at multiple companies. Tedizolid (pre-
viously torezolid, discovered at Dong-A and developed by Trius and Cubist) was
the next member of the class to be approved, and several others are currently in
clinical trials [3].
18 The New Versus Old Target Debate for Drug Discovery 567

18.2.3 DNA Synthesis: Inhibitors of Type II Topoisomerases

The largest class of synthetic antibiotics is the fluoroquinolones, which inhibit two
targets, the GyrA subunit of DNA gyrase and the ParC subunit of topoisomerase
IV. Many fluoroquinolones have been approved, the most recent being Baxdela
(delafloxacin) from Melinta Therapeutics; and there are others currently in various
stages of clinical development. Moreover, clinical candidates have been identified
for several new scaffolds of Type II topoisomerase inhibitors. Gepotidacin
(GlaxoSmithKline) and Zoliflodacin (Entasis Therapeutics), both in phase 2, both
inhibit gyrase A with mechanisms different from fluoroquinolones [12].
The lethality of fluoroquinolones is the result of the fact that they not only inhibit
the ATPase activity of gyrase but also stabilize the covalent enzyme–DNA complex.
Balanced inhibition of both gyrase and topoisomerase IV is important for the char-
acteristically low frequency of spontaneous mutation to high-level fluoroquinolone
resistance. Novobiocin has a much higher frequency of resistance, and this has been
attributed to its disproportionate inhibition of gyrase (GyrB) vs topoisomerase IV
(ParE). This insight inspired programs at both Vertex Pharmaceuticals and Trius
Therapeutics aimed at discovery of compounds with balanced inhibition of GyrB
and ParE. Both groups employed structure-based drug design and successfully iden-
tified dual-active compounds with efficacy in animal models that had the predicted
low frequency of resistance. Tricyclic GyrB/ParE (TriBE) inhibitors active against
both Gram-negative and Gram-positive bacteria were discovered by Trius before
their acquisition by Cubist and then by Merck. Vertex gyrase inhibitors are primar-
ily active against Gram-positive pathogens; their technology was recently licensed
by Spero Therapeutics. Using an antisense whole-cell screen, Merck scientists iden-
tified a natural product, kibdelomycin, that is also a dual inhibitor of GyrB and
ParE. Kibdelomycin is active primarily against Gram-positive bacteria and showed
efficacy in a hamster model of Clostridium difficile infection [7, 32, 37, 62, 97].

18.3 What Is Good About These Old Targets?

18.3.1 These Targets Work and Keep Working

As noted, there are multiple classes of antibiotics targeting each of the mechanisms
described above. Each has been shown to be amenable to rational drug design as
well as to the empiric methods that led to the first members of each scaffold.
Fluoroquinolones were probably the first class of antibiotics for which data on
potency toward the target enzyme as well as antibacterial activity were used to
inform design of new analogs [23]. Structural information on the interaction of com-
pounds with their targets was critical for discovery of the dual GyrB/ParE inhibitors
and was likely to have been important for the new classes of β-lactamase inhibitors
also. Ribosomal crystal structures now allow understanding of binding specificity of
new inhibitors but have not yet led to completely novel scaffolds [111].
568 A. L. Erwin

18.3.2 Low Resistance Frequency

As the impetus for discovery of new antibiotics has usually been the emergence of
resistance to existing antibiotics, it may seem paradoxical to describe drugs with
these old targets as having a low frequency of resistance. The key distinction is
between endogenous resistance, resulting from spontaneous mutations that produce
a substantial loss in susceptibility to an antibiotic, and exogenous resistance, result-
ing from acquisition of genes encoding resistance determinants. Mutations that con-
fer high-level resistance in a single step can lead to rapid selection for resistance,
sometimes within a single course of treatment. Drugs for which such mutations are
frequent cannot be used as single antibiotics because treatment failure is common.
This phenomenon was recognized in the 1950s, leading to the decision that strepto-
mycin and rifampin could safely be used in combination (typically in tuberculosis)
but neither could be used as monotherapy. In an influential paper, Lynn Silver
described the concept that the reason monotherapy is possible for many of the com-
mon broad-spectrum antibiotics is that they effectively have multiple targets [87].
For the classes of antibiotics described above, nearly all resistance is exogenous.
Acquired mechanisms of resistance include aminoglycoside-modifying enzymes,
β-lactamases, efflux pumps specific for macrolides or tetracyclines, rRNA methyl-
transferases, alternative PBPs, and other target-modifying enzymes. These resis-
tance determinants nearly always have origins in bacterial populations other than
the pathogens being treated by the antibiotic. The existence of transmissible resis-
tance does not prevent a new antibiotic from being useful, sometimes for many
decades after the first appearance of a resistance determinant. Understanding the
molecular basis of acquired resistance to an existing drug has been extremely valu-
able, allowing the discovery of new drugs with the same or similar mechanism of
action but which avoid resistance.

18.3.3 The Old Targets Gave Us Confidence in New Targets

The success of antibiotics with these old targets has given us confidence that we
know how antibiotics work, how to use them, and how to find new ones. Antibiotics
are probably the most successful and best understood type of drug. Broadly speak-
ing, all antibiotics have the same, extremely simple mechanism: stop growth of the
infecting organism and let the patient recover. Bacterial infections can be consid-
ered as either the invasion of the patient by a foreign pathogen or the intrusion of a
commensal organism into a normally sterile site. In either scenario, the vast major-
ity of such invasions are easily controlled by the patient’s inflammatory system
without any problem. If the bacteria somehow get away from the normal host
defenses, the result is symptomatic disease. A drug that slows growth of the invader
will often allow the patient to regain the upper hand and clear the infection. Although
it might seem desirable for antibiotics to kill bacteria outright, in actuality many
successful antibiotics are simply bacteriostatic.
18 The New Versus Old Target Debate for Drug Discovery 569

Study of the antibiotics discovered during the twentieth century led to standard
in vitro methods for determining the antimicrobial susceptibility of bacteria isolated
from patients and to understanding how antibacterial activity and pharmacokinetics
of an antibiotic contribute to its efficacy in an animal model. The developing field
of PK/PD not only allows accurate prediction of effective dosing in animal studies
but allows extension of those data to human trials.
For most existing antibiotics, we have a fairly good understanding of mechanism
of action, usually inhibition of an essential bacterial process. Collectively, the study
of existing antibiotics has given us the feeling that a chemical compound will pre-
vent bacterial growth if it is able to enter bacterial cells and inhibit an essential
bacterial process with sufficient potency. If in vitro antibacterial activity is suffi-
cient, the compound will probably have efficacy in mouse models of infection as
long as it is not toxic and has appropriate pharmacokinetics to provide sufficient
exposure to bacteria at the site of infection.
These features should make antibiotics far easier to discover and develop than
drugs for other therapeutic areas. Consider drugs for hypertension, cardiac disease,
Alzheimer’s, rheumatoid arthritis, etc. Discovery of a new drug often means simul-
taneously developing new understanding of the molecular basis of the disease and
finding compounds that affect that aspect of human biology while leaving the rest of
human physiology intact. For many human diseases, animal models are not very
useful for testing efficacy of new drugs. If a new drug is approved, it is necessary to
teach physicians which patients will benefit and how to use the new drug. In con-
trast, the principles of antibiotics are well understood by scientists, by physicians,
and by regulatory agencies. We know how to evaluate candidate drugs in the lab and
in patients, and physicians understand how to use the new drug if it is approved.

18.4 Seeking Antibiotics with New Targets: The Genome Era

At the turn of the century, it was recognized that the so-called Golden Age of
Antibiotics had come to an end. Although new members of existing antibiotic
classes were being developed, the only new class that had been discovered for
decades was the oxazolidinones. Discovery of promising natural products had
stalled. The search for new targets was driven by a feeling that as antibiotic resis-
tance was continuing to increase, new experimental approaches would be required.
The explosion of target-directed projects in the early twenty-first century, par-
ticularly the widespread use of in vitro enzyme assays for high-throughput screen-
ing of synthetic libraries, is often described as the “genome era” of antibiotic
research. In addition to the bacterial genome sequences that were appearing,
­industrial research programs at this time made use of other components of target-
directed drug discovery that were already in place in biotechnology and pharmaceu-
tical companies, having been developed for other therapeutic areas. There was
increasing interest in enzymes as targets, as it would be possible to screen large
chemical libraries with in vitro assays using robotics. If the enzyme could be crys-
570 A. L. Erwin

tallized, structural data would be available to guide chemistry. Rational drug design
seemed feasible and more attractive than the empiric methods by which most anti-
biotic classes had previously been developed.
The critical value of the genome sequences available by 2000 or so was not
the thousands of previously unknown genes but the incredible level of detail on
the genes for which functions were known or could be proposed. A gene previ-
ously known only in E. coli or Bacillus subtilis could be amplified from P. aeru-
ginosa or S. aureus in order to produce recombinant protein. Transposon
mutagenesis and other molecular biology techniques allowed genome-wide
assessment of genes required for in vitro growth or for full virulence in animals.
Research programs focused on targeting enzymes for which in vitro assays could
easily be developed would compile a list of 100–150 potential targets that were
considered to be essential in vitro, absent from mammalian cells, and conserved
across the pathogens of interest. A complementary approach used antisense tech-
nology to downregulate individual genes, generating strains that were hypersen-
sitive to inhibitors of the target enzyme (or pathway). Such strains could be used
in initial screening or in evaluating compounds that were known to be potent
inhibitors of the target [24, 59, 63].
The use of new, genome-wide studies together with robotics and combinational
chemistry seemed very exciting at the time but did not lead to the rapid discovery of
new antibiotics, as had been hoped. Some of the limitations of these programs will
be discussed in a later section.

18.5 New Antibiotics with Other, Well-Defined Targets

Considering antibacterial compounds with novel mechanisms, it is remarkable that

very few were discovered by choosing an enzyme target and screening for inhibi-
tors, though this might be considered by some to be the standard approach. Indeed,
no antibiotic ever approved was discovered by this approach. This section describes
several of the most promising target-directed projects. Not all were initiated by
screening for inhibitors of a preselected target. In most cases the target was identi-
fied early in the project, and potency toward the target was used in medicinal chem-
istry. For most of the targets listed below, at least one compound has reached clinical
trials; all have inhibitors with reported efficacy in animal models of infection.

18.5.1  rotein Synthesis: Targets Other than the 30S and 50S
P
Ribosomal Subunits

Several inhibitors of protein synthesis with mechanisms different from those dis-
cussed above have reached clinical trials.
18 The New Versus Old Target Debate for Drug Discovery 571

EF-Tu The thiopeptide LFF571 (Novartis) is a semisynthetic derivative of a natu-

ral product, GE2270 A, that binds elongation factor Tu. LFF571 was reported to be
effective in treating C. difficile infection, based on a phase 2 study [65].

tRNA-synthetases A series of boronate compounds with antimicrobial activity

was found to inhibit leucine-tRNA synthetase, leading to discovery of AN3365
(later GSK2251052) at Anacor. This compound appeared very promising because of
its Gram-negative activity. However, a phase 2 study for complicated urinary tract
infection was halted when resistant mutants were isolated from several patients
[70]. A chemical series of methionine-tRNA synthetase inhibitors discovered at
GlaxoSmithKline was licensed to Replidyne [40], who discovered the diaryldiamine
series, including REP3123 (now CRS3123), being developed by Crestone for C.
difficile infection [67].

Peptide deformylase (PDF) PDF was a very popular target because it was thought
there is no equivalent mammalian enzyme and that initiation of bacterial proteins
with formyl-methionine was a universal feature of bacterial proteins. An antibacte-
rial natural product, actinonin, was found to inhibit PDF, providing chemical valida-
tion. However, actinonin-resistant mutants of S. aureus were easily isolated and
found to have inactivated fmt, the gene encoding formyl-methionine transferase.
Finding that fmt-null mutants can be viable suggested that PDF inhibitors would be
effective only for species in which fmt is essential. These include Streptococcus
pneumoniae and Haemophilus influenzae. There is also some evidence that fmt-null
mutants of S. aureus are reduced in virulence. PDF is a metalloenzyme, so it was
expected that finding potent inhibitors would be straightforward. Indeed, at least
three PDF inhibitors reached clinical trials [15, 48, 110]. Two of these, BB-83698
and LBM-415, were discontinued after phase 1. GSK1322322 entered phase 2 and
has since been discontinued.

18.5.2 Lipid Synthesis

FabI The fatty acid biosynthetic enzyme FabI (enoyl-ACP reductase) has been
targeted in several programs. This enzyme had been expected to be conserved across
bacterial species. However, the availability of genome sequences made it apparent
that some species contain alternative enzymes, designated FabK, FabL, or
FabV. These catalyze the same reaction as FabI but are so different in structure that
inhibition of all by a single compound appears impossible. Moreover, some patho-
gens are able to use exogenous fatty acids during infection and are therefore
expected to be intrinsically resistant to inhibitors of endogenous fatty acid synthesis
[108]. Two FabI inhibitors are currently in clinical trials, CG400549 from
CrystalGenomics and Debio 1450 from Debiopharm [12].
572 A. L. Erwin

LpxC Lipopolysaccharides (LPS) are specific to Gram-negative bacteria. The lipid

A moiety is an essential component of LPS and is the most conserved part of the
molecule. In principle, most of the enzymes involved in lipid A synthesis could be
considered as potential antibiotic targets. In practice, only LpxC (UDP-3-O-(R-3-­
hydroxymyristoyl)-N-acetylglucosamine deacetylase) has been pursued seriously.
Nearly all LpxC programs can be traced back to a chemical series discovered at
Merck in the 1980s using a bacterial cell-based screen for inhibitors of any step in
the LPS biosynthetic pathway. Activity of the Merck series was limited to E. coli
and other enterics. A program at Chiron and the University of Washington discov-
ered analogs with activity for both E. coli and P. aeruginosa. This led to work at
many companies, and over a dozen institutions have filed patents on LpxC inhibi-
tors. All of these compounds are hydroxamic acid derivatives, and potency depends
on chelating zinc in the LpxC active site. Only one of these has so far reached the
clinic: ACHN-975 from Achaogen was discontinued after phase 1. Forge
Therapeutics has recently described efficacy of a series of non-hydroxamate LpxC
inhibitors, though structures have not yet been disclosed [27].

18.5.3 Antifolates

The antifolates are the oldest synthetic antibiotics still in use today. Sulfonamides
were discovered empirically in the 1930s by Gerhard Domagk, with antibacterial
activity determinations used to drive medicinal chemistry of compounds derived
from azo dyes. After introduction of the first sulfa drugs, it was realized that they
mimic para-aminobenzoic acid (PABA), thereby inhibiting dihydropterate synthase
(DHPS) [92]. Trimethoprim was discovered more rationally. As described by van
Miert, George Hitchings hypothesized that analogs of purine and pyrimidine bases
might serve as antimetabolite drugs. He discovered several 2,4-diaminopyrimidine
inhibitors of the folate pathway that differed in specificity, including trimethoprim
for bacteria, methotrexate for cancer, and pyrimethamine for protozoal infections
(e.g., malaria); these are all inhibitors of dihydrofolate reductase (DHFR) [102].
The DHFR inhibitor iclaprim was designed to address trimethoprim resistance.
Development was stalled in 2008, but it is again in phase 3 clinical trials, sponsored
by Motif Bio [12]. Both DHPS and DHFR are being pursued in academic research
labs, but no clinical candidates have yet been identified [28, 35].

18.5.4 RNA Polymerase

Discovered in the 1950s, rifamycins are the major antibiotic class targeting RNA
polymerase. Fidaxomicin, another inhibitor of RNA polymerase, is a natural prod-
uct described in 1975 that was approved in 2011 for C. difficile infection. A limita-
tion of both these classes is their very high frequency of resistance due to mutations
18 The New Versus Old Target Debate for Drug Discovery 573

in rpoB. Rifamycins are used primarily in combination with other antibiotics and
occasionally as monotherapy in situations where the number of bacteria to be treated
is very low, such as for prophylaxis. A recent report described the discovery of pseu-
douridimycin by screening a library of natural product extracts for inhibitors of
RNA polymerase. Pseudouridimycin is active against several Gram-positive species
and fastidious Gram-negative bacteria and was reported to be efficacious in a mouse
model of Streptococcus pyogenes infection. Of note, the frequency of resistance to
pseudouridimycin was reported to be tenfold lower than that of rifampin, apparently
because it has a different binding site that is less tolerant of mutations [47, 52, 78].

18.6 Did the New-Targets Programs Fail?

One of the risks of any new target is a high likelihood of failure, but the same can be
said of any drug discovery effort. Even with well-established antibiotic classes, it is
rare for a new clinical candidate to emerge, and the time between discovery and
approval is many years. It is thus not surprising that few of the antibiotics approved
in the past 5 years came from projects that were initiated after 2000 or so. However,
it is surprising that there are so few scaffolds in the pipeline from the target-directed
projects started in the genome era.
An influential review from scientists at GlaxoSmithKline described the dismal
failure of 67 high-throughput screens of bacterial targets, with very few producing
hits that were worth exploratory chemistry. A more recent review from the antibac-
terial program at AstraZeneca was somewhat more positive, in that they had identi-
fied hit scaffolds for 57 of the 65 targets they screened. Lead series were identified
for 19 targets. It seems likely that differences in the synthetic libraries at the two
pharmaceutical companies accounted for the differing hit rates. Despite initial
promising results, the AZ team dropped nearly all their targets when they were
unable to find compounds with broad-spectrum antibacterial activity. A contributing
factor was that at the time, Gram-negative activity was seen as essential. If they had
been primarily interested in drugs for staphylococcal and enterococcal infections,
they might have considered a smaller proportion of projects to be failures. Both
reviews emphasized, as others have, that pursuing such high numbers of targets
might have contributed to failure of the overall programs [74, 99].
The prevailing culture held that all genetically validated targets were equally
good and that screening would identify those that were worth pursuing. There was
high pressure to run as many screens as possible, leading to an emphasis on targets
for which assays would be easy to develop. In practice, these were nearly all
enzymes. Knowledge of the biochemical reaction and structure of the active site
aided prioritization of “druggable” targets. Despite the idea of pursuing novel tar-
gets identified in genome sequences, most work was on genes whose function had
already been identified in the previous decades of academic research. Protein–pro-
tein interactions were generally not considered druggable.
574 A. L. Erwin

Hit evaluation and early chemistry focused on improving affinity for the tar-
get, sometimes without evidence that any antibacterial activity observed was
mechanism-­based. The mantra “fail early and often” meant that one project after
another was terminated without a chance to learn how to improve the chances of
success.
A common experience is that biochemical and structural data could allow
the chemists to improve potency of a scaffold substantially but without a cor-
responding improvement in antibacterial activity. This is in large part because
we lack information on how to improve the accumulation of compounds within
bacterial cells – an important subject beyond the scope of this review. It is
often suggested that chemical libraries in most companies are not well suited
for antibacterial discovery and that we would be more successful with different
starting points.
Because it is difficult to confer antibacterial activity on chemical scaffolds that
lack it, some programs have run target-focused screens that use bacterial cells. A
number of approaches using reporter genes, differential growth, morphology, or
labeling by a dye or radioisotope have been described [63, 90]. Even in programs
that placed a strong emphasis on cell-based assays rather than in vitro enzyme
assays, the overall success rate has been low, considered in terms of drugs entering
clinical trials.
For some targets, we have some information about why development of a spe-
cific compound was halted. Inhibitors of two metalloenzymes (PDF and LpxC)
have failed to progress after phase 1, and we can guess that these compounds had
toxicity that might have been mechanism-based. Whether toxicity will keep other
LpxC inhibitors from reaching the clinic is difficult to determine. The rapid selec-
tion of mutants resistant to the LeuRS inhibitor GSK2251052 is consistent with our
general understanding of antibiotics with single targets. However, preclinical data
suggest that resistance to the MetRS inhibitor CRS3123 may be less of an issue
[17]. Perhaps surprisingly, the frequency of resistance to LpxC inhibitors is so far
very low [27].
For projects that never reach the clinic, there is rarely any public information
about the status of any project or the reason it was terminated. It is therefore
impossible to determine why any target has not yielded new drugs, or even to say
conclusively that it has failed. A scaffold may be active vs a few isolates of S.
aureus or E. coli and perhaps show efficacy in a mouse model but never reach the
desired antibacterial spectrum; or it may have poor pharmacokinetics or show
toxicity in preclinical studies. Most of these liabilities are characteristics of the
compound or the chemical scaffold but shed no light on the validity of the
molecular target. Often resources are diverted from one project to another for
reasons that are strategic rather than scientific. Should we conclude in such a
case that the target has now been validated, not only genetically but chemically
and pharmacologically? If no clinical candidate appears, should we conclude the
target is a failure?
18 The New Versus Old Target Debate for Drug Discovery 575

18.7 Limitations to Classical Target-Directed Discovery

Today there is increasing recognition that there are two kinds of unmet need for
bacterial infections.
Traditional antibiotics We need new traditional antibiotics to replace those for
which resistance has become common. The discovery programs described in the
previous sections are intended to meet this need. As has been discussed in numerous
conferences and white papers, new antibiotics are not emerging at the rate that is
required. However, even if we had a full pipeline of broad-spectrum antibiotics for
Gram-positive and Gram-negative infections, there would still be a need for other
types of drug.
One issue is that the desired target-product profiles change over time. At the turn
of the century, a common criticism of LpxC projects was that the resulting drugs
would not be useful for treating MRSA. It was difficult to convince senior manage-
ment that Gram-negative resistance was on the rise. Infectious disease physicians
generally did see value in a new antibiotic that would be limited to Gram-negative
bacteria – as long as it was active against P. aeruginosa as well as enteric bacteria.
Today we need drugs for A. baumannii as well as for P. aeruginosa. At the same
time, antibiotic resistance in Klebsiella and Enterobacter spp. and E. coli has
increased to the point where an enteric-only drug would be valuable. C. difficile
infection, long recognized as a complication of treatment with broad-spectrum anti-
biotics, became a much greater problem as antibiotic-resistant strains became prev-
alent. It has become common for new Gram-positive agents with poor PK and oral
bioavailability to be developed as C. difficile drugs. For some of these, microbiome
studies are incorporated into clinical trials.
The idea of narrow-spectrum antibiotics is becoming much more attractive. It is
thought that resistance to narrow-spectrum drugs may emerge more slowly, as the
normal flora will not provide a reservoir for resistance determinants. Further,
narrow-­spectrum antibiotics are not expected to induce the dysbiosis that is associ-
ated with broad-spectrum antibiotics. The development and use of narrow-spectrum
antibiotics are far from straightforward but may become more feasible in the next
several years [9]. It is therefore reasonable to reconsider some of the targets that
were rejected during the genome era because they were not broadly conserved
across pathogens.
Second unmet need We also need drugs for prevention and treatment of bacterial
infections for which antibiotics have never been fully effective. These include recur-
rent, latent, or persistent infections, often associated with biofilms. Such infections
have plagued humans for millennia but are much more prominent today. Examples
are ventilator-associated pneumonia, infections in immunocompromised patients
with organ transplants or cancer chemotherapy, infected joint replacements or other
orthopedic devices, diabetic foot ulcers, respiratory infections in patients with cys-
tic fibrosis, and recurrent urinary tract infections.
576 A. L. Erwin

The antibiotics we have today act mainly on growing bacteria. They are probably
most effective in treating acute infections in previously healthy individuals. The
same is likely to be true for new antibiotics discovered with either the old targets or
the enzyme targets of the genome era. We need drugs that are active against non-
growing bacteria and against bacteria that have become tolerant to antibiotics.
Some aspects of infection are not affected by antibiotics at all. These include the
direct toxic effects of secreted cytotoxins, some of which are treatable with antibod-
ies, and the uncontrolled inflammatory response to lipopolysaccharides or toxic-­
shock syndrome toxin. Research on sepsis has not yet produced effective drugs, in
part because animal models are not very predictive.
Overall, new antibiotics with standard kinds of targets are likely to have the same
limitations as existing antibiotics. Different targets, or different experimental
approaches, will be needed in order to discover drugs that are not only new but bet-
ter anti-infective therapeutic agents. Some of these are described in the next
section.

18.8 New Drugs with New, Often Complex Mechanisms

A previous section described promising antibacterial compounds that were discov-

ered in target-directed programs, meaning that medicinal chemistry was guided not
only by antibacterial activity but by experimental data on the interaction of com-
pounds with the molecular target. For each of the previously described drugs, the
mechanism of action is direct inhibition of a process that is essential for bacterial
growth.
In contrast, this section describes compounds with a variety of mechanisms, dis-
covered through experimental approaches that do not fit this paradigm. The com-
pounds described below have one or more of the following features: empiric
discovery, nonspecific mechanisms, affecting the target at the level of regulation
rather than enzymatic activity, and no in vitro antibacterial activity.

18.8.1 Empiric Discovery of New Drugs with New Mechanisms

Although most classes of existing antibiotics were discovered empirically, recent

experience with screens for antibacterial activity have been frustrating. Commonly
a high proportion of “hit” compounds are nonspecific, affecting both eukaryotic and
prokaryotic cells. Often these hits are found to be membrane active (either lysing or
depolarizing membranes), behaving like poisons or detergents. Attempts to improve
selectivity are rarely successful, so such compounds are routinely discarded [71, 74,
88, 90]. This does not mean that it is impossible to find drugs by tracking antibacte-
rial activity alone. In the examples below, if medicinal chemistry was used, the
18 The New Versus Old Target Debate for Drug Discovery 577

driver was antibacterial activity rather than potency vs target. At the time these
compounds were discovered, the target was not known (and may still be unknown).
Teixobactin A platform at NovoBiotic for seeking antibacterial compounds pro-
duced by previously uncultured microorganisms has revived hope in empiric natural
products discovery. In 2015, they described a new Gram-positive antibiotic with
efficacy in mouse models of S. aureus and pneumococcal infection. Teixobactin
inhibits cell wall biosynthesis, binding glycolipids lipid II and lipid III (precursors
of peptidoglycan and teichoic acids, respectively). This mechanism of action is
similar to that of vancomycin, which binds a different site on lipid II. Because it is
not the product of a single gene, it is difficult for spontaneous mutation to alter the
structure of lipid II. The frequency of endogenous resistance to vancomycin is thus
extremely low, and the same appears to be true for teixobactin. Resistance to vanco-
mycin became common only after a resistance cassette derived from the natural
producer was transferred into enterococci, encoding enzymes that produced an
altered lipid II. Teixobactin binds to both forms of lipid II and is thus active vs both
vancomycin-sensitive and vancomycin-resistant bacteria. Several other antibacte-
rial compounds that bind lipid II are known, including the peptides plectasin and
nisin. Apart from the vancomycin derivatives telavancin, dalbavancin, and orita-
vancin, none of these other lipid II-binding compounds is used clinically [49, 81].

Bedaquiline This review does not otherwise discuss tuberculosis drugs, but two
classes of antimycobacterial antibiotics are mentioned here as successful recent
examples of using antibacterial activity to guide medicinal chemistry, without
knowledge of the molecular target. The recently approved tuberculosis drug beda-
quiline targets ATP synthase and is bactericidal for both growing and nongrowing
mycobacteria. Bedaquiline was discovered at Janssen by screening for compounds
that inhibit growth of Mycobacterium smegmatis and then optimizing by medicinal
chemistry [1]. It is of interest because of the success of this empiric approach and
also because of the unprecedented nature of the target. No other antibacterial drug
has a well-defined direct effect on the respiratory pathway.

PA-824 and OPC-67683 A second class of new mycobacterial drugs is the

2-­nitroimidazoles, PA-824 (now pretomanid, discovered at PathoGenesis and in
development by the TB Alliance) and OPC-67683 (now delamanid, discovered at
Otsuka and approved in Europe and Japan in 2014). The scaffold is not entirely new,
being related to previously known nitroimidazoles. Metronidazole, a 5-­nitroimidazole
derived from the natural product azomycin, was first used for anaerobic protozoal
infections (e.g., trichomoniasis) and then realized to be active against anaerobic
bacteria. Metronidazole is a prodrug, converted within the microbial cell to produce
a nitro radical anion and a variety of reactive nitrogen intermediates that kill the cell
by damaging cellular components, particularly DNA. In the 1990s, metronidazole
was recognized to be active against dormant or anaerobically adapted M. tuberculo-
sis but not against actively growing mycobacteria. 2-imidazoles were recognized in
the early 1970s to have modest activity against a variety of bacteria. During discov-
578 A. L. Erwin

ery of PA-824, scientists at PathoGenesis monitored antimycobacterial activity of

compounds, resulting in development of the nitroimidazopyran scaffold. PA-824 is
active against both actively growing and nonreplicating mycobacteria and has little
activity against other bacterial species. Determining potency against the target was
not part of compound evaluation, and indeed the mechanism of action is complex
and not well understood. Both PA-824 and OPC-67683 are like previous nitroimid-
azoles in requiring activation within bacterial cells. Activation of PA-824 is thought
to produce reactive nitrogen species, similar to metronidazole. However, both
PA-824 and OPC-67683 also appear to have a more specific mechanism, affecting
synthesis of mycolic acids [64].

Ridinilazole (SMT 19969) A final example of discovering a drug with a novel

mechanism by chemical optimization of antibacterial activity is the bis-­
benzimidazole SMT-19969. The bis-benzimidazole scaffold was initially designed
to target specific DNA sequences, by binding in the minor groove of duplex DNA. A
subset of compounds were found to possess antibacterial activity, apparently medi-
ated by gyrase inhibition. Optimization for C. difficile activity at Summit
Therapeutics led to SMT-19969, which lacks activity against other bacterial species,
does not inhibit gyrase, and does not bind duplex DNA. Ridinilazole is currently in
phase 2 for treatment of C. difficile infection, and its target is not known [55].

18.8.2  mpiric Discovery of Compounds That Bind

E
Membranes or DNA

As noted above, disruption of bacterial membranes is not considered a desirable

mechanism of action, because it so often is associated with toxicity for eukaryotic
cells. However, if it is possible to make the compounds highly specific for bacteria,
then membrane activity has some attractive features. The frequency of resistance is
usually very low. Also, membrane-active compounds act rapidly and are active vs
both nongrowing and growing bacteria.
Brilacidin (formerly PMX-30063) Brilacidin is a non-peptide compound
designed to mimic human host-defense peptides. While antimicrobial peptides have
so far been limited to topical uses because of their toxicity and poor pharmacokinet-
ics, brilacidin is being developed as a systemic drug [83]. It is currently in a phase
2 study of acute bacterial skin and skin structure infection, sponsored by Cellceutix.

POL-7080 POL-7080, discovered at Polyphor and now in phase 2, is also a pepti-

domimetic, from a scaffold initially designed to mimic the host-defense peptide
protegrin I. Remarkably, during optimization for P. aeruginosa activity, the series
lost activity against other bacterial species. Its P. aeruginosa activity involves bind-
ing to the outer membrane protein LptD, and it is thought that its mechanism of
action is inhibition of the LPS export system, Lpt [93].
18 The New Versus Old Target Debate for Drug Discovery 579

Phage lysins The idea of using lytic peptides derived from phage or bacteria as
antibacterial drugs is not new. A key issue is finding lysins with appropriate speci-
ficity. Many such peptides are active against only certain isolates of a bacterial spe-
cies; others, like nisin and gramicidin, are so broad in spectrum that they are
generally toxic. Two phage lysins are being developed for staphylococcal infection.
N-Rephasin (SAL200) from iNtRON is in phase 2 studies. It was derived from bac-
teriophage SAP-1 and has broad activity against staphylococci but is not active
against other genera [42]. CF-301 (previously PlySs2) from Contrafect is currently
in phase 1. CF-301 was derived from a Streptococcus suis phage and is broadly
cidal for staphylococci and streptococci, including biofilms [82].

Pheromonicins An intriguing approach for generating lytic peptides of desired

specificity was proposed by Qiu et al. They fused a staphylococcal AgrD1 phero-
mone to the channel-forming domain of colicin Ia; the resulting peptide (“phero-
monicin”) had antibacterial activity toward strains of S. aureus able to recognize
that pheromone and was able to protect mice from S. aureus infection. A fusion
protein containing a scrambled pheromone was inactive [75].

Minor groove binders Drugs that bind the minor groove of DNA, such as the
antiparasitic drug pentamidine, have a long history. While there is some specificity
in the binding sites, these are not generally designed to affect transcription of a
particular gene (as for the antisense approach described below). Analogs are evalu-
ated by their antimicrobial activity and by selectivity for the desired target organ-
isms. Minor groove binders have the advantage that they kill rapidly and have low
frequency of resistance. However, the nonspecific nature of their activity makes
toxicity a concern. MGB-BP-3, currently in phase 1, is derived from the natural
product distamycin and is being developed for C. difficile infection. A series of bis-­
indole compounds with activity in mouse models of Gram-positive infection dis-
covered at Microbiotix was recently reported to have a mechanism that involves
DNA binding [5, 72, 95].

18.8.3  arget-Specific, with a Mechanism at the Level

T
of Transcription or Translation Rather than Binding
the Protein Target

Antisense The idea of using antisense molecules to prevent synthesis of a target

protein is attractive. Oligonucleotides with high affinity for a specific RNA sequence
can be designed, manufactured, and tested much more easily than small-molecule
enzyme inhibitors. Specificity for a single bacterial species is feasible. A potential
problem is the likely high frequency of resistance due to point mutations that affect
binding of the oligonucleotide without impairing fitness. Instability in serum is a
problem for RNA drugs that has been addressed using mimics such as phosphoro-
580 A. L. Erwin

diamidate morpholino oligomers (PMOs). Conjugation of peptides to PMOs can

improve their uptake by bacteria. The Greenberg group at University of Texas
Southwestern has described peptide-conjugated PMOs for several targets in a vari-
ety of bacterial species and in some cases achieved efficacy in mice [38].

Riboswitches A different approach to inhibiting production of a target is using

small molecules to bind riboswitches in mRNA. The Breaker lab at Yale first
described riboswitches several years ago. Breaker and colleagues recently described
riboflavin analogs that bind the FMN riboswitch and are active against C. difficile
both in vitro and in vivo. Researchers at Merck described a small molecule, ribocil-
­C, with antibacterial activity for several Gram-positive and Gram-negative species.
Like the natural product roseoflavin, ribocil-C acts by binding FMN riboswitches.
A recent paper from Merck reported that in S. aureus, both these compounds are
dual binders of two FMN riboswitches, one controlling riboflavin synthesis and the
other riboflavin uptake [8, 104].

18.8.4 Drugs Lacking Antibacterial Activity In Vitro

The idea that studying host–pathogen interaction will lead to discovery of improved
therapeutic agents is intellectually attractive but has not yet been very successful
with regard to standard antibiotics like those discussed above. In contrast, nearly all
the agents discussed below resulted from research on how bacteria cause disease or
respond to therapy.

18.8.4.1  gent Increases Vulnerability of Bacteria to Antibiotics and/or

A
to Host Defenses

Surface-binding mAbs Several monoclonal antibodies that bind to bacterial cell

surface antigens of P. aeruginosa or S. aureus are currently in clinical trials.
Antibodies are today considered as alternative or nontraditional approaches because
they do not kill bacteria directly, but they are far from new. Administration of
pathogen-­specific antibodies was used successfully well over a century ago for
treatment of pneumococcal pneumonia, meningococcal meningitis, and many other
bacterial infections [13].
Three monoclonal antibodies to surface antigens of P. aeruginosa are in phase 2
clinical studies. MEDI3902 from MedImmune is a bispecific antibody that binds
the exopolysaccharide Psl and also PcrV, the needle of the Type III secretion system
[22]. It is being tested for prevention of hospital-acquired or ventilator-associated
pneumonia (HAP/VAP) in high-risk patients colonized with P. aeruginosa. Aridis
has two monoclonal antibodies being tested as adjuncts to antibiotic therapy in
18 The New Versus Old Target Debate for Drug Discovery 581

patients with P. aeruginosa HAP or VAP. Aerucin binds alginate, and Aerumab
(previously AR-101 or Panobacumab) binds LPS of serotype O11 [76].
Two antibody-based agents are being studied in patients with S. aureus bactere-
mia: monoclonal antibody 514G3 from XBiotech and an antibody-antibiotic conju-
gate (DSTA4637S, RG7861) from Genentech.
Antibiotic potentiators Given the difficulty in discovering new antibiotics, an
obvious potential alternative is to find drugs that can either restore the susceptibility
of resistant strains or extend the spectrum of Gram-positive antibiotics to Gram-­
negative bacteria. Apart from the β-lactamase inhibitors discussed in a previous
section, no such drugs have reached the market. The only clinical candidate in this
category is SPR741.
SPR741(previously NAB741) is a polymyxin derivative being developed by
Spero Therapeutics as an antibiotic-potentiating agent, currently in phase 1 [101].
Because its mechanism is permeabilization of the outer membrane, it is possible
that SPR741 might also increase the susceptibility of bacteria to host defenses such
as defensins and complement.
Many academic and industrial labs have screened for compounds that increase
the activity of β-lactams for MRSA. Where mechanisms of active compounds have
been identified, they have been surprisingly diverse. A more systematic approach
has been taken by Merck scientists, seeking inhibitors of wall teichoic acid synthe-
sis [39, 66, 98, 107].
Efflux pump inhibition is another area that has been extremely frustrating,
despite advances in the assembly and function of multidrug efflux pumps and in the
interaction of pump components with their substrates [36, 51, 53, 91, 96].

18.8.4.2 Agent Counteracts a Defined Virulence Mechanism

Targeting pathogenesis is often suggested but rarely has been very successful. One
issue is that potential targets are identified by demonstrating the inability of mutants
to initiate infection. It is by no means certain that drugs inhibiting the production or
function of those virulence determinants would be able to reverse an established
function. In addition to these concerns, an additional argument against anti-viru-
lence drugs is that they would be narrow in spectrum. That is now seen as less of a
disadvantage than it was several years ago. It must be admitted that there has been
very little serious effort by the pharmaceutical industry to discover anti-­virulence
drugs.
Antitoxin mAbs Like the surface-binding antibodies discussed in a previous sec-
tion, passively administered toxin-neutralizing antibodies have an extremely long
history of successful use. Shigamab, being developed by Taro, consists of two anti-
bodies against Shiga toxin [60]. ASN100 from Arsanis is also a mixture of two
antibodies, one active against five cytotoxins and the other active against LukGH
leukocidin [2, 21]. ASN100 is being tested as a prophylactic agent in patients colo-
582 A. L. Erwin

nized with S. aureus and at high risk of HAP/VAP, as is MEDI4893, an α-toxin-­

neutralizing mAb from MedImmune. A second mAb that neutralizes staphylococcal
α-toxin is AR-301 (Salvecin) from Aridis, being tested for therapeutic use in HAP
and VAP, in combination with antibiotics [31].

MfvR One of the difficulties in targeting virulence determinants is that many

pathogens have several mechanisms of virulence. With some exceptions (such as
certain secreted toxins and major surface components like pneumococcal capsular
polysaccharides), inactivation of a single virulence factor has relatively little effect
on infection. In P. aeruginosa, several virulence factors are controlled by a single
regulatory protein MvfR. In one of the most promising examples of an anti-­virulence
approach, the Rahme lab at Harvard identified compounds binding to MvfR that are
effective in treating infected mice [57, 94].

18.8.4.3 Immunomodulatory Agents

Some aspects of infectious disease result not from the bacteria per se but from an
overwhelming inflammatory response to bacterial components. This can persist
even if the infection is treated with appropriate antibiotics. Attempts to rescue septic
patients with interleukins or other immunomodulators have been largely unsuccess-
ful, even though some such agents appeared to be effective in mice. Two immuno-
modulatory agents are now in clinical trials.
AB103 from Atox Bio, now in phase 3, is an octapeptide that attenuates the sig-
naling through CD28 that is involved in induction of many proinflammatory cyto-
kines [77]. It is being tested in patients with necrotizing soft tissue infections.
CAL02, now in phase 2 from Combioxin, is a liposomal drug that neutralizes
bacterial toxins. It is being tested in addition to standard of care in intensive care
unit patients with severe pneumococcal pneumonia [6].
The Kranz lab, University of Illinois, engineered T-cell receptor domains to bind
staphylococcal enterotoxins B and C with high affinity and reported efficacy in rab-
bit models of necrotizing pneumonia and infective endocarditis [56, 86].

18.9  ow Well Do We Understand Antibiotics

H
and Resistance?

The previous section described a number of experimental approaches that have the
possibility of improving treatment of infectious disease. Several of the antibacterial
agents are bactericidal for nongrowing bacteria, making it possible that they will be
more effective against persistent infection than current antibiotics are. Many of
them were discovered by empiric methods and/or have mechanisms that would not
have been considered appropriate for target-directed drug discovery. With the
18 The New Versus Old Target Debate for Drug Discovery 583

exception of the tuberculosis drugs, these are all very early in development. It can-
not be concluded that these nonclassical approaches will always be more successful
than classical target-directed antibiotic discovery.
The target-directed programs of the genome era focused on enzymes. As dis-
cussed above, there are a number of reasons why any given project might have
failed. There is no obvious reason to think that enzymes as a group should be less
suitable as antibiotic targets than the macromolecular synthesis machines that most
existing antibiotics inhibit. Indeed, the first broad-spectrum synthetic antibiotics,
the antifolates, are enzyme inhibitors. It is puzzling that even for a pathway as well
validated as peptidoglycan synthesis, it has been impossible to find good inhibitors
of the soluble enzymes (GlmU, MurA-MurF) apart from fosfomycin. One limita-
tion of enzymes as targets is the risk that mutants with endogenous resistance occur
at high frequency [87]. This risk is not simply theoretical, as indicated by the failure
of GSK2251052 during phase 2 [70]. It is unlikely, however, that high frequency of
resistance was the characteristic that killed most of the genome-era projects before
a clinical candidate was identified.
It is more likely that after initial identification of a hit series with mechanism-­
based antibacterial activity, most projects progressed for a while but failed to come
close to a solid lead within an allotted period of time. A common experience is that
chemistry can produce very potent inhibitors, as assessed in an in vitro biochemical
assay, but that potency does not translate into antibacterial activity. The standard
explanation is that these compounds “don't get in.” This is certainly true to a great
extent. Compounds active vs staphylococci and streptococci but not vs Gram-­
negative bacteria are often active against an efflux-deficient or hyperpermeable
mutant of E. coli or P. aeruginosa. In that case, it is appropriate to conclude that the
intrinsic defenses of Gram-negative bacteria are a major limitation to antibacterial
activity. Poor antibacterial activity against Gram-positive bacteria is harder to
explain, although again failure to cross the cytoplasmic membrane and accumulate
against the concentration gradient probably contributes to the problem [89].
We must consider the possibility that we simply don’t know enough about bacte-
ria to choose good targets. The assumption of the new-targets programs of the
genome era was that we understand antibiotics and infectious disease well enough
to be able to do this rationally. Discovery programs at that time were often driven
not by new insight into how antibiotics work but by new technology, combined with
a feeling that no new insight was needed. I would argue that not all essential genes
are equally good targets. Moreover, research into the mechanisms by which bacteria
avoid the effects of antibiotics has suggested that drugs with apparently similar
mechanisms can have very different effects on bacterial physiology.
One potential problem is that inhibition of the target may have relatively little
immediate effect on bacterial physiology. It is very difficult to determine “how
essential” a bacterial process is. Will growth stop if catalytic activity is reduced by
25%? Will it be necessary to inhibit 99.9% of activity? Very few studies have
attempted to address this, and indeed there are no general methods. Tuberculosis
researchers have developed methods for targeted degradation of specific proteins in
order to assess the impact on growth and survival of the cell [43, 105]. Extending
584 A. L. Erwin

these methods beyond mycobacteria would be useful, although reducing the amount
of a protein within the cell may have a different effect from chemical inhibition of
that protein’s activity. For a few enzymes, mutants with partial activity can give us
some information as to the level of inhibition that is tolerated. The envA1 mutant of
E. coli has an 18-fold reduction in LpxC activity, compared to wild-type strains
[109]. Similarly, the E. coli mutant ligA251 has a point mutation that reduces activ-
ity of DNA ligase by 20- to 60-fold [46].
Systems biology may allow improved prioritization of targets, with the caveat
that we may not yet have the information needed to generate predictive models. One
example is the modeling of the lipid A synthetic pathway. A combination of compu-
tational and experimental methods led to the suggestion that LpxK would be a better
target for inhibition than LpxC if previous knowledge of the pathway’s regulation is
incorporated into the model. If regulation was not considered, then LpxC appeared
to be a better target, as suggested by previous enzymology [26].
While target-directed discovery programs tend to focus on the effect of chemical
compounds on the target, it may be more useful to focus on the effect of compounds
on the bacterial cell. The most obvious such effects are bacterial stasis or death,
morphological changes such as filamentation or spheroplasting, or changes in pro-
cesses like macromolecular synthesis that can be monitored easily. Bacterial
responses to differing antibiotic stresses may also be important. It may be useful to
look for targets that, when inhibited, induce the same stress responses as one or
more of the well-established existing classes of antibiotics. The SOS response to
DNA damage is one example. Fluoroquinolines induce the SOS response, and it is
also involved in the thymineless death induced by trimethoprim [30].
Reporters of the cell wall stress response have been used by many groups to
screen chemical libraries, finding hits with diverse targets in envelope biogenesis –
not only peptidoglycan synthesis but also LPS synthesis and lipoprotein export [19,
68, 100]. The machines involved in export of proteins and lipopolysaccharide to the
outer membrane and in maintenance of its permeability barrier could be considered
as the only truly novel bacterial pathways that have been discovered in recent
decades [58, 79, 80].
The stringent response was first studied decades ago in the context of amino acid
starvation. Much more recently it was identified as a critical issue in antibiotic toler-
ance and biofilm formation. A better understanding of persistence and tolerance
may allow us to design more effective antibiotics. One might expect all ­bacteriostatic
inhibitors of protein synthesis to induce similar responses in bacteria. However, a
recent study of the effect of bacteriostatic agents on induction of tolerance to
β-lactams found that the Met-tRNA synthetase inhibitor mupiricin activates RelA,
while the ribosome binders tetracycline and chloramphenicol inhibit induction of
the stringent response [45, 54, 85].
Finally, several lines of evidence suggest that the mechanisms of existing antibi-
otics are more complicated than the simple picture presented above. In 2007, the
Collins lab at Boston University proposed that the bactericidal activities of β-lactams,
aminoglycosides, and fluoroquinolones have a common mechanism involving pro-
duction of hydroxyl radicals, as a consequence of the immediate effects of these
18 The New Versus Old Target Debate for Drug Discovery 585

antibiotics on their respective molecular targets [44]. This theory is controversial,

with not all researchers agreeing with the details of the “common death” pathway
[50] (see also Chap. 20). The effects of sublethal concentrations of antibiotics have
suggested that halting cell growth has effects that differ from one class of antibiotic
to another and are not easily predictable [18, 34]. Similarly, genome-wide studies on
changes in bacterial growth and susceptibility when genes are over- or underex-
pressed have revealed a network of interactions that is not readily explained in terms
of single effects of antibacterial drugs [4, 16]. A school of thought suggests that
polypharmacology (not promiscuity) is a feature of many successful drugs and
should be considered as an advantage [10, 11, 61]. It is possible that in optimizing
the potency of a series of antibacterial compounds for a single target, we are remov-
ing minor activities that contributed to the antibacterial activity initially observed.

18.10 Concluding Remarks

Old targets, benefits and risks The field has been successful in identifying new
antibiotics for the old targets, utilizing current structural biology, and making use of
new understanding of mechanisms and resistance determinants as they emerge.
Even for the old antibiotic classes, new synthetic methods are allowing increased
chemical diversity. However, the side effects and other liabilities of these antibiotic
classes must be considered in developing new members. A further problem is that
inhibitors of the old targets tend to be broad-spectrum antibiotics, with inherent
effects on normal flora. Resistance may be slow to emerge but is likely to spread
rapidly, just as for existing antibiotics. Moreover, new inhibitors of the old targets
are unlikely to be better than the old drugs at treating persistent infections or toxic
effects.

New targets, benefits and risks New targets have the possibility of yielding better
drugs – more diverse in chemical structure and mechanism, possibly narrower in
spectrum, and/or providing improved treatment of conditions not well handled by
existing antibiotics. However, focusing narrowly on targets we think we understand
has not been very successful. We need to take a variety of approaches, allowing use
of technology and incorporating new knowledge as it becomes available.

We don’t know enough about compound uptake One of the major problems in
antibiotic discovery is our inability to design compounds with good access to their
targets. Aminoglycosides and tetracyclines owe their success in part to their ability
to achieve concentrations within a bacterial cell that are higher than external con-
centrations. β-lactams and other inhibitors of late stages of peptidoglycan synthesis
do not need to cross the cytoplasmic membrane, but for Gram-negative bacteria the
outer membrane is a barrier to these drugs. Academic research has been very suc-
cessful in elucidating the structure and synthesis of the outer membrane and of the
multidrug efflux pumps that constitute the intrinsic resistance mechanisms of Gram-­
586 A. L. Erwin

negative bacteria. This knowledge has not yet provided concepts or tools to guide
medicinal chemists in improving the antibacterial activity of potent compounds.

Commitment is necessary In a recent review, Bisacchi and Manchester made the

striking observation that even during the Golden Age of Antibiotics, when many
large pharmaceutical companies were heavily invested in anti-infective research,
discovery of a new class of antibiotics was a rare event [7].

Major Points
• Most target-directed drug discovery projects focus on seeking inhibitors of indi-
vidual enzymes
• No antibiotic that has ever been approved was discovered through this approach
• Only a few target-directed projects have yielded clinical candidates
• A critical problem that is not well appreciated is that not all essential bacterial
functions are equally good drug targets
• A second important problem is that optimization of a chemical scaffold for
potency in a cell-free assay often does not improve its antibacterial activity or
drug characteristics
• The antibacterial drug development pipeline is sparse, but includes small mole-
cules and biologics with a wide variety of mechanisms
• Some of these were discovered empirically, others through rational methods
including structure-based drug design

References

1. Andries K, Verhasselt P, Guillemont J, Göhlmann HWH, Neefs J-M, Winkler H, et al. A

diarylquinoline drug active on the ATP synthase of Mycobacterium tuberculosis. Science.
2005;307(5707):223–7. https://doi.org/10.1126/science.1106753.
2. Badarau A, Rouha H, Malafa S, Battles MB, Walker L, Nielson N, et al. Context mat-
ters: the importance of dimerization-induced conformation of the LukGH leukocidin of
Staphylococcus aureus for the generation of neutralizing antibodies. MAbs. 2016;8(7):1347–
60. https://doi.org/10.1080/19420862.2016.1215791.
3. Barbachyn MR. Oxazolidinone antibacterial agents. In: Dougherty TJ, Pucci MJ, editors.
Antibiotic discovery and development. Boston: Springer US; 2012. p. 271–99.
4. Barker CA, Farha MA, Brown ED. Chemical genomic approaches to study model microbes.
Chem Biol. 2010;17(6):624–32. https://doi.org/10.1016/j.chembiol.2010.05.010.
5. Barrett MP, Gemmell CG, Suckling CJ. Minor groove binders as anti-infective agents.
Pharmacol Ther. 2013;139(1):12–23. https://doi.org/10.1016/j.pharmthera.2013.03.002.
6. Baumgartner D, Aebi S, Grandgirard D, Leib SL, Draeger A, Babiychuk E, et al. Clinical
Streptococcus pneumoniae isolates induce differing CXCL8 responses from human naso-
pharyngeal epithelial cells which are reduced by liposomes. BMC Microbiol. 2016;16:154.
https://doi.org/10.1186/s12866-016-0777-5.
7. Bisacchi GS, Manchester JI. A new-class antibacterial—almost. Lessons in drug discovery
and development: a critical analysis of more than 50 years of effort toward ATPase inhibi-
18 The New Versus Old Target Debate for Drug Discovery 587

tors of DNA gyrase and topoisomerase IV. ACS Infect Dis. 2015;1(1):4–41. https://doi.
org/10.1021/id500013t.
8. Blount KF, Megyola C, Plummer M, Osterman D, O'Connell T, Aristoff P, et al. Novel
riboswitch-­binding flavin analog that protects mice against Clostridium difficile infection
without inhibiting cecal flora. Antimicrob Agents Chemother. 2015;59(9):5736–46. https://
doi.org/10.1128/AAC.01282-15.
9. Boucher HW, Ambrose PG, Chambers HF, Ebright RH, Jezek A, Murray BE, et al. White
paper: developing antimicrobial drugs for resistant pathogens, narrow-spectrum indications,
and unmet needs. J Infect Dis. 2017. https://doi.org/10.1093/infdis/jix211.
10. Bourne CR. Utility of the biosynthetic folate pathway for targets in antimicrobial discovery.
Antibiotics (Basel). 2014;3(1):1–28. https://doi.org/10.3390/antibiotics3010001.
11. Brötz-Oesterhelt H, Brunner NA. How many modes of action should an antibiotic have? Curr
Opin Pharmacol. 2008;8(5):564–73. https://doi.org/10.1016/j.coph.2008.06.008.
12. Bush K, Page MGP. What we may expect from novel antibacterial agents in the pipeline
with respect to resistance and pharmacodynamic principles. J Pharmacokinet Pharmacodyn.
2017;44(2):113–32. https://doi.org/10.1007/s10928-017-9506-4.
13. Casadevall A. The third age of antimicrobial therapy. Clin Infect Dis. 2006;42(10):1414–6.
https://doi.org/10.1086/503431.
14. Coleman K. Diazabicyclooctanes (DBOs): a potent new class of non-β-lactam
β-lactamase inhibitors. Curr Opin Microbiol. 2011;14(5):550–5. https://doi.org/10.1016/j.
mib.2011.07.026.
15. Corey R, Naderer OJ, O‘Riordan WD, Dumont E, Jones LS, Kurtinecz M, et al. Safety, toler-
ability, and efficacy of GSK1322322 in the treatment of acute bacterial skin and skin structure
infections. Antimicrob Agents Chemother. 2014;58(11):6518–27. https://doi.org/10.1128/
AAC.03360-14.
16. Côté J-P, French S, Gehrke SS, MacNair CR, Mangat CS, Bharat A, et al. The genome-wide
interaction network of nutrient stress genes in Escherichia coli. MBio. 2016;7(6):e01714–6.
https://doi.org/10.1128/mBio.01714-16.
17. Critchley IA, Green LS, Young CL, Bullard JM, Evans RJ, Price M, et al. Spectrum of activ-
ity and mode of action of REP3123, a new antibiotic to treat Clostridium difficile infections.
J Antimicrob Chemother. 2009;63(5):954–63. https://doi.org/10.1093/jac/dkp041.
18. Davies J, Spiegelman GB, Yim G. The world of subinhibitory antibiotic concentrations. Curr
Opin Microbiol. 2006;9(5):445–53. https://doi.org/10.1016/j.mib.2006.08.006.
19. DeCenzo M, Kuranda M, Cohen S, Babiak J, Jiang Z-D, Su D, et al. Identification of
compounds that inhibit late steps of peptidoglycan synthesis in bacteria. J Antibiot.
2002;55(3):288–95.
20. Deng Y, Sun C, Hunt DK, Fyfe C, Chen C-L, Grossman TH, et al. Heterocyclyl tetracyclines.
1. 7-Trifluoromethyl-8-pyrrolidinyltetracyclines: potent, broad spectrum antibacterial agents
with enhanced activity against Pseudomonas aeruginosa. J Med Chem. 2017;60(6):2498–
512. https://doi.org/10.1021/acs.jmedchem.6b01903.
21. Diep BA, Le VTM, Visram ZC, Rouha H, Stulik L, Dip EC, et al. Improved protection in
a rabbit model of CA-MRSA necrotizing pneumonia upon neutralization of leukocidins in
addition to alpha-hemolysin. Antimicrob Agents Chemother. 2016:AAC.01213-16. https://
doi.org/10.1128/AAC.01213-16.
22. DiGiandomenico A, Keller AE, Gao C, Rainey GJ, Warrener P, Camara MM, et al. A mul-
tifunctional bispecific antibody protects against Pseudomonas aeruginosa. Sci Transl Med.
2014;6(262):262ra155. https://doi.org/10.1126/scitranslmed.3009655.
23. Domagala JM, Hanna LD, Heifetz CL, Hutt MP, Mich TF, Sanchez JP, et al. New structure-­
activity relationships of the quinolone antibacterials using the target enzyme. The develop-
ment and application of a DNA gyrase assay. J Med Chem. 1986;29(3):394–404.
24. Donald RGK, Skwish S, Forsyth RA, Anderson JW, Zhong T, Burns C, et al. A Staphylococcus
aureus fitness test platform for mechanism-based profiling of antibacterial compounds. Chem
Biol. 2009;16(8):826–36. https://doi.org/10.1016/j.chembiol.2009.07.004.
588 A. L. Erwin

25. Ehmann DE, Jahić H, Ross PL, Gu R-F, Hu J, Kern G, et al. Avibactam is a covalent, revers-
ible, non–β-lactam β-lactamase inhibitor. Proc Natl Acad Sci U S A. 2012;109(29):11663–8.
https://doi.org/10.1073/pnas.1205073109.
26. Emiola A, George J, Andrews SS. A complete pathway model for lipid A biosynthesis in
Escherichia coli. PLoS One. 2014. https://doi.org/10.1371/journal.pone.0121216.
27. Erwin AL. Antibacterial drug discovery targeting the lipopolysaccharide biosynthetic enzyme
LpxC. Cold Spring Harb Perspect Med. 2016. https://doi.org/10.1101/cshperspect.a025304.
28. Estrada A, Wright DL, Anderson AC. Antibacterial antifolates: from development through
resistance to the next generation. Cold Spring Harb Perspect Med. 2016;6(8). https://doi.
org/10.1101/cshperspect.a028324.
29. Fernandes P, Martens E, Pereira D. Nature nurtures the design of new semi-synthetic macro-
lide antibiotics. J Antibiot. 2017;70(5):527–33. https://doi.org/10.1038/ja.2016.137.
30. Fonville NC, Bates D, Hastings PJ, Hanawalt PC, Rosenberg SM. Role of RecA and the SOS
response in thymineless death in Escherichia coli. PLoS Genet. 2010;6(3):e1000865. https://
doi.org/10.1371/journal.pgen.1000865.
31. Giersing BK, Dastgheyb SS, Modjarrad K, Moorthy V. Status of vaccine research and devel-
opment of vaccines for Staphylococcus aureus. Vaccine. 2016;34(26):2962–6. https://doi.
org/10.1016/j.vaccine.2016.03.110.
32. Grillot A-L, Le Tiran A, Shannon D, Krueger E, Liao Y, O’Dowd H, et al. Second-generation
antibacterial benzimidazole ureas: discovery of a preclinical candidate with reduced meta-
bolic liability. J Med Chem. 2014;57(21):8792–816. https://doi.org/10.1021/jm500563g.
33. Grossman TH. Tetracycline antibiotics and resistance. Cold Spring Harb Perspect Med.
2016;6(4):a025387. https://doi.org/10.1101/cshperspect.a025387.
34. Gullberg E, Cao S, Berg OG, Ilbäck C, Sandegren L, Hughes D, et al. Selection of resis-
tant bacteria at very low antibiotic concentrations. PLoS Pathog. 2011;7(7). https://doi.
org/10.1371/journal.ppat.1002158.
35. Hammoudeh DI, Zhao Y, White SW, Lee RE. Replacing sulfa drugs with novel DHPS inhibi-
tors. Future Med Chem. 2013;5(11):1331–40. https://doi.org/10.4155/fmc.13.97.
36. Haynes KM, Abdali N, Jhawar V, Zgurskaya HI, Parks JM, Green AT, et al. Identification and
structure-activity relationships of novel compounds that potentiate the activities of antibiotics
in Escherichia coli. J Med Chem. 2017. https://doi.org/10.1021/acs.jmedchem.7b00453.
37. Hooper DC. Emerging mechanisms of fluoroquinolone resistance. Emerg Infect Dis.
2001;7(2):337–41. https://doi.org/10.3201/eid0702.700337.
38. Howard JJ, Sturge CR, Moustafa DA, Daly SM, Marshall-Batty KR, Felder CF, et al.
Inhibition of Pseudomonas aeruginosa by peptide-conjugated phosphorodiamidate mor-
pholino oligomers. Antimicrob Agents Chemother. 2017:AAC.01938-16. https://doi.
org/10.1128/AAC.01938-16.
39. Huber J, Donald RGK, Lee SH, Jarantow LW, Salvatore MJ, Meng X, et al. Chemical
genetic identification of peptidoglycan inhibitors potentiating carbapenem activity against
methicillin-resistant Staphylococcus aureus. Chem Biol. 2009;16(8):837–48. https://doi.
org/10.1016/j.chembiol.2009.05.012.
40. Hurdle JG, O’Neill AJ, Chopra I. Prospects for aminoacyl-tRNA synthetase inhibitors as
new antimicrobial agents. Antimicrob Agents Chemother. 2005;49(12):4821–33. https://doi.
org/10.1128/AAC.49.12.4821-4833.2005.
41. Janardhanan J, Chang M, Mobashery S. The oxadiazole antibacterials. Curr Opin Microbiol.
2016;33:13–7. https://doi.org/10.1016/j.mib.2016.05.009.
42. Jun SY, Jang IJ, Yoon S, Jang K, Yu K-S, Cho JY, et al. Pharmacokinetics and tolerance
of the phage endolysin-based candidate drug SAL200 after a single intravenous adminis-
tration among healthy volunteers. Antimicrob Agents Chemother. 2017;61(6). https://doi.
org/10.1128/AAC.02629-16.
43. Kim J-H, O’Brien KM, Sharma R, Boshoff HIM, Rehren G, Chakraborty S, et al. A genetic
strategy to identify targets for the development of drugs that prevent bacterial persistence.
Proc Natl Acad Sci. 2013;110(47):19095–100. https://doi.org/10.1073/pnas.1315860110.
18 The New Versus Old Target Debate for Drug Discovery 589

44. Kohanski MA, Dwyer DJ, Hayete B, Lawrence CA, Collins JJ. A common mechanism of
cellular death induced by bactericidal antibiotics. Cell. 2007;130(5):797–810. https://doi.
org/10.1016/j.cell.2007.06.049.
45. Kudrin P, Varik V, Oliveira SRA, Beljantseva J, Santos TDP, Dzhygyr I, et al. Subinhibitory
concentrations of bacteriostatic antibiotics induce relA-dependent and relA-independent
tolerance to β-lactams. Antimicrob Agents Chemother. 2017;61(4):e02173–16. https://doi.
org/10.1128/AAC.02173-16.
46. Lavesa-Curto M, Sayer H, Bullard D, MacDonald A, Wilkinson A, Smith A, et al.
Characterization of a temperature-sensitive DNA ligase from Escherichia coli. Microbiology.
2004;150(Pt 12):4171–80. https://doi.org/10.1099/mic.0.27287-0.
47. Leeds JA. Antibacterials developed to target a single organism: mechanisms and frequen-
cies of reduced susceptibility to the novel anti-Clostridium difficile compounds fidaxomicin
and LFF571. Cold Spring Harb Perspect Med. 2016;6(2):a025445. https://doi.org/10.1101/
cshperspect.a025445.
48. Leeds JA, Dean CR. Peptide deformylase as an antibacterial target: a critical assessment.
Curr Opin Pharmacol. 2006;6(5):445–52. https://doi.org/10.1016/j.coph.2006.06.003.
49. Ling LL, Schneider T, Peoples AJ, Spoering AL, Engels I, Conlon BP, et al. A new antibi-
otic kills pathogens without detectable resistance. Nature. 2015;advance online publication.
https://doi.org/10.1038/nature14098.
50. Liu Y, Imlay JA. Cell death from antibiotics without the involvement of reactive oxygen spe-
cies. Science. 2013;339(6124):1210–3. https://doi.org/10.1126/science.1232751, https://doi.
org/10.1126/science.1232688.
51. Lomovskaya O, Bostian KA. Practical applications and feasibility of efflux pump inhibitors
in the clinic–a vision for applied use. Biochem Pharmacol. 2006;71(7):910–8. https://doi.
org/10.1016/j.bcp.2005.12.008.
52. Maffioli SI, Zhang Y, Degen D, Carzaniga T, Gatto GD, Serina S, et al. Antibacterial
nucleoside-­analog inhibitor of bacterial RNA polymerase. Cell. 2017;169(7):1240–8.e23.
https://doi.org/10.1016/j.cell.2017.05.042.
53. Mahmood HY, Jamshidi S, Sutton JM, Rahman KM. Current advances in developing inhibi-
tors of bacterial multidrug efflux pumps. Curr Med Chem. 2016;23(10):1062–81.
54. Maisonneuve E, Castro-Camargo M, Gerdes K. (p) ppGpp controls bacterial persistence
by stochastic induction of toxin-antitoxin activity. Cell. 2013;154(5):1140–50. https://doi.
org/10.1016/j.cell.2013.07.048, https://doi.org/10.1128/mBio.02267-16.
55. Mann J, Taylor PW, Dorgan CR, Johnson PD, Wilson FX, Vickers R, et al. The discovery
of a novel antibiotic for the treatment of Clostridium difficile infections: a story of an effec-
tive academic-industrial partnership. Med Chem Commun. 2015;6(8):1420–6. https://doi.
org/10.1039/c5md00238a.
56. Mattis DM, Spaulding AR, Chuang-Smith ON, Sundberg EJ, Schlievert PM, Kranz
DM. Engineering a soluble high-affinity receptor domain that neutralizes staphylococcal
enterotoxin C in rabbit models of disease. Protein Eng Des Sel. 2013;26(2):133–42. https://
doi.org/10.1093/protein/gzs094.
57. Maura D, Ballok AE, Rahme LG. Considerations and caveats in anti-virulence drug develop-
ment. Curr Opin Microbiol. 2016;33:41–6. https://doi.org/10.1016/j.mib.2016.06.001.
58. May KL, Silhavy TJ. Making a membrane on the other side of the wall. Biochim Biophys
Acta. 2016. https://doi.org/10.1016/j.bbalip.2016.10.004.
59. McLeod SM, Dougherty TJ, Pucci MJ. Novel antibacterial targets/identification of new tar-
gets by comparative genomics. In: Dougherty TJ, Pucci MJ, editors. Antibiotic discovery and
development. Boston: Springer US; 2012. p. 881–900.
60. Melton-Celsa AR, OʼBrien AD. New therapeutic developments against shiga toxin-­producing
Escherichia coli. Microbiol Spectr. 2014;2(5). https://doi.org/10.1128/microbiolspec.
EHEC-0013-2013.
61. Metz JT, Hajduk PJ. Rational approaches to targeted polypharmacology: creating and navi-
gating protein-ligand interaction networks. Curr Opin Chem Biol. 2010;14(4):498–504.
https://doi.org/10.1016/j.cbpa.2010.06.166.
590 A. L. Erwin

62. Miesel L, Hecht DW, Osmolski JR, Gerding D, Flattery A, Li F, et al. Kibdelomycin is a potent
and selective agent against toxigenic Clostridium difficile. Antimicrob Agents Chemother.
2014;58(4):2387–92. https://doi.org/10.1128/AAC.00021-14.
63. Mills SD, Dougherty TJ. Cell-based screening in antibacterial discovery. In: Dougherty
TJ, Pucci MJ, editors. Antibiotic discovery and development. Boston: Springer US; 2012.
p. 901–29.
64. Mukherjee T, Boshoff H. Nitroimidazoles for the treatment of TB: past, present and future.
Future Med Chem. 2011;3(11):1427–54. https://doi.org/10.4155/fmc.11.90.
65. Mullane K, Lee C, Bressler A, Buitrago M, Weiss K, Dabovic K, et al. A multi-center, ran-
domized clinical trial to compare the safety and efficacy of LFF571 and vancomycin for
Clostridium difficile infections. Antimicrob Agents Chemother. 2014. https://doi.org/10.1128/
AAC.04251-14.
66. Nair DR, Monteiro JM, Memmi G, Thanassi J, Pucci M, Schwartzman J, et al. Characterization
of a novel small molecule that potentiates β-lactam activity against Gram positive and
Gram negative pathogens. Antimicrob Agents Chemother. 2015. https://doi.org/10.1128/
AAC.04164-14.
67. Nayak SU, Griffiss JM, Blumer J, O’Riordan MA, Gray W, McKenzie R, et al. Safety, toler-
ability, systemic exposure and metabolism of CRS3123, a methionyl-tRNA synthetase inhib-
itor developed for treatment of Clostridium difficile infections, in a Phase I study. Antimicrob
Agents Chemother. 2017. https://doi.org/10.1128/AAC.02760-16.
68. Nayar AS, Dougherty TJ, Ferguson KE, Granger BA, McWilliams L, Stacey C, et al. Novel
antibacterial targets and compounds revealed by a high-throughput cell wall reporter assay.
J Bacteriol. 2015;197(10):1726–34. https://doi.org/10.1128/JB.02552-14.
69. Novak R. Are pleuromutilin antibiotics finally fit for human use? Ann N Y Acad Sci.
2011;1241(1):71–81. https://doi.org/10.1111/j.1749-6632.2011.06219.x.
70. O’Dwyer K, Spivak AT, Ingraham K, Min S, Holmes DJ, Jakielaszek C, et al. Bacterial resis-
tance to leucyl-tRNA synthetase inhibitor GSK2251052 develops during treatment of com-
plicated urinary tract infections. Antimicrob Agents Chemother. 2015;59(1):289–98. https://
doi.org/10.1128/AAC.03774-14.
71. O’Neill AJ, Chopra I. Preclinical evaluation of novel antibacterial agents by microbiological
and molecular techniques. Expert Opin Investig Drugs. 2004;13(8):1045–63.
72. Opperman TJ, Kwasny SM, Li JB, Lewis MA, Aiello D, Williams JD, et al. DNA target-
ing as a likely mechanism underlying the antibacterial activity of synthetic bis-indole anti-
biotics. Antimicrob Agents Chemother. 2016;60(12):7067–76. https://doi.org/10.1128/
AAC.00309-16.
73. Page MGP. Beta-lactam antibiotics. In: Dougherty TJ, Pucci MJ, editors. Antibiotic discovery
and development. Boston: Springer US; 2012. p. 79–117.
74. Payne DJ, Gwynn MN, Holmes DJ, Pompliano DL. Drugs for bad bugs: confronting the
challenges of antibacterial discovery. Nat Rev Drug Discov. 2007;6(1):29–40. https://doi.
org/10.1038/nrd2201.
75. Qiu X-Q, Wang H, Lu X-F, Zhang J, Li S-F, Cheng G, et al. An engineered multidomain bac-
tericidal peptide as a model for targeted antibiotics against specific bacteria. Nat Biotechnol.
2003;21(12):1480–5. https://doi.org/10.1038/nbt913.
76. Que YA, Lazar H, Wolff M, François B, Laterre PF, Mercier E, et al. Assessment of pan-
obacumab as adjunctive immunotherapy for the treatment of nosocomial Pseudomonas
aeruginosa pneumonia. Eur J Clin Microbiol Infect Dis. 2014;33(10):1861–7. https://doi.
org/10.1007/s10096-014-2156-1.
77. Ramachandran G, Kaempfer R, Chung C-S, Shirvan A, Chahin AB, Palardy JE, et al. Cd28
homodimer interface mimetic peptide acts as a preventive and therapeutic agent in models of
severe bacterial sepsis and Gram-negative bacterial peritonitis. J Infect Dis. 2015;211(6):995–
1003. https://doi.org/10.1093/infdis/jiu556.
78. Rothstein DM. Rifamycins, alone and in combination. Cold Spring Harb Perspect Med.
2016;6(7). https://doi.org/10.1101/cshperspect.a027011.
18 The New Versus Old Target Debate for Drug Discovery 591

79. Ruiz N, Kahne D, Silhavy TJ. Advances in understanding bacterial outer-membrane biogen-
esis. Nat Rev Microbiol. 2006;4(1):57–66. https://doi.org/10.1038/nrmicro1322.
80. Ruiz N, Kahne D, Silhavy TJ. Transport of lipopolysaccharide across the cell envelope:
the long road of discovery. Nat Rev Microbiol. 2009;7(9):677–83. https://doi.org/10.1038/
nrmicro2184.
81. Schneider T, Sahl H-G. An oldie but a goodie – cell wall biosynthesis as antibiotic target path-
way. Int J Med Microbiol. 2010;300(2–3):161–9. https://doi.org/10.1016/j.ijmm.2009.10.005.
82. Schuch R, Khan BK, Raz A, Rotolo JA, Wittekind M. Bacteriophage lysin CF-301: a
potent anti-staphylococcal biofilm agent. Antimicrob Agents Chemother. 2017. https://doi.
org/10.1128/AAC.02666-16.
83. Scott RW, Tew GN. Mimics of host defense proteins; strategies for translation to therapeutic
applications. Curr Top Med Chem. 2017;17(5):576–89.
84. Seiple IB, Zhang Z, Jakubec P, Langlois-Mercier A, Wright PM, Hog DT, et al. A platform
for the discovery of new macrolide antibiotics. Nature. 2016;533(7603):338–45. https://doi.
org/10.1038/nature17967.
85. Shan Y, Gandt AB, Rowe SE, Deisinger JP, Conlon BP, Lewis K. ATP-dependent per-
sister formation in Escherichia coli. MBio. 2017;8(1):e02267–16. https://doi.org/10.1128/
mBio.02267-16.
86. Sharma P, Wang N, Kranz DM. Soluble T cell receptor Vβ domains engineered for high-­affinity
binding to staphylococcal or streptococcal superantigens. Toxins (Basel). 2014;6(2):556–74.
https://doi.org/10.3390/toxins6020556.
87. Silver LL. Multi-targeting by monotherapeutic antibacterials. Nat Rev Drug Discov.
2007;6(1):41–55. https://doi.org/10.1038/nrd2202.
88. Silver LL. Challenges of antibacterial discovery. Clin Microbiol Rev. 2011;24(1):71–109.
https://doi.org/10.1128/CMR.00030-10.
89. Silver LL. A Gestalt approach to Gram-negative entry. Bioorg Med Chem. 2016. https://doi.
org/10.1016/j.bmc.2016.06.044.
90. Singh SB, Young K, Miesel L. Screening strategies for discovery of antibacterial natural
products. Expert Rev Anti-Infect Ther. 2011;9(8):589–613. https://doi.org/10.1586/eri.11.81.
91. Sjuts H, Vargiu AV, Kwasny SM, Nguyen ST, Kim H-S, Ding X, et al. Molecular basis for
inhibition of AcrB multidrug efflux pump by novel and powerful pyranopyridine derivatives.
Proc Natl Acad Sci U S A. 2016;113(13):3509–14. https://doi.org/10.1073/pnas.1602472113.
92. Sköld O. Sulfonamides and trimethoprim. Expert Rev Anti-Infect Ther. 2010;8(1):1–6.
https://doi.org/10.1586/eri.09.117.
93. Srinivas N, Jetter P, Ueberbacher BJ, Werneburg M, Zerbe K, Steinmann J, et al.
Peptidomimetic antibiotics target outer-membrane biogenesis in Pseudomonas aeruginosa.
Science. 2010;327(5968):1010–3. https://doi.org/10.1126/science.1182749.
94. Starkey M, Lepine F, Maura D, Bandyopadhaya A, Lesic B, He J, et al. Identification of anti-­
virulence compounds that disrupt quorum-sensing regulated acute and persistent pathogenic-
ity. PLoS Pathog. 2014;10(8). https://doi.org/10.1371/journal.ppat.1004321.
95. Suckling CJ. The antibacterial drug MGB-BP3 : from discovery to clinical trial. Chem Biol
Interface. 2015;5(3):166–74.
96. Takatsuka Y, Chen C, Nikaido H. Mechanism of recognition of compounds of diverse struc-
tures by the multidrug efflux pump AcrB of Escherichia coli. Proc Natl Acad Sci U S A.
2010;107(15):6559–65. https://doi.org/10.1073/pnas.1001460107.
97. Tari LW, Li X, Trzoss M, Bensen DC, Chen Z, Lam T, et al. Tricyclic GyrB/ParE (TriBE)
inhibitors: a new class of broad-spectrum dual-targeting antibacterial agents. PLoS One.
2013;8(12):e84409. https://doi.org/10.1371/journal.pone.0084409.
98. Therien AG, Huber JL, Wilson KE, Beaulieu P, Caron A, Claveau D, et al. Broadening the
spectrum of β-lactam antibiotics through inhibition of signal peptidase type I. Antimicrob
Agents Chemother. 2012;56(9):4662–70. https://doi.org/10.1128/AAC.00726-12.
99. Tommasi R, Brown DG, Walkup GK, Manchester JI, Miller AA. ESKAPEing the labyrinth of
antibacterial discovery. Nat Rev Drug Discov. 2015;14(8):529–42. https://doi.org/10.1038/
nrd4572.
592 A. L. Erwin

100. Utaida S, Dunman PM, Macapagal D, Murphy E, Projan SJ, Singh VK, et al. Genome-wide
transcriptional profiling of the response of Staphylococcus aureus to cell-wall-active antibiot-
ics reveals a cell-wall-stress stimulon. Microbiology. 2003;149(Pt 10):2719–32. https://doi.
org/10.1099/mic.0.26426-0.
101. Vaara M, Siikanen O, Apajalahti J, Fox J, Frimodt-Møller N, He H, et al. A novel polymyxin
derivative that lacks the fatty acid tail and carries only three positive charges has strong syn-
ergism with agents excluded by the intact outer membrane. Antimicrob Agents Chemother.
2010;54(8):3341–6. https://doi.org/10.1128/AAC.01439-09.
102. van Miert AS. The sulfonamide-diaminopyrimidine story. J Vet Pharmacol Ther.
1994;17(4):309–16.
103. Wang DY, Abboud MI, Markoulides MS, Brem J, Schofield CJ. The road to avibactam: the
first clinically useful non-β-lactam working somewhat like a β-lactam. Future Med Chem.
2016;8(10):1063–84. https://doi.org/10.4155/fmc-2016-0078.
104. Wang H, Mann PA, Xiao L, Gill C, Galgoci AM, Howe JA, et al. Dual-targeting small-­
molecule inhibitors of the Staphylococcus aureus FMN riboswitch disrupt riboflavin
homeostasis in an infectious setting. Cell Chem Biol. 2017;24(5):576–88.e6. https://doi.
org/10.1016/j.chembiol.2017.03.014.
105. Wei J-R, Krishnamoorthy V, Murphy K, Kim J-H, Schnappinger D, Alber T, et al.
Depletion of antibiotic targets has widely varying effects on growth. Proc Natl Acad Sci.
2011;108(10):4176–81. https://doi.org/10.1073/pnas.1018301108.
106. Wright PM, Seiple IB, Myers AG. The evolving role of chemical synthesis in antibacterial
drug discovery. Angew Chem Int Ed Engl. 2014;53(34):8840–69. https://doi.org/10.1002/
anie.201310843.
107. Xu W, Wang W, Wang X. Gold-catalyzed cyclization leads to a bridged tetracyclic indolenine
that represses β-lactam resistance. Angew Chem Int Ed. 2015;54(33):9546–9. https://doi.
org/10.1002/anie.201503736.
108. Yao J, Rock CO. Bacterial fatty acid metabolism in modern antibiotic discovery. Biochim
Biophys Acta. 2016. https://doi.org/10.1016/j.bbalip.2016.09.014.
109. Young K, Silver LL, Bramhill D, Cameron P, Eveland SS, Raetz CR, et al. The envA perme-
ability/cell division gene of Escherichia coli encodes the second enzyme of lipid A biosyn-
thesis. UDP-3-O-(R-3-hydroxymyristoyl)-N-acetylglucosamine deacetylase. J Biol Chem.
1995;270(51):30384–91.
110. Yuan Z, White RJ. The evolution of peptide deformylase as a target: contribution of bio-
chemistry, genetics and genomics. Biochem Pharmacol. 2006;71(7):1042–7. https://doi.
org/10.1016/j.bcp.2005.10.015.
111. Zhou J, Bhattacharjee A, Chen S, Chen Y, Duffy E, Farmer J, et al. Design at the atomic level:
generation of novel hybrid biaryloxazolidinones as promising new antibiotics. Bioorg Med
Chem Lett. 2008;18(23):6179–83. https://doi.org/10.1016/j.bmcl.2008.10.014.
Chapter 19
Non-quinolone Topoisomerase Inhibitors

Anthony Maxwell, Natassja G. Bush, Thomas Germe, and Shannon J. McKie

19.1 Introduction

There is no question that the development of antibiotics has been among the most
important advances of the twentieth century, saving a countless number of lives. Not
only are antibiotics used in the direct treatment of bacterial infections, but they are
also important in surgical situations (e.g., transplant surgery, joint replacement) by
preventing infections. However, the twenty-first century has seen increasing con-
cern due to the rise in antimicrobial-resistant bacterial infections [1]. Antibiotic
resistance is a growing global threat, with resistance to all classes of antibiotics now
reported worldwide. The resistance problem is compounded by a lack of innovation
and few new structural classes of antibiotics being brought to the clinic [2–4]. To
tackle antibiotic resistance, we need to review our stewardship of existing antibiot-
ics and expand efforts to discover new agents that are not susceptible to known
resistance mechanisms.
Quinolones (specifically fluoroquinolones (FQs); Fig. 19.1) are a potent class of
synthetic antibiotics that target DNA gyrase and DNA topoisomerase IV, essential
enzymes that are ubiquitous among bacterial species. The quinolones were discov-
ered in the early 1960s, and they are now the most successful class of topoisomerase
inhibitors; fluoroquinolones are widely prescribed in the USA, Europe, and most
regions of the world [5, 6]. This heavy consumption of fluoroquinolones has led to
an increase in resistance that derives from a variety of processes including upregula-
tion of efflux pumps, reduced ability to take up the drug, plasmid-based resistance,
or mutations in the gyrase and/or topo IV genes [7, 8]. This widespread resistance
has resulted in revised stewardship guidelines for quinolones [6] as well as the

A. Maxwell (*) · N. G. Bush · T. Germe · S. J. McKie

Department of Biological Chemistry, John Innes Centre, Norwich Research Park,
Norwich NR4 7UH, UK
e-mail: [email protected]

© Springer International Publishing AG, part of Springer Nature 2018 593

I. W. Fong et al. (eds.), Antimicrobial Resistance in the 21st Century, Emerging Infectious
Diseases of the 21st Century, https://doi.org/10.1007/978-3-319-78538-7_19
594 A. Maxwell et al.

Fig. 19.1 Fluoroquinolone compounds. The basic FQ skeleton is shown along with three FQ
compounds in clinical use

WHO categorizing quinolones as highest priority critically important antimicrobi-

als [9].
Due to the success of DNA gyrase (and topo IV) as a clinical target for antimi-
crobials and to the troubling issues associated with quinolone resistance, other com-
pounds, particularly novel inhibitors of topoisomerases, are logical alternatives to
quinolones. These novel agents are the subject of this chapter.
19 Non-quinolone Topoisomerase Inhibitors 595

19.2 DNA Topoisomerases

DNA topoisomerases are enzymes that can interconvert different topological forms
of DNA; their reactions include relaxation of supercoils, decatenation, and unknot-
ting [10, 11]. They are present and essential in all organisms and are involved in
DNA replication and transcription, preventing the buildup of unwanted supercoils
and resolving catenated products [12]. Topoisomerases are classified into two types,
I and II, depending on whether they catalyze reactions involving single (I)- or dou-
ble (II)-stranded breaks in DNA; they can also be further divided into subtypes IA,
IB, and IC and IIA and IIB, dependent on mechanistic and evolutionary consider-
ations [13, 14]. The subjects of this chapter, DNA gyrase (Fig. 19.2) and DNA
topoisomerase (topo) IV, are type IIA enzymes. Type I enzymes catalyze their reac-
tions by transiently breaking one strand of DNA and forming a covalent bond
between either the 5′ (IA) or 3′ (IB) phosphate at the break site and the active-site
tyrosine of the enzyme. The topoisomerase reaction occurs via a swivel mechanism
(type IB) or by strand passage, where a single- or double-stranded segment of DNA
is passed through the break (type IA). Type II topoisomerases make a transient
double-strand break in DNA, forming covalent bonds with the 5′-phosphates at the
break site and passing a double-stranded segment of DNA (the so-called “T” or
“transported” segment) through the break [10, 11] in the “G” or “gate” segment.
Although this is the case for both IIA and IIB enzymes, the two subtypes are differ-
ent in several other structural and mechanistic aspects [15]. The principal member
of the type IIB subtype is topo VI, originally discovered in archaea and now shown
to occur in plants and plasmodial parasites [16, 17]. The members of the IIA sub-
type, gyrase, topo IV, and eukaryotic topo II, are better known and have been more
extensively studied. All these enzymes can relax supercoiled DNA; gyrase is dis-
tinct in also being able to catalyze the introduction of negative supercoils into DNA.
Structural and mechanistic aspects of these enzymes have been extensively reviewed

Fig. 19.2 Schematic representation of DNA gyrase (A2B2) complexed with G-segment
DNA. NTD, N-terminal domain; CTD, C-terminal domain; TOPRIM, topoisomerase-primase
domain; WHD, winged-helix domain
596 A. Maxwell et al.

elsewhere [10, 11, 18]. The main topic of this chapter concerns their role as targets
for antibacterial agents; other recent reviews also address this topic [10, 19–21].

19.3  opoisomerases as Targets for Antibacterial

T
Chemotherapy

Because of their essentiality and the fact that their reactions proceed via transient
breaks in DNA, topoisomerases have become targets for both anticancer and anti-
bacterial chemotherapy [10, 21, 22]. Many topoisomerase-targeted drugs act by sta-
bilizing the DNA-protein covalent complexes that form between the enzymes and
DNA during the topoisomerase reaction cycle. However, there are other agents (see
below) that act by interrupting the reaction cycle in other ways; these have been
dubbed “catalytic inhibitors.” Quinolone compounds, and more specifically fluoro-
quinolones (FQs), such as ciprofloxacin, moxifloxacin, and levofloxacin (Fig. 19.1),
are highly successful antibiotics that are used for a wide range of clinical conditions
[23, 24]. However, due to resistance to these compounds, new agents that match
their clinical characteristics, but are not affected by quinolone-resistance mutations,
are urgently needed. This chapter reviews such compounds.
Although, in principle, DNA topo I is a valid target for antibacterial compounds
and a number have been investigated [25], there are currently no topo I inhibitors in
clinical use, although recent work suggests that there is scope to develop antibiotics
targeted to this enzyme [26, 27]; topo I will not be further discussed in this chapter.
The scope of DNA gyrase and DNA topo I as targets for tuberculosis therapy has
recently been reviewed [20].
There are two main mechanisms of inhibition of topoisomerases: catalytic
inhibition and topoisomerase poisons. Catalytic inhibitors arrest enzyme activity
and need to bind to their targets with reasonably high affinity to be effective.
Poisons are compounds that stabilize the topoisomerase-DNA cleavage complex
and tend to be more effective, as relatively low occupancy of the inhibitor bound
to its target can lead to cell death, which involves chromosome fragmentation,
induction of the SOS response, and possibly the induction of reactive oxygen
species [28, 29] (see Chap. 20).

19.4 Cleavage Complex-Stabilizing Agents

Binding at the DNA Cleavage Site

The high degree of success of FQs is at least in part due to their mode of action: the
ability to arrest their targets (gyrase and/or topo IV) at the stage in the topoisomer-
ase reaction cycle where the enzyme is covalently bound to a double-stranded DNA
break. This so-called cleavage complex results in chromosomal breaks and cell
19 Non-quinolone Topoisomerase Inhibitors 597

death; it is also the basis for the mode of action of a number of successful antitumor
drugs that target human topo II [30]. Finding agents that act in a similar way that can
substitute for FQs has been a major challenge.
As we now know, from X-ray crystallography, how FQs work at the molecular
level [31, 32], we can use this information to develop other molecules that act in a
similar way but that may avoid the problems of quinolone resistance. As will be
seen from some of the examples below, it is possible to find compounds that bind at
sites distinct from the quinolone-binding site and that can also stabilize cleavage
complexes. This type of compound that can avoid quinolone cross-resistance per-
haps represents the best opportunities for finding novel gyrase−/topo IV-targeted
antibiotics with clinical potential. The term NBTI (novel (non-fluoroquinolone)
bacterial type II topoisomerase inhibitor) has been introduced [31] to encompass
compounds that demonstrate these properties.
FQs work through intercalation of two compounds, one at each cleavage site,
into the DNA (Fig. 19.3) [31, 32]; the compounds also interact with the enzyme
through a water-metal-ion bridge [33]. Mutations affecting the residues involved in
contacting the metal ion are found in clinically-isolated strains of FQ-resistant
pathogenic bacteria. It is likely that disrupting the water-metal ion bridge destabi-
lizes the ternary complex, mitigating cleavage-complex formation (see Chap. 16).
Efforts have therefore been made to develop compounds that do not rely on this
water-metal ion bridge. Several compounds share a scaffold with the FQs, for exam-
ple, the quinazolinediones (QZDs) [34–36] and the imidazopyrazinones (IPYs)
[37, 38]. Compounds of these classes are able to intercalate at the FQ-binding site
and stabilize cleavage complexes without relying on the water-metal ion bridge [37,
39, 40]. The water-metal ion bridge is a major contributor to the stable binding of
bacterial topoisomerase by the FQs. Removing it requires other contacts to be estab-
lished by putative intercalating compounds of similar scaffold to afford the same
affinity and, by extension, equivalent potency (although the relationship between
efficiency of poisoning in vitro and potency against bacterial pathogens is far from
straightforward). Indeed in the case of QZDs, efficiency of poisoning tends to be
significantly lower than the FQs [41–43], and some of the medicinal chemistry
effort has focused on developing alternative contacts with the protein, for instance,
by modifying the C7 substituent [44]. In the case of the IPYs, the absence of a water-
metal ion bridge is compensated for by direct contact with one of the residues
involved in the bridge, which unfortunately affords some degree of cross-resistance
between quinolones and these compounds [37]. The IPYs also establish contact
with the arginine situated next to the catalytic tyrosine, which is presumably essen-
tial for catalytic activity [37]. In the case of the QZDs, attempts have been made to
develop such contacts in order to improve the activity of the compounds while mini-
mizing resistance [45].
Structural information on the protein-DNA-drug complexes has shown that they
all share a characteristic DNA extension, increasing the distance between the two
scissile phosphates compared to an uncleaved “binary” enzyme-DNA complex [31,
33, 37, 46]. In gyrase, this extension is associated with conformational movement
involving sliding of the two GyrA subunits against each other and tilting of the
598 A. Maxwell et al.

Fig. 19.3 Binding sites of cleavage complex-stabilizing compounds. Two structures are super-
posed to show binding by moxifloxacin (PDB 5CDQ), GSK945237, and thiophene2 (PDB 5NPP)
to the S. aureus gyrase core (C-terminal region of GyrB fused to the N-terminal region of GyrA).
The TOPRIM domains from 5CDQ are superposed and displayed instead of the ones from 5NPP;
the DNA and GyrA domain displayed are from 5NPP. An orientation cartoon shows the color-­
coded compounds binding at distinct sites on the enzyme. While moxifloxacin intercalates directly
at the cleavage site (as with the IPYs and QZDs), the triazaacenaphthylene and thiophene com-
pounds bind at different allosteric sites and stabilize the cleaved intermediate. Triazaacenaphthylenes
block the sliding of the GyrA subunits against one another, thereby blocking resealing, whereas the
thiophene blocks the TOPRIM domain in a tilted conformation against the DNA

TOPRIM domains toward the dyad axis, suggesting that the enzyme can manipulate
DNA geometry to favor cleavage. The dynamic nature of the DNA gate also allows
for a variety of structurally unrelated compounds to bind at the FQ site without
requiring a water-metal ion bridge and thereby bypassing most FQ-resistance muta-
tions. For instance, etoposide, a eukaryotic topoisomerase II inhibitor [46], is also
able to stabilize gyrase-DNA cleavage complexes through intercalation into DNA
and interaction with residues conserved in the related human topoisomerase
II. Likewise, the spiropyrimidinetriones (such as QPT-1) can stabilize cleavage
complexes by intercalation much like the FQs despite being structurally different
[46]. However, unlike etoposide, QPT-1 is not active on human topoisomerase II
despite interacting with conserved residues. It was thought that the water-metal ion
bridge conferred bacterial specificity to the FQs, but the case of QPT-1 shows that
more is at play. This is an important issue, as development of bacterial ­topoisomerase
19 Non-quinolone Topoisomerase Inhibitors 599

II poisons incurs the risk of developing genotoxic compounds by cross-­reactivity

with human topoisomerase II.
Compounds Binding Outside the DNA Cleavage Site
The extensive conformational changes involved in DNA cleavage open up the theo-
retical possibility that compounds binding away from the cleavage site affect cleav-
age by stabilizing the cleavage-prone conformation of the enzyme. Indeed, the
triazaacenaphthylenes (the original NBTIs), e.g., gepotidacin (GSK2140944
[47]), bind at the interface between the two sliding GyrA subunits and act as a
“locking pin” to freeze the enzyme into the DNA-extended state that favors cleav-
age [31]. Moreover, an allosteric pocket has been identified that is targeted by thio-
phene compounds [48]. Binding to this pocket, which is remote from the DNA
cleavage site, results in cleavage-complex stabilization, presumably by allosteri-
cally locking the enzyme in the cleavage-prone conformation. This binding pocket
is located at the base of the GyrAα1 helix hinging the tilting of the TOPRIM domain
[37] (Figs. 19.2 and 19.3). This is similar to the mode of action of the NBTIs but
involves the other segment of the enzyme implicated in the conformational transi-
tion to the cleaved state (see above). The utilization of novel pockets that exploit the
natural ability of the enzyme to transition into the cleaved state has the advantage of
bypassing existing resistance to FQs. Moreover, given the conformational changes
involved in cleavage, it can be expected that mutations restricting access to these
pockets would reduce the ability of the enzyme to cleave DNA and thereby reduce
its activity. Consistent with this idea, the frequency of resistance is low for the thio-
phene compounds [48]. However, it is theoretically possible for a mutation to affect
the energy of transition to the cleaved state rather than the binding of the compound.
Indeed, mutations conferring resistance to different series of compounds have been
located away from the compound’s binding site. This raises the possibility of “uni-
versal” mutations that would confer resistance to all cleavage-stabilizing agents.
This idea remains to be tested. Candidates for such universal mutation include Asp82
to Asn (Escherichia coli numbering) that confers resistance to both FQs and IPY
[37] and the Val96 to Ala (in Bacillus anthracis) which confers resistance to both
FQs and QZDs [49].
This thermodynamic view of poisoning as the stabilization of a natural, cleavage-­
prone conformation can also help explain the bacterial specificity of compounds
like QPT-1 that interact with residues conserved in human topo II. The binding
energy of a compound must offset the energetic cost of the transition to the cleaved
state for cleavage to be stabilized. Therefore, if the cost of the transition differs
between the bacterial and the human enzyme, we can envision a situation whereby
a similar binding energy is sufficient to offset the cost of cleavage for one but not the
other. This effect could also contribute to the variation in activity of a given com-
pound against different bacterial species.
Non-small Molecule Inhibitors
Although not a small molecule, the bacterial toxin microcin B17 (MccB17) is also
able to stabilize the cleavage complex between gyrase and DNA. MccB17 is a
3.1 kDa posttranslationally modified peptide that contains 8 or 9 oxazole and
600 A. Maxwell et al.

thiazole heterocycles [50]. MccB17 targets bacterial gyrase and can stabilize the
cleavage complex but in a manner distinct from quinolones [51, 52]. The only
known mutation in gyrase that confers resistance to MccB17 is at the C-terminal
end of GyrB (Trp751 to Arg) [52, 53]; no other drug-resistance mutations map here.
Although MccB17 is a potentially attractive option as an antibacterial compound,
its poor physicochemical properties have hampered its development as a drug can-
didate. Despite significant work on this toxin, its binding site and mode of action on
gyrase are not known. However, the toxin and fragments thereof have been chemi-
cally synthesized, and fragments have also been made using molecular biology/
biochemical methods [54–57]. Some of these fragments, with molecular weights
<2 kDa, show activity suggesting that it may be possible to generate smaller ver-
sions of MccB17 with more attractive physicochemical properties that might have
potential as antibacterial agents in the future.
Other non-small molecule inhibitors of gyrase include the phytotoxin albicidin,
the CcdB protein toxin, the E. coli GyrI protein, and the pentapeptide-repeat pro-
teins, such as Qnr and MfpA; these agents have been reviewed elsewhere [10, 19].
Although the toxin protein CcdB (MW ~12 kDa) is outside the scope of this review,
it is interesting to note that peptides based on CcdB as short as 18 amino acids can
retain inhibitory activity on gyrase [58].

19.5 Catalytic Inhibitors

The term “catalytic inhibitors” refers to agents that do not inhibit gyrase/topo IV by
stabilizing the DNA cleavage complex but affect another aspect of the catalytic
cycle. The majority of these are ATPase inhibitors, e.g., aminocoumarins and cyclo-
thialidines, but other types, such as simocyclinones, which inhibit DNA binding,
also exist. Although arguably catalytic inhibitors are less likely than cleavage
complex-­stabilizing compounds to be effective antibiotics, such catalytic inhibition
is effective with other antibiotic targets (e.g., rifamycin on RNA polymerase, trim-
ethoprim on dihydrofolate reductase [59]); thus, there is no reason a priori that they
should not succeed as antibiotics. Indeed novobiocin, see below, has been utilized
as a clinical antibiotic.
Aminocoumarin antibiotics (Fig. 19.4) that target DNA gyrase were discovered
as Streptomyces natural products in the 1950s; these “classical” agents are novobio-
cin, clorobiocin, and coumermycin A1 [60, 61]. Early on it was established that
these compounds are competitive inhibitors of the gyrase ATPase reaction [62]; they
also have activity against topo IV [63]. In this sense, they are classic catalytic inhibi-
tors that affect ATP-dependent topoisomerase reactions without stabilizing the
cleavage complex. Specifically, they bind to the ATPase (N-terminal) domain of
GyrB and block the binding of ATP, a process that was definitively established using
X-ray crystallography [64, 65]. This conclusion was somewhat surprising given that
aminocoumarins do not obviously resemble ATP (Fig. 19.4). In fact, it is the sugar
ring of the aminocoumarins that overlaps the adenine-binding site in the ATPase
19 Non-quinolone Topoisomerase Inhibitors 601

Fig. 19.4 Catalytic inhibitors of gyrase and topo IV. (a) Novobiocin, clorobiocin, novclobiocin
401, and coumermycin A1, (b) cyclothialidine, (c) simocyclinone D8. ATPase inhibitors are boxed
in orange, DNA-binding inhibitors in green

pocket, thereby preventing the binding of ATP. Several crystal structures of gyrase
B fragments bound to aminocoumarin compounds now exist (Fig. 19.5 [19, 61]);
these fragments have potentiated the design of alternative compounds that can bind
at the same site (see below).
Although aminocoumarins are very effective inhibitors of gyrase and topo IV,
with Kd values in the 1–20 nM range [66], they have not had a high degree of suc-
cess as clinical antibiotics. Although novobiocin has been used on its own and in
combination as an antibiotic, safety concerns have led to discontinuation of its
usage [19]. Its toxicity issues may stem in part from its binding site, the ATPase
domain of GyrB/ParE, which is part of the GHKL ATPase/kinase superfamily [67],
and secondary eukaryotic targets are therefore likely. Indeed it has been possible to
“redesign” novobiocin to target Hsp90 (see below) [68]. Furthermore, the amino-
coumarins suffer from solubility issues, making it difficult to develop them as drugs.
To circumvent these problems, attempts have been made to prepare aminocou-
marin derivatives with superior properties. These efforts have been made possible
by the identification, sequencing, and annotation of the gene clusters for the three
“classical” aminocoumarins [69]. This has led to a detailed understanding of the
biosynthetic pathways for these compounds [60]. With this information it has been
possible to use various technologies: genetic engineering, combinatorial biosynthe-
sis, and mutasynthesis, to generate novel aminocoumarins, which can then be tested
for antibacterial activity and for their effects on the target enzymes [70]. A signifi-
cant amount of this effort has been carried out by Heide and coworkers, who coined
602 A. Maxwell et al.

Fig. 19.5 Binding site for aminocoumarins and other ATPase inhibitors. The superposition of
structures of the ATP-binding domain of DNA gyrase bound to ADPNP (5′-adenylyl-β,γ-­
imidodiphosphate), novobiocin, and cyclothialidine. The nucleotide binds inside a furrow and pro-
motes dimerization through binding of a projection from the cognate monomer covering the
opening of the furrow. Both novobiocin and cyclothialidine impinge on nucleotide binding as their
binding sites overlap with the nucleotide-binding site

the name “novclobiocins” for molecules that are hybrids between clorobiocin and
novobiocin [71]. Many novclobiocins have been produced and their activities
assessed [69, 72]. One specific example is novclobiocin 401 (Fig. 19.4), which con-
tains the catechol moiety 3,4-dihydroxybenzoic acid, which assists import across
the bacterial cell envelope [73]. This modification improved the penetration into E.
coli, and the analog also retained good activity against gyrases from E. coli and
Staphylococcus aureus. Whether further improvements to the aminocoumarins can
generate compounds with clinical potential remains to be seen.
Cyclothialidines (Fig. 19.4) are Streptomyces-derived natural products and, like
aminocoumarins, are also competitive inhibitors of the gyrase ATPase reaction and
bind at essentially the same site in GyrB [65]. Although their antibacterial potency
is generally poor, their novel structures and good in vitro activity against gyrase
warranted further investigation [74–76]. For example, a chemistry program at
19 Non-quinolone Topoisomerase Inhibitors 603

F. Hoffmann-La Roche generated a large number of cyclothialidine analogs, includ-

ing compounds with improved in vivo efficacy [77, 78]. It remains to be seen
whether such compounds can lead to clinically successful antibiotics.
During work on the classical aminocoumarins (described above), the related
simocyclinones (Fig. 19.4) were discovered [79, 80]. These are also Streptomyces
natural products and contain an aminocoumarin group, but they also comprise a
polyketide moiety and linker. The biosynthetic gene cluster for simocyclinones
shares genes related to those found in the gene clusters of classical aminocoumarins
[81]. Several simocyclinones were discovered [82, 83], with simocyclinone D8
(SD8) being the best studied. It was expected that simocyclinones would bind at the
same site as the aminocoumarins, i.e., the ATPase site of GyrB, and it came as a real
surprise when it was found that these compounds did not inhibit the gyrase ATPase
reaction but prevented the enzyme from binding DNA [84]. This unexpected result
was confirmed by X-ray crystallography, which showed SD8 bound to the N-terminal
domain of GyrA at the DNA G-segment binding site [85]. Subsequent mass spec-
trometry and further X-ray crystallography [86, 87] generated a modified model for
the SD8-GyrA complex that satisfied all the mutant data and biophysical analyses
(Fig. 19.6). Although simocyclinones are thought not to be particularly potent anti-
bacterials (but see below) and this mode of action is arguably less effective than
cleavage-complex stabilization, this work nonetheless showed that there are alterna-
tive modes of inhibition of gyrase aside from the quinolone and aminocoumarin
mechanisms.
The crystallography work [85, 87], backed up by mutational analysis, firmly
established that the simocyclinone-binding site lies in the N-terminal domain of
GyrA, but there is evidence from circular dichroism studies of a second binding site
in the C-terminal domain of GyrB [88]. The existence of this second site was cor-
roborated by isothermal titration calorimetry experiments [87], but its affinity for
SD8 was found to be ~1000-fold weaker than the GyrA site, suggesting that the
latter is likely to be the primary target. However, the existence of a second site sug-
gests promiscuousness in simocyclinone binding, which has been reflected in other
work (see below).
Although the antibacterial potency of simocyclinones is thought to be relatively
weak, particularly against Gram-negative bacteria [82], it has been pointed out that
these tests are generally carried out on laboratory strains. Experiments assessing the
potency of SD8 against clinical isolates of E. coli and Klebsiella pneumoniae sug-
gested that these compounds may be more active in a clinical setting [89]. However,
the reported activity of simocyclinones against human topoisomerases [90, 91] sug-
gests that they may lack the selectivity required to be effective antibiotics. Perhaps
the most promising aspect of simocyclinones is the identification of a novel mode of
action that is distinct from those of quinolones and aminocoumarins. Such a mode
presents the prospect of developing more “drug-like” molecules that can exploit this
binding mode.
As the site of SD8 binding was found to be close to that of the FQs (Fig. 19.6),
this raised the possibility of making hybrid compounds, i.e., quinolone-based com-
pounds whose affinity would be enhanced by being further anchored to the
604 A. Maxwell et al.

Fig. 19.6 Binding site of simocyclinone. The compound binds in a “saddle” formed by the two
GyrA subunits that normally accommodates the G-DNA; two molecules of simocyclinone D8 bind
per GyrA dimer. GyrA55 is a truncated version of the GyrA-NTD

a­ minocoumarin pocket of simocyclinones. A series of ciprofloxacin-coumarin ana-

logs was synthesized, suggesting that it is possible to generate compounds that bind
to the FQ-binding site and the coumarin pocket of SD8 and still retain potency [92].
It remains to be seen whether such compounds can be further developed as potential
antibiotics. Interestingly, in other work, flavone-based analogs of simocyclinone
were synthesized in order to create additional binding opportunities [93]. Two of
these compounds were found to inhibit DNA gyrase, but as they also stabilized the
DNA cleavage complex, inhibition was probably via a different mechanism; these
compounds were shown to be DNA intercalators [93].
Simocyclinones have proved to be fascinating natural product compounds, both
in terms of their unexpected mode of action and their biosynthetic pathways.
Additional simocyclinones have been discovered from genomic-driven identifica-
tion using Streptomyces and Kitasatospora species [94]. These new simocyclinones
(D9, D10, and D11) inhibit DNA gyrase, but they show unexpectedly different bio-
synthetic gene cluster arrangements from simocyclinone D8. The availability of
several simocyclinones and the identification of their gene clusters raise the possi-
bility of carrying out engineering experiments to generate novel molecular entities
going forward.
Other natural products – aminocoumarins, simocyclinones, and cyclothiali-
dines are natural products produced by Streptomyces species. Actinomycetes in
general have proven to be rich sources of antibiotic compounds, so it is not a sur-
prise that natural products active against gyrase and topo IV have been found.
19 Non-quinolone Topoisomerase Inhibitors 605

Fig. 19.7 Natural product inhibitors of gyrase and topo IV

However, apart from the limited success of novobiocin, these have yet to be devel-
oped as clinically useful compounds. But, given their high degree of chemical diver-
sity and the great success of natural products directed at other targets, it is realistic
to expect that gyrase/topo IV-targeted compounds will be successful in the future.
There are a whole host of compounds that have been discovered that target gyrase/
topo IV and that may have clinical potential going forward. On account of space
limitations, only a few examples are given here to illustrate this type of work.
In 2007, a novel inhibitor, identified in an in vivo E. coli screen, demonstrated
inhibitory effects on cellular division [95]: N-benzyl-3-sulfonamidopyrrolidine,
then referred to as “534F6” and now as gyramide A (Fig. 19.7). The target of gyra-
mide A was shown to be DNA gyrase, reportedly with a unique binding site and
mechanism of inhibition [96]. Gyramides B and C were synthesized through modi-
fication of gyramide A. Using an E. coli strain having an inactive AcrAB-TolC mul-
tidrug efflux pump, gyramide A gave an MIC of 10 μM. The sequencing of
spontaneous gyramide A-resistant mutants revealed amino acid substitutions in both
GyrA and GyrB that clustered adjacent to the cleavage-religation site and were dis-
tinct from those that produce quinolone resistance [96]. Cross-resistance to quino-
lones was also absent in these gyramide A-resistant mutants. When paired with an
efflux pump inhibitor, gyramides A, B, and C demonstrated effective antibacterial
action against Gram-negative bacteria, as well as activity against some Gram-­
positive species [96].
In 2014, gyramide A was suggested to be a competitive inhibitor of ATP hydro-
lysis [97], in contrast to the earlier report, although this was later withdrawn [98]. It
is possible that gyramide A reduces the rate of ATP hydrolysis indirectly by perturb-
ing the binding to DNA. The 2014 paper also showed that gyramide A is a specific
inhibitor of gyrase, showing no activity against topo IV in vitro.
606 A. Maxwell et al.

Recently, through chemical modifications, gyramide analogs (D–F) have been

synthesized and their antibacterial characteristics explored. One drawback with
gyramides A, B, and C is that they are readily pumped out of the cell. Thus, to
achieve antibacterial activity, these compounds must be used in combination with
an efflux pump inhibitor. The new gyramide analogs demonstrate reduced sensitiv-
ity to efflux along with an increased inhibitory effect on gyrase (MICs in the
1 μg mL−1 range) and an extended spectrum of sensitive species [98]. These promis-
ing data, along with the unique binding site and no cross-resistance to ciprofloxacin
and novobiocin, suggest that this new class of gyrase inhibitor may have potential
as clinically useful antibiotics.
Naphthoquinones have been implicated for usage in the treatment of a variety
of human diseases [99, 100]. Extracts from Euclea natalensis (the “toothbrush
tree”), used in South African traditional medicine for various indications, were
found to contain several naphthoquinones, in particular diospyrin (Fig. 19.7) [101].
This compound, and related compounds, was found to inhibit DNA gyrase from E.
coli, S. aureus, and Mycobacterium tuberculosis [102]. Diospyrin seems to bind at
a novel, currently uncharacterized site in GyrB without being an ATPase inhibitor.
Although instability issues have hampered further development of this compound
(unpublished data), the existence of another potentially exploitable ligand-binding
site in gyrase is of potential interest.
Kibdelomycin is a natural product synthesized by Kibdelosporangium sp. strain,
MA7385, which was isolated from a Central African forest soil sample.
Kibdelomycin was discovered using an antisense-induced strain sensitivity (AISS)
profiling technique in S. aureus [103]. This method involves the reduction of expres-
sion of individual genes essential to growth using 245 S. aureus strains with induc-
ible antisense RNA. The AISS profile for kibdelomycin showed strong depletions in
growth in the strains containing antisense RNA to gyrB, parC and parE, and weak
depletions using the antisense gyrA strain. This was similar to the profile produced
by novobiocin, and much less so to the quinolone profile, although not incompara-
ble. Kibdelomycin has a complex chemical structure (comprehensively described in
ref. [104]), which was elucidated using 2D-NMR and mass spectrometry
techniques.
Kibdelomycin has demonstrated broad-spectrum potency against aerobic bacte-
ria that include the Gram-positive MRSA, Streptococcus pneumoniae, Enterococcus
faecium, and Enterococcus faecalis and the Gram-negative Moraxella catarrhalis,
Haemophilus influenza, and Acinetobacter baumannii but not E. coli (due to both
increased efflux and reduced membrane permeability) or Pseudomonas aeruginosa
(due to reduced membrane permeability) [105]. Unfortunately, the MIC of kibdelo-
mycin increased 256-fold in the presence of 50% human serum, raising concerns as
to its value in vivo as a systemic antibiotic. However, it has been shown to be a
potent inhibitor of Clostridium difficile growth in an in vivo hamster model, provid-
ing 100% protection against infection when dosed orally at 12.5–6.25 mg/kg, twice
a day for 4 days [106].
The presence of kibdelomycin has been shown in vitro to potently inhibit E. coli
gyrase supercoiling (IC50 0.06 μM) and both S. aureus gyrase supercoiling and topo
19 Non-quinolone Topoisomerase Inhibitors 607

IV decatenation (IC50 0.009 and 0.5 μM, respectively), but it only weakly inhibits
E. coli topo IV decatenation (IC50 29 μM). Also, producing an effect very similar to
novobiocin, kibdelomycin is a potent inhibitor of the ATPase activity of both E. coli
gyrase and topo IV (IC50 0.011 and 0.9 μM). When crystallized bound to both GyrB
and ParE, kibdelomycin exhibited a unique U-shaped binding mode having exten-
sive hydrophobic and polar interactions with surface residues of both proteins,
while the pyrrolamide moiety extended deep into the ATP-binding pocket. This
behavior is distinct from the binding of aminocoumarins and is consistent in the
lack of cross-resistance between the two [107].
Closthioamide, a member of the polythioamide class of DNA gyrase inhibitors
(Fig. 19.7), was the first secondary metabolite isolated from the obligate anaerobe
Clostridium cellulolyticum. While genome mining identified genes involved in
polyketide and peptide synthesis, no bioactive compounds were found until the
addition of aqueous soil extract to the culture led to the production of closthioamide
[108]. Closthioamide is symmetrical in structure with six thioamide groups flanked
on either side by a phenol group. It has demonstrated potent activity against Gram-­
positive strains including MRSA and VRE, giving MICs of 0.14 and 0.4 mg/L,
respectively. Only moderate inhibitory effects were found using wild-type E. coli
strains (MICs of 2.5–3.5 mg/L); however, activity could be increased dramatically
using the membrane permeability enhancer polymyxin B nonapeptide (PMBN) or
using a drug efflux pump-deficient strain with and without PMBN, to MICs of
0.625, 0.035, and 0.31 mg/L, respectively. These data clearly suggest that the outer
membrane and efflux pumps of Gram-negative bacteria are responsible for the
reduction of closthioamide’s efficacy [109].
Closthioamide’s mode of action is unlikely to involve cleavage-complex stabili-
zation, as very little linear DNA is detected when the agent is present in gyrase-­
DNA reaction mixtures. However, closthioamide did reduce the ATPase activity of
gyrase and topo IV by 80% and 60%, respectively. Although the compound also
inhibits the ATP-independent relaxation activity of gyrase, it is doubtful that it is a
competitive ATPase inhibitor. Rather, it is more likely to allosterically interfere with
ATP hydrolysis, inhibiting the enzyme using a novel binding mode, one that has
been likened to the mode of action diospyrin [109].
Using the soil-dwelling actinomycete, Amycolatopsis sp., amycolamicin
(Fig. 19.7) was isolated and found to have potent, broad-spectrum antibiotic activ-
ity. Its structure, determined using a combination of NMR spectroscopy, chemical
degradation, X-ray analysis, and functional group modification, is described in
detail in ref. [110]. The compound has shown promise against the Gram-positive
MRSA, VRE, and penicillin-resistant S. pneumoniae (all with MICs in the range of
0.25–1 μg/mL) as well as against the Gram-negative ampicillin-resistant and beta-­
lactamase-­positive amoxicillin-clavulanate-resistant strains of H. influenzae (MIC
0.5 and 2 μg/mL, respectively). The target was determined to be bacterial type II
topoisomerases, with amycolamicin inhibiting gyrase and topo IV with IC50s of
0.024 and 6.2 μg/mL, respectively.
608 A. Maxwell et al.

The binding region of amycolamicin was explored using an amycolamicin-­

resistant S. aureus mutant. Upon sequencing of both gyrA and gyrB genes, muta-
tions conferring resistance were found corresponding to substitutions in the B
subunit, involving Thr173 to Ile and Glu201 to Ala changes. Using known novobiocin-
and coumermycin-resistant mutations, it was found that some mutations, but not all,
affected the binding of amycolamicin, indicating the binding region to be in the
vicinity of the GHKL ATP-binding domain [110].
Other GyrB Inhibitors
The crystal structures of the N-terminal domain of GyrB, complexed with ATP ana-
logs [19, 64, 111–113] and with aminocoumarins and cyclothialidines [114–116],
have potentiated many drug-design programs using in silico methods, fragment
screening, and related approaches that have included work from a number of com-
panies, such as Cubist and F. Hoffmann-La Roche. The result has been a large num-
ber of publications describing novel compounds designed to bind to this site [117].
This type of approach has been referred to as bioisosterism [19]. While this is a
valid approach, none of these compounds have so far become a clinically useful
antibiotic. It is possible that a successful compound may emerge in the future, but it
is worth bearing in mind that this ATP-binding site shares similarities with that of a
number of other proteins, the GHKL ATPases [67]; thus, mammalian toxicity is
always a danger. Indeed, we noted above that it has been possible to engineer novo-
biocin, the archetypal GyrB ATPase inhibitor, such that it is more specific to the
human anticancer target Hsp90 [68, 118], and further modifications have led to
compounds that are MAPK pathway inhibitors [119]. In addition, this type of target-­
based modeling approach does not take into account bacterial permeability and
efflux issues. Nonetheless, target-based approaches can be successful (e.g., thio-
phenes, see above), and modeling may well lead to exploration of new chemical
space.
Another potential drawback of targeting the ATP-binding site of gyrase/topo IV
is that this mode of inhibition does not generally lead to cleavage-complex stabiliza-
tion, which is a key feature of the success of topoisomerase-targeted drugs, includ-
ing the quinolones. However, it has been shown that the ATP analog ADPNP
(5'-adenylyl-β,γ-imidodiphosphate) can lead to cleavage-complex stabilization by
gyrase [120], and the anticancer compound ICRF-193 stabilizes cleavage com-
plexes with eukaryotic topo II through binding at the ATPase site [121]. Therefore,
the possibility of compounds binding at or near the ATPase site and stabilizing the
cleavage complex should not be disregarded.
Since this area has been extensively reviewed recently [19, 122], just a few illus-
trative examples are given here. In one, F. Hoffmann-La Roche embarked on a
screen using low-molecular-weight (<300 Da) entities (“needle screening”) coupled
with a high-throughput gyrase ATPase assay and biophysical validation, followed
by a 3D-guided optimization process [123]. Selected “hit” compounds were crystal-
lized with the N-terminal sub-domain (24 kDa) of S. aureus GyrB to verify the
binding mode. Seven new classes of inhibitor were found, including one compound
that was ten times more potent than novobiocin [123].
19 Non-quinolone Topoisomerase Inhibitors 609

In other examples, Cubist utilized a fragment-based screening method using

NMR to assess the binding of >5000 diverse small chemical entities to the N-terminal
sub-domain of S. aureus GyrB [124]. Compounds were further evaluated using
X-ray crystallography and IC50 determination. Pyrazolopyridones were developed
using this approach [124]; they were subjected to further optimization using medici-
nal chemistry guided by structure-based drug design [125]. Some of the compounds
that emerged showed activity against both gyrase and topo IV, plus antibacterial
activity against S. aureus [125]. Furthermore, using de novo design based on the
GyrB ATPase site, Cubist discovered azaindole ureas [126], compounds that show
in vitro activity against gyrase and Gram-positive bacteria, including fluoroquinolone-­
resistant MRSA. AstraZeneca used an NMR screening approach followed by design
and synthesis to develop pyrrolamide inhibitors [127]; efficacy of a representative
pyrrolamide was demonstrated against Streptococcus pneumoniae in a mouse lung
infection model.
In terms of academic efforts, a number of laboratories have utilized these types
of approaches to discover new gyrase/topo IV inhibitors. For example, Sriram and
coworkers (Hyderabad, India) have published extensively on the use of structure-­
guided approaches to find new molecules, particularly to target M. tuberculosis
gyrase [128, 129]. This work has yielded a number of molecular scaffolds, e.g.,
quinolines [130] and phenylthiophene-carboxamide derivatives [131]. Work from
Kikelj and others (Ljubljana, Slovenia), using ligand-based and structure-based
approaches, has generated a number of different inhibitors targeted to the GyrB
ATPase site; these inhibitors include benzothiazole and oxadiazole compounds
[132, 133]. In other work, using docking simulations with the structure of the S.
aureus GyrB ATPase domain, a series of dihydropyrazole compounds have been
synthesized and evaluated [134]. Whether these or any other compounds developed
from these approaches can be successfully developed into clinically useful antibiot-
ics remains to be seen.

19.6 Concluding Remarks

A clear message that emerges from this review is that gyrase and topo IV are excel-
lent targets for antibacterial chemotherapy. The success of the fluoroquinolones
attests to the value of these targets. However, what is clearly needed are new agents,
ideally cleavage complex-stabilizing compounds, that can replace the quinolones.
The inhibitors described in this chapter have been discovered and developed using
a variety of approaches, including screening chemical libraries, following natural
product leads, and fragment-based and in silico approaches; they include both
target-­led and phenotypic-led methodologies. Although arguments can be advanced
that favor one or another of these approaches, it is likely that we need to retain a
diverse range of approaches to discover the types of agent we seek. Increased chem-
ical diversity, perhaps through investigating novel sources of natural products, is
likely to be a key component to success going forward.
610 A. Maxwell et al.

While the approaches to new compound discovery are an important issue for
discussion, a more challenging question is who will carry out this work? The vast
majority of antibiotics available for clinical use have been developed and produced
by large pharma companies. While academics and small companies may have dis-
covered and researched compounds, it is only large pharma that has the know-how
and resources to bring them to market. However, in the current economic and politi-
cal climate, large pharma is pulling out of antibiotic R & D, mainly on account of
profitability issues, with a consequent reduction in effort in some cases and com-
plete withdrawal in others [135, 136]. We are now faced with potential significant
shortcomings in the discovery pipeline [137]. This is leading to the perilous situa-
tion of increasing antimicrobial-resistant bacterial infections and a paucity of new
agents to treat them [138]. It is probably essential that governments, ideally working
in cooperation, confront this challenge and provide the necessary resources and
incentives to sustain the antibiotic discovery effort. It is likely that this will be
increasingly carried out by the academic and SME sectors, resourced through public
financing [139, 140]; it is clear that governments need to take action to address this
crisis.
Government action would be starting with a solid base, as some pharma compa-
nies are still developing novel quinolones, particularly for niche markets (e.g., dela-
floxacin produced by Melinta Therapeutics [141]). Moreover, there are many
examples of non-quinolone agents being developed, some of which have been
described in this chapter. Among the active companies are AstraZeneca (e.g., spiro-
pyrimidinetriones [142]), Cubist (pyrazolopyridones [125]), GSK (NBTIs and thio-
phenes [31, 48]), Pfizer (quinazolinediones [35]), Sanofi (IPYs [38]), and Vertex
(benzimidazoles [143]). It is to be hoped that mechanisms will be found to sustain
these efforts and ensure that the considerable expertise in this area of investigation
is not lost.
Major Points
• Quinolones are highly successful antibiotics, but resistance is a serious
problem.
• DNA topoisomerases (particularly gyrase and topo IV) are important targets for
antimicrobial chemotherapy that should continue to be exploited.
• Cleavage-complex stabilization is an excellent mode of action for antibiotics; it
is possible to find new, non-quinolone, compounds that work via this
mechanism.
• Catalytic inhibitors of topoisomerases can, in principle, be developed as antibiot-
ics of the future.
• It is important to sustain a variety of approaches for discovering new
antibiotics.
• Big pharma companies cannot necessarily be relied upon to develop new antibi-
otics going forward; other ways of developing new drugs need to be explored.
19 Non-quinolone Topoisomerase Inhibitors 611

References

1. Boucher HW, Talbot GH, Bradley JS, Edwards JE, Gilbert D, Rice LB, et al. Bad bugs, no
drugs: no ESKAPE! An update from the Infectious Diseases Society of America. Clin Infect
Dis. 2009;48(1):1–12.
2. Bush K, Courvalin P, Dantas G, Davies J, Eisenstein B, Huovinen P, et al. Tackling antibiotic
resistance. Nat Rev Microbiol. 2011;9(12):894–6.
3. Lewis K. Antibiotics: Recover the lost art of drug discovery. Nature. 2012;485(7399):439–40.
4. Walsh CT, Wencewicz TA. Prospects for new antibiotics: a molecule-centered perspective.
J Antibiot (Tokyo). 2014;67(1):7–22.
5. Linder JA, Huang ES, Steinman MA, Gonzales R, Stafford RS. Fluoroquinolone prescribing
in the United States: 1995 to 2002. Am J Med. 2005;118(3):259–68.
6. Pitiriga V, Vrioni G, Saroglou G, Tsakris A. The impact of antibiotic stewardship programs in
combating quinolone resistance: a systematic review and recommendations for more efficient
interventions. Adv Ther. 2017;34(4):854–65.
7. Hooper DC. Mechanisms of quinolone resistance. In: Hooper DC, Rubinstein E, editors.
Quinolone antimicrobial agents. 3rd ed. Washington, DC: ASM Press; 2003. p. 41–67.
8. Redgrave LS, Sutton SB, Webber MA, Piddock LJ. Fluoroquinolone resistance: mechanisms,
impact on bacteria, and role in evolutionary success. Trends Microbiol. 2014;22(8):438–45.
9. WHO. Critically important antimicrobials for human medicine – 5 rev. June 2017 Ed.
Geneva: World Health Organisation – WHO Advisory Group on Integrated Surveillance of
Antimicrobial Resistance (AGISAR); 2017.
10. Bush NG, Evans-Roberts K, Maxwell A. DNA topoisomerases. EcoSal Plus. 2015;6(2).
11. Vos SM, Tretter EM, Schmidt BH, Berger JM. All tangled up: how cells direct, manage and
exploit topoisomerase function. Nat Rev Mol Cell Biol. 2011;12(12):827–41.
12. Wang JC. Cellular roles of DNA topoisomerases: a molecular perspective. Nat Rev Mol Cell
Biol. 2002;3(6):430–40.
13. Forterre P, Gribaldo S, Gadelle D, Serre MC. Origin and evolution of DNA topoisomerases.
Biochimie. 2007;89(4):427–46.
14. Wang JC. DNA topoisomerases. Annu Rev Biochem. 1996;65:635–92.
15. Gadelle D, Filee J, Buhler C, Forterre P. Phylogenomics of type II DNA topoisomerases.
BioEssays. 2003;25(3):232–42.
16. Aravind L, Iyer LM, Wellems TE, Miller LH. Plasmodium biology: genomic gleanings. Cell.
2003;115(7):771–85.
17. Sugimoto-Shirasu K, Stacey NJ, Corsar J, Roberts K, McCann MC. DNA topoisomerase VI
is essential for endoreduplication in Arabidopsis. Curr Biol. 2002;12(20):1782–6.
18. Wang JC. Moving one DNA double helix through another by a type II DNA topoisomerase:
the story of a simple molecular machine. Q Rev Biophys. 1998;31(2):107–44.
19. Mayer C, Janin YL. Non-quinolone inhibitors of bacterial type IIA topoisomerases: a feat of
bioisosterism. Chem Rev. 2014;114(4):2313–42.
20. Nagaraja V, Godbole AA, Henderson SR, Maxwell A. DNA topoisomerase I and DNA gyrase
as targets for TB therapy. Drug Discov Today. 2017;22(3):510–8.
21. Collin F, Karkare S, Maxwell A. Exploiting bacterial DNA gyrase as a drug target: current
state and perspectives. Appl Microbiol Biotechnol. 2011;92(3):479–97.
22. Pommier P, Capranico G, Orr A, Kohn KW. Distribution of topoisomerase II cleavage sites in
simian virus 40 DNA and the effects of drugs. J Mol Biol. 1991;222:909–24.
23. De Souza MV. New fluoroquinolones: a class of potent antibiotics. Mini Rev Med Chem.
2005;5(11):1009–17.
24. Oliphant CM, Green GM. Quinolones: a comprehensive review. Am Fam Physician.
2002;65(3):455–64.
612 A. Maxwell et al.

25. Tse-Dinh YC. Bacterial topoisomerase I as a target for discovery of antibacterial compounds.
Nucleic Acids Res. 2009;37(3):731–7.
26. Sandhaus S, Annamalai T, Welmaker G, Houghten RA, Paz C, Garcia PK, et al. Small-­
molecule inhibitors targeting topoisomerase I as novel antituberculosis agents. Antimicrob
Agents Chemother. 2016;60(7):4028–36.
27. Cheng B, Cao S, Vasquez V, Annamalai T, Tamayo-Castillo G, Clardy J, et al. Identification
of anziaic acid, a lichen depside from Hypotrachyna sp., as a new topoisomerase poison
inhibitor. PLoS One. 2013;8(4):e60770.
28. Aldred KJ, Kerns RJ, Osheroff N. Mechanism of quinolone action and resistance.
Biochemistry. 2014;53(10):1565–74.
29. Drlica K, Hiasa H, Kerns R, Malik M, Mustaev A, Zhao X. Quinolones: action and resistance
updated. Curr Top Med Chem. 2009;9(11):981–98.
30. Pommier Y, Leo E, Zhang H, Marchand C. DNA topoisomerases and their poisoning by anti-
cancer and antibacterial drugs. Chem Biol. 2010;17(5):421–33.
31. Bax BD, Chan PF, Eggleston DS, Fosberry A, Gentry DR, Gorrec F, et al. Type IIA topoisom-
erase inhibition by a new class of antibacterial agents. Nature. 2010;466(7309):935–40.
32. Laponogov I, Sohi MK, Veselkov DA, Pan X-S, Sawhney R, Thompson AW, et al. Structural
insight into the quinolone-DNA cleavage complex of type IIA topoisomerases. Nat Struct
Mol Biol. 2009;16(6):667–9.
33. Wohlkonig A, Chan PF, Fosberry AP, Homes P, Huang J, Kranz M, et al. Structural basis of
quinolone inhibition of type IIA topoisomerases and target-mediated resistance. Nat Struct
Mol Biol. 2010;17(9):1152–3.
34. Ellsworth EL, Tran TP, Showalter HD, Sanchez JP, Watson BM, Stier MA, et al.
3-­aminoquinazolinediones as a new class of antibacterial agents demonstrating excellent
antibacterial activity against wild-type and multidrug resistant organisms. J Med Chem.
2006;49(22):6435–8.
35. Huband MD, Cohen MA, Zurack M, Hanna DL, Skerlos LA, Sulavik MC, et al. In vitro
and in vivo activities of PD 0305970 and PD 0326448, new bacterial gyrase/topoisomerase
inhibitors with potent antibacterial activities versus multidrug-resistant gram-positive and
fastidious organism groups. Antimicrob Agents Chemother. 2007;51(4):1191–201.
36. Tran TP, Ellsworth EL, Sanchez JP, Watson BM, Stier MA, Showalter HD, et al. Structure-­
activity relationships of 3-aminoquinazolinediones, a new class of bacterial type-2 topoisom-
erase (DNA gyrase and topo IV) inhibitors. Bioorg Med Chem Lett. 2007;17(5):1312–20.
37. Germe T, Voros J, Jeannot F, Tailler T, Stavenger RA, Bacqué E, et al. A new class of antibac-
terials, the imidizopyrazinones, reveal structural transitions involved in DNA gyrase poison-
ing and mechanisms of resistance. Nucleic Acids Res. (in press).
38. Jeannot F, Taillier T, Despeyroux P, Renard S, Rey A, Mourez M, et al. Imidazopyrazinones
(IPYs): novel non-quinolone bacterial topoisomerase inhibitors showing partial cross-­
resistance with quinolones. (in press).
39. Aldred KJ, McPherson SA, Wang P, Kerns RJ, Graves DE, Turnbough CL Jr, et al. Drug
interactions with Bacillus anthracis topoisomerase IV: biochemical basis for quinolone action
and resistance. Biochemistry. 2012;51(1):370–81.
40. Laponogov I, Pan XS, Veselkov DA, McAuley KE, Fisher LM, Sanderson MR. Structural
basis of gate-DNA breakage and resealing by type II topoisomerases. PLoS One.
2010;5(6):e11338.
41. Aldred KJ, Schwanz HA, Li G, McPherson SA, Turnbough CL Jr, Kerns RJ, et al.
Overcoming target-mediated quinolone resistance in topoisomerase IV by introducing metal-
ion-­independent drug-enzyme interactions. ACS Chem Biol. 2013;8(12):2660–8.
42. German N, Malik M, Rosen JD, Drlica K, Kerns RJ. Use of gyrase resistance mutants to
guide selection of 8-methoxy-quinazoline-2,4-diones. Antimicrob Agents Chemother.
2008;52(11):3915–21.
19 Non-quinolone Topoisomerase Inhibitors 613

43. Oppegard LM, Streck KR, Rosen JD, Schwanz HA, Drlica K, Kerns RJ, et al. Comparison of
in vitro activities of fluoroquinolone-like 2,4- and 1,3-diones. Antimicrob Agents Chemother.
2010;54(7):3011–4.
44. Mustaev A, Malik M, Zhao X, Kurepina N, Luan G, Oppegard LM, et al. Fluoroquinolone-­
gyrase-­DNA complexes: two modes of drug binding. J Biol Chem. 2014.
45. Drlica K, Mustaev A, Towle TR, Luan G, Kerns RJ, Berger JM. Bypassing fluoroquinolone
resistance with quinazolinediones: studies of drug-gyrase-DNA complexes having implica-
tions for drug design. ACS Chem Biol. 2014;9(12):2895–904.
46. Chan PF, Srikannathasan V, Huang J, Cui H, Fosberry AP, Gu M, et al. Structural basis of
DNA gyrase inhibition by antibacterial QPT-1, anticancer drug etoposide and moxifloxacin.
Nat Commun. 2015;6:10048.
47. Biedenbach DJ, Bouchillon SK, Hackel M, Miller LA, Scangarella-Oman NE, Jakielaszek
C, et al. In vitro activity of gepotidacin, a novel triazaacenaphthylene bacterial topoisomerase
inhibitor, against a broad spectrum of bacterial pathogens. Antimicrob Agents Chemother.
2016;60(3):1918–23.
48. Chan PF, Germe T, Bax BD, Huang J, Thalji RK, Bacque E, et al. Thiophene antibacterials
that allosterically stabilize DNA-cleavage complexes with DNA gyrase. Proc Natl Acad Sci
U S A. 2017;114(22):E4492–E500.
49. Aldred KJ, Breland EJ, McPherson SA, Turnbough CL Jr, Kerns RJ, Osheroff N. Bacillus
anthracis GrlAV96A topoisomerase IV, a quinolone resistance mutation that does not affect
the water-metal ion bridge. Antimicrob Agents Chemother. 2014;58(12):7182–7.
50. Parks WM, Bottrill AR, Pierrat OA, Durrant MC, Maxwell A. The action of the bacterial
toxin, microcin B17, on DNA gyrase. Biochimie. 2007;89:500–7.
51. Heddle JG, Blance SJ, Zamble DB, Hollfelder F, Miller DA, Wentzell LM, et al. The antibi-
otic microcin B17 is a DNA gyrase poison: characterisation of the mode of inhibition. J Mol
Biol. 2001;307(5):1223–34.
52. Vizan JL, Hernandez-Chico C, del Castillo I, Moreno F. The peptide antibiotic microcin
B17 induces double-strand cleavage of DNA mediated by E. coli DNA gyrase. EMBO
J. 1991;10(2):467–76.
53. del Castillo FJ, del Castillo I, Moreno F. Construction and characterization of mutations at
codon 751 of the Escherichia coli gyrB gene that confer resistance to the antimicrobial pep-
tide microcin B17 and alter the activity of DNA gyrase. J Bacteriol. 2001;183(6):2137–40.
54. Thompson RE, Collin F, Maxwell A, Jolliffe KA, Payne RJ. Synthesis of full length
and truncated microcin B17 analogues as DNA gyrase poisons. Org Biomol Chem.
2014;12(10):1570–8.
55. Collin F, Thompson RE, Jolliffe KA, Payne RJ, Maxwell A. Fragments of the bacterial toxin
microcin b17 as gyrase poisons. PLoS One. 2013;8(4):e61459.
56. Thompson RE, Jolliffe KA, Payne RJ. Total synthesis of Microcin B17 via a fragment con-
densation approach. Org Lett. 2011;13(4):680–3.
57. Videnov G, Kaiser D, Brooks M, Jung G. Synthesis of the DNA gyrase inhibitor microcin
B17, a 43-peptie antibiotic with eight heterocycles in its backbone. Agnew Chem Int Ed Engl.
1996;35(13/14):1506–8.
58. Trovatti E, Cotrim CA, Garrido SS, Barros RS, Marchetto R. Peptides based on CcdB protein
as novel inhibitors of bacterial topoisomerases. Bioorg Med Chem Lett. 2008;18(23):6161–4.
59. Walsh C, Wencewicz T. Antibiotics challenges mechanisms opportunites. Washington, DC:
ASM Press; 2016. 477 p.
60. Heide L. The aminocoumarins: biosynthesis and biology. Nat Prod Rep. 2009;26(10):1241–50.
61. Maxwell A, Lawson DM. The ATP-binding site of type II topoisomerases as a target for
antibacterial drugs. Curr Top Med Chem. 2003;3(1):283–303.
62. Mizuuchi K, O’Dea MH, Gellert M. DNA gyrase: subunit structure and ATPase activity of
the purified enzyme. Proc Natl Acad Sci U S A. 1978;75:5960–3.
614 A. Maxwell et al.

63. Kato J, Suzuki H, Ikeda H. Purification and characterization of DNA topoisomerase IV in

Escherichia coli. J Biol Chem. 1992;267(36):25676–84.
64. Wigley DB, Davies GJ, Dodson EJ, Maxwell A, Dodson G. Crystal structure of an N-terminal
fragment of the DNA gyrase B protein. Nature. 1991;351(6328):624–9.
65. Lewis RJ, Singh OM, Smith CV, Skarzynski T, Maxwell A, Wonacott AJ, et al. The nature
of inhibition of DNA gyrase by the coumarins and the cyclothialidines revealed by X-ray
crystallography. EMBO J. 1996;15(6):1412–20.
66. Kampranis SC, Gormley NA, Tranter R, Orphanides G, Maxwell A. Probing the binding of
coumarins and cyclothialidines to DNA gyrase. Biochemistry. 1999;38(7):1967–76.
67. Dutta R, Inouye M. GHKL, an emergent ATPase/kinase superfamily. Trends Biochem Sci.
2000;25(1):24–8.
68. Burlison JA, Neckers L, Smith AB, Maxwell A, Blagg BS. Novobiocin: redesigning a DNA
gyrase inhibitor for selective inhibition of Hsp90. J Am Chem Soc. 2006;128(48):15529–36.
69. Heide L. Genetic engineering of antibiotic biosynthesis for the generation of new aminocou-
marins. Biotechnol Adv. 2009;27(6):1006–14.
70. Heide L. New aminocoumarin antibiotics as gyrase inhibitors. Int J Med Microbiol.
2014;304(1):31–6.
71. Eustaquio AS, Gust B, Luft T, Li SM, Chater KF, Heide L. Clorobiocin biosynthesis in
Streptomyces. Identification of the halogenase and generation of structural analogs. Chem
Biol. 2003;10(3):279–88.
72. Anderle C, Stieger M, Burrell M, Reinelt S, Maxwell A, Page M, et al. Biological activi-
ties of novel gyrase inhibitors of the aminocoumarin class. Antimicrob Agents Chemother.
2008;52(6):1982–90.
73. Alt S, Burkard N, Kulik A, Grond S, Heide L. An artificial pathway to 3,4-dihydroxybenzoic
acid allows generation of new aminocoumarin antibiotic recognized by catechol transporters
of E. coli. Chem Biol. 2011;18(3):304–13.
74. Kamiyama T, Shimma N, Ohtsuka T, Nakayama N, Itezono Y, Nakada N, et al. Cyclothialidine,
a novel DNA gyrase inhibitor. II. Isolation, characterization and structure elucidation.
J Antibiot (Tokyo). 1994;47(1):37–45.
75. Watanabe J, Nakada N, Sawairi S, Shimada H, Ohshima S, Kamiyama T, et al. Cyclothialidine,
a novel DNA gyrase inhibitor. I. Screening, taxonomy, fermentation and biological activity.
J Antibiot (Tokyo). 1994;47(1):32–6.
76. Goetschi E, Angehrn P, Gmuender H, Hebeisen P, Link H, Masciadri R, et al. Cyclothialidine
and its congeners: a new class of DNA gyrase inhibitors. Pharmacol Ther. 1993;60(2):367–80.
77. Angehrn P, Buchmann S, Funk C, Goetschi E, Gmuender H, Hebeisen P, et al. New anti-
bacterial agents derived from the DNA gyrase inhibitor cyclothialidine. J Med Chem.
2004;47(6):1487–513.
78. Angehrn P, Goetschi E, Gmuender H, Hebeisen P, Hennig M, Kuhn B, et al. A new DNA
gyrase inhibitor subclass of the cyclothialidine family based on a bicyclic dilactam-lactone
scaffold. Synthesis and antibacterial properties. J Med Chem. 2011;54(7):2207–24.
79. Theobald U, Schimana J, Fiedler HP. Microbial growth and production kinetics of
Streptomyces antibioticus Tu 6040. Antonie Van Leeuwenhoek. 2000;78(3–4):307–13.
80. Trefzer A, Pelzer S, Schimana J, Stockert S, Bihlmaier C, Fiedler HP, et al. Biosynthetic gene
cluster of simocyclinone, a natural multihybrid antibiotic. Antimicrob Agents Chemother.
2002;46(5):1174–82.
81. Galm U, Schimana J, Fiedler HP, Schmidt J, Li SM, Heide L. Cloning and analysis of
the simocyclinone biosynthetic gene cluster of Streptomyces antibioticus Tu 6040. Arch
Microbiol. 2002;178(2):102–14.
82. Schimana J, Fiedler HP, Groth I, Sussmuth R, Beil W, Walker M, et al. Simocyclinones,
novel cytostatic angucyclinone antibiotics produced by Streptomyces antibioticus Tu
6040. I. Taxonomy, fermentation, isolation and biological activities. J Antibiot (Tokyo).
2000;53(8):779–87.
19 Non-quinolone Topoisomerase Inhibitors 615

83. Schimana J, Walker M, Zeeck A, Fiedler P. Simocyclinones: diversity of metabolites is

dependent on fermentation conditions. J Ind Microbiol Biotechnol. 2001;27(3):144–8.
84. Flatman RH, Howells AJ, Heide L, Fiedler HP, Maxwell A. Simocyclinone D8, an
inhibitor of DNA gyrase with a novel mode of action. Antimicrob Agents Chemother.
2005;49(3):1093–100.
85. Edwards MJ, Flatman RH, Mitchenall LA, Stevenson CE, Le TB, Clarke TA, et al. A crystal
structure of the bifunctional antibiotic simocyclinone D8, bound to DNA gyrase. Science.
2009;326(5958):1415–8.
86. Edwards MJ, Williams MA, Maxwell A, McKay AR. Mass spectrometry reveals that the
antibiotic simocyclinone D8 binds to DNA gyrase in a “bent-over” conformation: evidence
of positive cooperativity in binding. Biochemistry. 2011;50(17):3432–40.
87. Hearnshaw SJ, Edwards MJ, Stevenson CE, Lawson DM, Maxwell A. A new crystal struc-
ture of the bifunctional antibiotic simocyclinone d8 bound to DNA gyrase gives fresh insight
into the mechanism of inhibition. J Mol Biol. 2014;426(10):2023–33.
88. Sissi C, Vazquez E, Chemello A, Mitchenall LA, Maxwell A, Palumbo M. Mapping simocycli-
none D8 interaction with DNA gyrase: evidence for a new binding site on GyrB. Antimicrob
Agents Chemother. 2010;54(1):213–20.
89. Richter SN, Frasson I, Palumbo M, Sissi C, Palu G. Simocyclinone D8 turns on against gram-­
negative bacteria in a clinical setting. Bioorg Med Chem Lett. 2010;20(3):1202–4.
90. Oppegard LM, Nguyen T, Ellis KC, Hiasa H. Inhibition of human topoisomerases I and II by
simocyclinone D8. J Nat Prod. 2012;75(8):1485–9.
91. Sadiq AA, Patel MR, Jacobson BA, Escobedo M, Ellis K, Oppegard LM, et al. Anti-­
proliferative effects of simocyclinone D8 (SD8), a novel catalytic inhibitor of topoisomerase
II. Investig New Drugs. 2009.
92. Austin MJ, Hearnshaw SJ, Mitchenall LA, McDermott PJ, Howell LA, Maxwell A, et al. A
natural product inspired fragment-based approach towards the development of novel anti-­
bacterial agents. MedChemComm. 2016;7(7):1387–91.
93. Verghese J, Nguyen T, Oppegard LM, Seivert LM, Hiasa H, Ellis KC. Flavone-based ana-
logues inspired by the natural product simocyclinone D8 as DNA gyrase inhibitors. Bioorg
Med Chem Lett. 2013;23(21):5874–7.
94. Bilyk O, Brotz E, Tokovenko B, Bechthold A, Paululat T, Luzhetskyy A. New simocyclinones:
surprising evolutionary and biosynthetic insights. ACS Chem Biol. 2016;11(1):241–50.
95. Mukherjee S, Robinson CA, Howe AG, Mazor T, Wood PA, Urgaonkar S, et al. N-Benzyl-­
3-sulfonamidopyrrolidines as novel inhibitors of cell division in E. coli. Bioorg Med Chem
Lett. 2007;17(23):6651–5.
96. Foss MH, Hurley KA, Sorto N, Lackner LL, Thornton KM, Shaw JT, et al. N-Benzyl-3-­
sulfonamidopyrrolidines are a new class of bacterial DNA gyrase inhibitors. ACS Med Chem
Lett. 2011;2(4):289–92.
97. Rajendram M, Hurley KA, Foss MH, Thornton KM, Moore JT, Shaw JT, et al. Gyramides
prevent bacterial growth by inhibiting DNA gyrase and altering chromosome topology. ACS
Chem Biol. 2014;9(6):1312–9.
98. Hurley KA, Santos TMA, Fensterwald MR, Rajendran M, Moore JT, Balmond EI, et al.
Targeting quinolone- and aminocoumarin-resistant bacteria with new gyramide analogs that
inhibit DNA gyrase. MedChemComm. 2017;8(5):942–51.
99. Pinto AV, de Castro SL. The trypanocidal activity of naphthoquinones: a review. Molecules.
2009;14(11):4570–90.
100. van der Kooy F, Meyer JJM, Lall N. Antimycobacterial activity and possible mode of action
of newly isolated neodiospyrin and other naphthoquinones from Euclea natalensis. S Afr
J Bot. 2006;72(3):349–52.
101. Lall N, Meyer JJM. Inhibition of drug-sensitive and drug-resistant strains of Mycobacterium
tuberculosis by diospyrin, isolated from Euclea natalensis. J Ethnopharmacol.
2001;78(2–3):213–6.
616 A. Maxwell et al.

102. Karkare S, Chung TT, Collin F, Mitchenall LA, McKay AR, Greive SJ, et al. The naphthoqui-
none diospyrin is an inhibitor of DNA gyrase with a novel mechanism of action. J Biol Chem.
2013;288(7):5149–56.
103. Phillips JW, Goetz MA, Smith SK, Zink DL, Polishook J, Onishi R, et al. Discovery of kibde-
lomycin, a potent new class of bacterial type II topoisomerase inhibitor by chemical-genetic
profiling in Staphylococcus aureus. Chem Biol. 2011;18(8):955–65.
104. Singh SB. Discovery and development of kibdelomycin, a new class of broad-spectrum anti-
biotics targeting the clinically proven bacterial type II topoisomerase. Bioorg Med Chem.
2016;24(24):6291–7.
105. Singh SB, Dayananth P, Balibar CJ, Garlisi CG, Lu J, Kishii R, et al. Kibdelomycin is a
bactericidal broad-spectrum aerobic antibacterial agent. Antimicrob Agents Chemother.
2015;59(6):3474–81.
106. Miesel L, Hecht DW, Osmolski JR, Gerding D, Flattery A, Li F, et al. Kibdelomycin is a
potent and selective agent against toxigenic Clostridium difficile. Antimicrob Agents
Chemother. 2014;58(4):2387–92.
107. Lu J, Patel S, Sharma N, Soisson SM, Kishii R, Takei M, et al. Structures of kibdelomycin
bound to Staphylococcus aureus GyrB and ParE showed a novel U-shaped binding mode.
ACS Chem Biol. 2014;9(9):2023–31.
108. Lincke T, Behnken S, Ishida K, Roth M, Hertweck C. Closthioamide: an unprecedented
polythioamide antibiotic from the strictly anaerobic bacterium Clostridium cellulolyticum.
Angew Chem Int Ed Engl. 2010;49(11):2011–3.
109. Chiriac AI, Kloss F, Kramer J, Vuong C, Hertweck C, Sahl HG. Mode of action of clos-
thioamide: the first member of the polythioamide class of bacterial DNA gyrase inhibitors.
J Antimicrob Chemother. 2015;70(9):2576–88.
110. Sawa R, Takahashi Y, Hashizume H, Sasaki K, Ishizaki Y, Umekita M, et al. Amycolamicin:
a novel broad-spectrum antibiotic inhibiting bacterial topoisomerase. Chemistry.
2012;18(49):15772–81.
111. Agrawal A, Roue M, Spitzfaden C, Petrella S, Aubry A, Hann M, et al. Mycobacterium
tuberculosis DNA gyrase ATPase domain structures suggest a dissociative mechanism that
explains how ATP hydrolysis is coupled to domain motion. Biochem J. 2013;456(2):263–73.
112. Hearnshaw SJ, Chung TT, Stevenson CE, Maxwell A, Lawson DM. The role of monova-
lent cations in the ATPase reaction of DNA gyrase. Acta Crystallogr D Biol Crystallogr.
2015;71(Pt 4):996–1005.
113. Stanger FV, Dehio C, Schirmer T. Structure of the N-terminal gyrase B fragment in com-
plex with ADPPi reveals rigid-body motion induced by ATP hydrolysis. PLoS One.
2014;9(9):e107289.
114. Lewis RJ, Singh OM, Smith CV, Maxwell A, Skarzynski T, Wonacott AJ, et al. Crystallization
of inhibitor complexes of an N-terminal 24 kDa fragment of the DNA gyrase B protein. J Mol
Biol. 1994;241(1):128–30.
115. Tsai FT, Singh OM, Skarzynski T, Wonacott AJ, Weston S, Tucker A, et al. The high-­
resolution crystal structure of a 24-kDa gyrase B fragment from E. coli complexed with one
of the most potent coumarin inhibitors, clorobiocin. Proteins. 1997;28(1):41–52.
116. Bellon S, Parsons JD, Wei Y, Hayakawa K, Swenson LL, Charifson PS, et al. Crystal struc-
tures of Escherichia coli topoisomerase IV ParE subunit (24 and 43 kilodaltons): a single resi-
due dictates differences in novobiocin potency against topoisomerase IV and DNA gyrase.
Antimicrob Agents Chemother. 2004;48(5):1856–64.
117. Oblak M, Kotnik M, Solmajer T. Discovery and development of ATPase inhibitors of DNA
gyrase as antibacterial agents. Curr Med Chem. 2007;14(19):2033–47.
118. Garg G, Zhao H, Blagg BSJ. Design, synthesis, and biological evaluation of ring-constrained
Novobiocin analogues as Hsp90 C-terminal inhibitors. ACS Med Chem Lett. 2014.
119. Hall JA, Seedarala S, Zhao H, Garg G, Ghosh S, Blagg BS. Novobiocin analogues that inhibit
the MAPK pathway. J Med Chem. 2016;59(3):925–33.
19 Non-quinolone Topoisomerase Inhibitors 617

120. Kampranis SC, Bates AD, Maxwell A. A model for the mechanism of strand passage by DNA
gyrase. Proc Natl Acad Sci U S A. 1999;96(15):8414–9.
121. Morris SK, Baird CL, Lindsley JE. Steady-state and rapid kinetic analysis of topoisomerase
II trapped as the closed-clamp intermediate by ICRF-193. J Biol Chem. 2000;275(4):2613–8.
122. Ehmann DE, Lahiri SD. Novel compounds targeting bacterial DNA topoisomerase/DNA
gyrase. Curr Opin Pharmacol. 2014;18:76–83.
123. Boehm HJ, Boehringer M, Bur D, Gmuender H, Huber W, Klaus W, et al. Novel inhibitors
of DNA gyrase: 3D structure based biased needle screening, hit validation by biophysical
methods, and 3D guided optimization. A promising alternative to random screening. J Med
Chem. 2000;43(14):2664–74.
124. Mesleh MF, Cross JB, Zhang J, Kahmann J, Andersen OA, Barker J, et al. Fragment-based
discovery of DNA gyrase inhibitors targeting the ATPase subunit of GyrB. Bioorg Med Chem
Lett. 2016;26(4):1314–8.
125. Cross JB, Zhang J, Yang Q, Mesleh MF, Romero JA, Wang B, et al. Discovery of pyrazo-
lopyridones as a novel class of gyrase B inhibitors using structure guided design. ACS Med
Chem Lett. 2016;7(4):374–8.
126. Zhang J, Yang Q, Cross JB, Romero JA, Poutsiaka KM, Epie F, et al. Discovery of Azaindole
Ureas as a novel class of bacterial gyrase B inhibitors. J Med Chem. 2015;58(21):8503–12.
127. Eakin AE, Green O, Hales N, Walkup GK, Bist S, Singh A, et al. Pyrrolamide DNA gyrase
inhibitors: fragment-based nuclear magnetic resonance screening to identify antibacterial
agents. Antimicrob Agents Chemother. 2012;56(3):1240–6.
128. Jeankumar VU, Renuka J, Kotagiri S, Saxena S, Kakan SS, Sridevi JP, et al. Gyrase ATPase
domain as an antitubercular drug discovery platform: structure-based design and lead optimi-
zation of nitrothiazolyl carboxamide analogues. ChemMedChem. 2014;9(8):1850–9.
129. Jeankumar VU, Saxena S, Vats R, Reshma RS, Janupally R, Kulkarni P, et al. Structure-guided
discovery of antitubercular agents that target the gyrase ATPase domain. ChemMedChem.
2016;11(5):539–48.
130. Jeankumar VU, Reshma RS, Vats R, Janupally R, Saxena S, Yogeeswari P, et al. Engineering
another class of anti-tubercular lead: hit to lead optimization of an intriguing class of gyrase
ATPase inhibitors. Eur J Med Chem. 2016;122:216–31.
131. Saxena S, Samala G, Renuka J, Sridevi JP, Yogeeswari P, Sriram D. Development of 2-amino-­
5-phenylthiophene-3-carboxamide derivatives as novel inhibitors of Mycobacterium tubercu-
losis DNA GyrB domain. Bioorg Med Chem. 2015;23(7):1402–12.
132. Gjorgjieva M, Tomasic T, Barancokova M, Katsamakas S, Ilas J, Tammela P, et al. Discovery of
benzothiazole scaffold-based DNA gyrase B inhibitors. J Med Chem. 2016;59(19):8941–54.
133. Jakopin Z, Ilas J, Barancokova M, Brvar M, Tammela P, Sollner Dolenc M, et al. Discovery
of substituted oxadiazoles as a novel scaffold for DNA gyrase inhibitors. Eur J Med Chem.
2017;130:171–84.
134. Sun J, Lv PC, Yin Y, Yuan RJ, Ma J, Zhu HL. Synthesis, structure and antibacterial activ-
ity of potent DNA gyrase inhibitors: N'-benzoyl-3-(4-bromophenyl)-1H-pyrazole-5-­
carbohydrazide derivatives. PLoS One. 2013;8(7):e69751.
135. O'Brien S. Meeting the societal need for new antibiotics: the challenges for the pharmaceuti-
cal industry. Br J Clin Pharmacol. 2015;79(2):168–72.
136. Bax R, Green S. Antibiotics: the changing regulatory and pharmaceutical industry paradigm.
J Antimicrob Chemother. 2015;70(5):1281–4.
137. Silver LL. Challenges of antibacterial discovery. Clin Microbiol Rev. 2011;24(1):71–109.
138. Cooper MA, Shlaes D. Fix the antibiotics pipeline. Nature. 2011;472(7341):32.
139. Payne DJ, Miller LF, Findlay D, Anderson J, Marks L. Time for a change: addressing R&D
and commercialization challenges for antibacterials. Philos Trans R Soc Lond Ser B Biol Sci.
2015;370(1670):20140086.
140. Kostyanev T, Bonten MJ, O'Brien S, Steel H, Ross S, Francois B, et al. The Innovative
Medicines Initiative's New Drugs for Bad Bugs programme: European public-private part-
618 A. Maxwell et al.

nerships for the development of new strategies to tackle antibiotic resistance. J Antimicrob
Chemother. 2016;71(2):290–5.
141. Candel FJ, Penuelas M. Delafloxacin: design, development and potential place in therapy.
Drug Des Devel Ther. 2017;11:881–91.
142. Basarab GS, Brassil P, Doig P, Galullo V, Haimes HB, Kern G, et al. Novel DNA gyrase
inhibiting spiropyrimidinetriones with a benzisoxazole scaffold: SAR and in vivo character-
ization. J Med Chem. 2014;57(21):9078–95.
143. Charifson PS, Grillot AL, Grossman TH, Parsons JD, Badia M, Bellon S, et al. Novel
dual-­targeting benzimidazole urea inhibitors of DNA gyrase and topoisomerase IV pos-
sessing potent antibacterial activity: intelligent design and evolution through the judi-
cious use of structure-­guided design and structure-activity relationships. J Med Chem.
2008;51(17):5243–63.
Chapter 20
Antimicrobial-Mediated Bacterial Suicide

Yuzhi Hong, Karl Drlica, and Xilin Zhao

20.1 Introduction

Controlling antibiotic resistance can be considered from two perspectives: limit the
emergence of new resistance and halt the dissemination/horizontal transfer of exist-
ing resistance. For both applications, we expect better results from compounds that
rapidly reduce bacterial burden. A lower pathogen burden will then reduce our reli-
ance on host immune responses, which are likely to decline as populations age, to
clear infection. Lower burden will also help control pathogens as we increase our
use of immunosuppressants. Thus, rapid killing by antimicrobials will become
increasingly important. The present chapter addresses a common mechanism of
rapid killing by focusing on the hypothesis that bacteria respond to severe stress by
accumulating toxic reactive oxygen species (ROS) and thereby self-destruct.

Y. Hong · K. Drlica
Public Health Research Institute, New Jersey Medical School, Rutgers Biomedical
and Health Sciences, Newark, NJ, USA
X. Zhao (*)
Public Health Research Institute, New Jersey Medical School, Rutgers Biomedical
and Health Sciences, Newark, NJ, USA
Department of Microbiology, Biochemistry, & Molecular Genetics, New Jersey Medical
School, Rutgers Biomedical and Health Sciences, Newark, NJ, USA
State Key Laboratory of Molecular Vaccinology and Molecular Diagnostics,
School of Public Health, Xiamen University, Xiamen, Fujian Province, China
e-mail: [email protected]

© Springer International Publishing AG, part of Springer Nature 2018 619

I. W. Fong et al. (eds.), Antimicrobial Resistance in the 21st Century, Emerging Infectious
Diseases of the 21st Century, https://doi.org/10.1007/978-3-319-78538-7_20
620 Y. Hong et al.

Understanding this process may help us better control bacterial populations. Readers
interested in earlier reviews on ROS and on programmed cell death are referred to
references [1–4].
We begin by distinguishing between antimicrobial lethality and blocking growth,
the usual measure of antimicrobial activity, because the terms are frequently mis-
used. Experimentally, blocking growth is measured with antimicrobial in the test
medium. The output is expressed as either minimal inhibitory concentration (MIC)
or as efficiency of plating. MIC is currently the basis for most antimicrobial consid-
erations, including diagnosis, resistance surveillance, new compound development,
and dosing. In contrast, killing is measured as survival after treatment with antimi-
crobial, usually by scoring colony formation on drug-free agar. Input cells that can-
not form colonies after removal of drug are commonly considered to be dead.
Several variations exist for measuring killing. One is to determine the concentra-
tion of drug that reduces survival by a particular amount following a long incuba-
tion, often overnight for rapidly growing bacteria. This measurement, when
performed under standardized conditions, is called minimal bactericidal concentra-
tion (MBC). Many secondary events can occur during the long incubation required
for MBC measurement, which makes it difficult to characterize direct, specific
lethal mechanisms using MBC. In contrast, rapid killing, which determines rate and
extent of killing occurring within a few hours of drug exposure, is measured to study
mechanism and find ways to improve the direct lethal activity of antimicrobials.
Work discussed in the present chapter focuses on rapid killing.
A key concept is that rapid antimicrobial-mediated killing occurs in two general
ways: (1) the primary lesion is sufficient to directly kill the cell, and (2) the primary
lesion induces a lethal, self-destructive stress response (see Fig. 20.1). The contribu-
tion of ROS to lethal activity falls in the second category [3]. Since the relevance of
ROS action relies on cell death, we begin with a brief discussion of how the impor-
tance of lethality is growing as the prevalence of antimicrobial resistance increases.

20.2 Importance of Lethal Action

Traditionally, lethal action has been thought to be important mostly for curing indi-
vidual infections, particularly for problematic diseases such as endocarditis and
meningitis. Dosing decisions have been based on relationships between incubation
conditions and killing, largely using measurements of MBC [5]. Compounds fall
into two general groups. Members of one group, represented by fluoroquinolones
and aminoglycosides, are said to be concentration-dependent killers because
increases in drug concentration result in increased killing. Members of the other,
which includes β-lactams and macrolides, are time-dependent killers – as long as
antimicrobial concentration exceeds MIC by a multiple of 2–5, further increases in
drug concentration have little effect on killing [6]. According to this approach, some
antimicrobials, in particular fluoroquinolones, are thought to be lethal enough to
cure most infections. That perspective questions the need for improving lethality.
20 Antimicrobial-Mediated Bacterial Suicide 621

Fig. 20.1 Relationships among resistance, tolerance, and killing. (a) Antimicrobials create pri-
mary damage that is specific to the compound class. The resulting lesions block growth; the effect
is quantified by the minimal inhibitory concentration (MIC). Resistance interferes with the forma-
tion of the primary lesion. (b) When the initial lesion is sufficiently damaging, a series of events
occur that lead to accumulation of ROS; those events can be perturbed in a variety of ways. (c)
When an ROS threshold is passed, the reactive species cause secondary damage that can then elicit
the accumulation of even more ROS (d) and cell death (e). (f) Some antimicrobials produce lesions
that are lethal independent of ROS production. Bacterial tolerance (g) occurs when lethal action is
inhibited even though the primary lesion occurs and growth is still blocked. Tolerant cells may
suppress lethal pathway (e) by inhibiting respiration and thereby ROS production/accumulation.
Whether tolerance also arises from blockage of lethal pathway (f) is likely compound-specific

The traditional approach worked fine before antimicrobial resistance became a

major problem. Now, as we move deeper into the era of antibiotic resistance, atten-
tion must shift to restricting the emergence and spread of resistance. Simply curing
most patients is unlikely to be adequate when many millions of doses are consumed,
especially with antimicrobials that are mutagenic and thereby lead to induced resis-
tance. Induced resistance is especially noticeable with the quinolone-class antibac-
terials and other DNA-damaging agents that derepress the mutagenic SOS response
(β-lactams also induce the SOS response [7]). The mutagenic action of quinolones
is readily demonstrated by adding the agent to agar, plating bacteria, and then count-
ing colonies daily during a 2-week incubation. A marked increase in colony number
occurs that is blocked by a mutation preventing induction of the SOS response [8,
9]. In this type of experiment, greater lethal activity correlates with reduced recov-
ery of induced mutants [8]. Rapid lethal action is needed to reduce the number of
induced mutants.
A related but distinct concept is that killing may lower antimicrobial concentra-
tions needed to suppress the enrichment of resistant mutant subpopulations present
before drug exposure [10, 11]. If concentrations can be kept above the MIC of the
622 Y. Hong et al.

least susceptible mutant subpopulation, resistance is less likely to emerge than if the
goal is only to keep concentrations above the MIC of the major portion of the popu-
lation [10–12]. Unfortunately, with most pathogen-antimicrobial combinations,
keeping antimicrobial concentrations high enough and exposure time long enough
to restrict the emergence of resistance is likely to have adverse effects on patients.
Thus, it is encouraging that studies with animal models show that some highly lethal
compounds restrict the emergence of resistance without needing to keep concentra-
tions above the MIC of mutant subpopulations throughout treatment [13–16].
Lethal activity is also an important consideration with drug-tolerant cells (see
Chap. 13 for discussion of tolerance). Such bacteria, when exposed to highly lethal
antimicrobials, fail to grow, but they are not killed (in contrast, resistant cells grow
in the presence of drug; see Fig. 20.1 for relationships). Tolerant cells (persisters),
which are usually a small fraction of the population, are a problem because they
may survive treatment and subsequently cause relapse. Since many antimicrobials
are active primarily with replicating cells, non-growing cells can display a form of
tolerance, as seen with ampicillin and first-generation quinolones. Antibiotic toler-
ance can also arise from mutation, and it can facilitate the emergence of resistance
[17]. Thus, finding ways to overcome tolerance, i.e., weakened lethal activity, is of
major importance for restricting the emergence of resistance.

20.3 Experimental Manipulation of ROS

20.3.1 Genetic Perturbations

Key evidence for the contribution of ROS to lethal antimicrobial action derives from
manipulating ROS levels and correlating those changes with bacterial survival. One
type of genetic approach involves increasing or decreasing the expression of cata-
lase/peroxidase. For example, deletion of katG increases ROS accumulation and
lethality arising from treatment of E. coli with quinolones [18–20], from thymine
starvation [21], and from exposure to UV irradiation (Y. Hong & X. Zhao, unpub-
lished observation). Similarly, deletion of ahpCF, which encodes a peroxidase,
increases the lethal activity of kanamycin and ampicillin [18]. In a reciprocal exper-
iment, overexpression of katG suppresses antimicrobial lethality [20].
Many other genes that protect from oxidative stress have more complex effects.
Among these are sodA and sodB, genes that encode superoxide dismutases (enzymes
that convert superoxide to peroxide). A deficiency of either gene has little effect on
the lethal activity of norfloxacin [18]; however, a sodA sodB double mutant protects
from the lethal stress arising from norfloxacin, ampicillin, and kanamycin, as if
elevation of superoxide concentration is protective (wild-type genes would reduce
superoxide levels). A similar conclusion emerged from a study of the DNA-­
damaging agent bleomycin [22]. In this case, a sodA sodB double mutant exhibited
reduced killing from bleomycin, while overexpression of superoxide dismutase
20 Antimicrobial-Mediated Bacterial Suicide 623

increased lethal activity. Likewise, treatment with a low concentration of plumba-

gin, a metabolic generator of superoxide, protected from the lethal activity of bleo-
mycin [22]. The protective effect of elevated superoxide was confirmed by a
subsequent finding in which low concentrations of plumbagin or paraquat, another
metabolic generator of superoxide, reduced the lethal activity of oxolinic acid,
kanamycin, and ampicillin [23]. When overall ROS levels were examined using a
fluorescent probe, the increase associated with oxolinic acid treatment was muted
by plumbagin and paraquat [23]. Apparently, spontaneously elevated concentrations
of superoxide in the sodA sodB double mutant induce genes that lower levels of
other ROS during antimicrobial treatment and thereby protect from lethal stress.
The identity of those genes is unknown.
Not every mutant has provided a simple yes or no answer to whether it stimulates
or restricts the lethal stress response. For example, preinduction of many protective,
oxidative stress genes may account for an oxyR deficiency and for the Hpx− triple
mutant (Hpx− contains deficiencies in ahpCF, katE, and katG) showing no effect on
antimicrobial lethality [24] – if these mutants have elevated levels of anti-oxidative
defenses before addition of antimicrobial, subsequent ROS production could be
dampened when stressor is added. In other cases, discussed below, a gene can
appear to be protective at low levels of stress and destructive at high levels. Thus,
stressor concentration or exposure time can be a crucial variable in antimicrobial-
ROS experiments.
Another idea emerging from early work was that iron is released from iron-sulfur
clusters during lethal antimicrobial treatment. That iron would then become avail-
able for conversion of peroxide to hydroxyl radical via the Fenton reaction [25]. The
hypothesis was based largely on a deficiency in iscS protecting from the lethal
action of norfloxacin, ampicillin, and kanamycin. The deficiency also suppressed
the increase in the hydroxyl radical signal from a fluorescent dye. However, this
conclusion needs to be revisited, because follow-up work with gentamicin, an ami-
noglycoside, argued that perturbing iron-sulfur clusters interferes with drug uptake
[26]. Since lethality cannot be studied if drug uptake is blocked (see Fig. 20.1), care
must be taken to assure that the lethal response is being studied, not the many steps
that lead to formation of the primary, stress-mediated lesion.

20.3.2 Chemical Perturbations

One chemical approach involves treating bacterial cells with antioxidants, such as
thiourea, dimethyl sulfoxide, ascorbic acid, glutathione, or resveratrol. Such com-
pounds are expected to scavenge hydroxyl radical. With chemical perturbation, an
important consideration is the concentration of the perturbing agent, since that con-
centration needs to be adjusted to avoid interfering with bacterial growth – growth
inhibition can create a type of tolerance. Another consideration is whether antioxi-
dants actually scavenge hydroxyl radical, since the rapid reaction of hydroxyl radi-
cal with other biomolecules might obscure any reaction with thiourea [24]. That
624 Y. Hong et al.

would make the observed protection from killing by thiourea [25] an off-target
effect. To our knowledge, no off-target effect of antioxidants has been identified.
Another chemical perturbation uses iron chelators, such as 2,2′ bipyridyl. These
agents are expected to act in two ways. One is by inhibiting the iron-requiring
Fenton reaction in which hydrogen peroxide produces hydroxyl radical. A second
involves iron-requiring proteins involved in respiration, a source of superoxide and
ultimately hydroxyl radical. Experiments employing bipyridyl have consistently
supported the conclusion that lethal activity arising from a variety of stressors
involves ROS [18, 25].
Ruling out off-target effects due to chelators and antioxidants is difficult. One
argument is that antioxidants that interfere with antimicrobial lethality are chemi-
cally diverse and unlikely to have the same off-target effect. Another argument
derives from consideration of catalase as an antioxidant. We found that E. coli cells,
thought to be killed by the lethal antimicrobial trimethoprim and then plated on
drug-free agar, are revived by addition of catalase to the plated cells (Y. Hong and
X. Zhao, unpublished observations). Since hydrogen peroxide readily enters and
exits from cells, extracellular catalase would lower the overall peroxide concentra-
tion and provide a specific protective effect that would not be attributed to either
off-target effects or growth inhibition. Overall, effects of chelators and antioxidants
fit well with their suppression of ROS accumulation.

20.4  actericidal Activity of Antimicrobials Mediated

B
by ROS

20.4.1 Development of the ROS-Lethality Hypothesis

Credit for the general nature of the ROS-lethality hypothesis is usually attributed to
a 2007 report from the Collins laboratory [25], although involvement of ROS with
lethal activity had been proposed earlier. For example, activation of the SoxRS regu-
lon conferred resistance to multiple antimicrobial classes [27–29]; moreover, anti-
oxidants, such as vitamin C and glutathione, raised minimal inhibitory concentration
(MIC) and efficiency of plating for quinolones and aminoglycosides [30, 31]. In
addition, elevated levels of oxidative stress signals were detected in cells treated
with antimicrobials [32, 33]. However, such observations do not establish a connec-
tion between ROS and cell death, because the measurements reported inhibition of
bacterial growth, not cell death. Thus, the work by Kohanski et al. [25] was a quali-
tative advance because killing was measured. We later showed that the killing was
separate from bacteriostatic effects by normalizing killing to MIC [18].
Among the central observations from the Kohanski et al. study [25] was that kill-
ing caused by norfloxacin, ampicillin, or kanamycin was suppressed by thiourea
and 2,2′ bipyridyl. Although the experimental conditions also suppressed bacterial
growth, which itself was known to interfere with killing, subsequent work [18]
showed that subinhibitory concentrations of thiourea and bipyridyl protect from
20 Antimicrobial-Mediated Bacterial Suicide 625

antimicrobial lethality. The latter study also described genetic perturbations that
supported the hypothesis [18].
It soon became clear that not all derivatives of an antimicrobial class depend on
ROS to kill cells [34]. For example, the quinolones were known to kill by two gen-
eral pathways, one that was blocked by chloramphenicol, an inhibitor of protein
synthesis, and one that was not [35–37]. This dichotomy also applied to the contri-
bution of ROS: 2,2′ bipyridyl treatment [34] and anaerobic conditions [38] paral-
leled the behavior of chloramphenicol. Overall, the quinolone experiments
established that first-generation compounds, such as nalidixic and oxolinic acids,
kill E. coli by a mechanism that relies heavily on ROS; the fluoroquinolones addi-
tionally kill bacteria by a pathway that appears to rely more on chromosome frag-
mentation than on ROS, although fluoroquinolones still trigger ROS accumulation
[34]. Within this scheme, norfloxacin is an outlier having an intermediate,
concentration-­dependent activity [38]. As pointed out below in Sect. 5, subsequent
challenges to the ROS-lethality hypothesis relied in part on work with norfloxacin.

20.4.2 Factors Involved in ROS Accumulation

Steps leading from stress to accumulation of ROS have been investigated by identi-
fying genes that, when defective, alter ROS levels and stress-mediated cell death.
The action of the gene products can be fit into a scheme (Fig. 20.2). One of the fac-
tors is the MazEF toxin-antitoxin pair. The MazF protein is an endoribonuclease
that during stress cleaves mRNA and thereby blocks protein synthesis. At low, bac-
teriostatic concentrations of antimicrobial, blocking gene expression and bacterial
growth by MazF would allow time for cells to efflux noxious molecules and repair
damage. However, at high, lethal levels of stress, MazF would produce toxic levels
of truncated mRNA and misfolded proteins that perturb cell membrane function and
elevate ROS levels. Thus, MazF is expected to be bifunctional with respect to stress.
Indeed, bifunctionality has been observed with Bacillus subtilis. For example, at
low doses of UV irradiation or low concentrations of moxifloxacin, a ΔndoA (mazF)
mutant is more readily killed than wild-type cells, but at high doses the opposite is
seen [39]. Bifunctionality is a key feature of a stress response that either allows
repair of minor damage or causes self-destruction when damage is severe.
EF4 is another factor that contributes to ROS accumulation. This ribosomal elon-
gation protein is normally sequestered in the cell membrane, but during stress, the
protein enters the cytosol, binds to stress-stalled ribosomes, and stimulates ribo-
somal back-translation. These events allow protein synthesis to recover from mod-
erate stress. But EF4 also blocks tagging of truncated proteins for degradation by
tmRNA – this activity would facilitate the accumulation of toxic peptides, presum-
ably arising from MazF action. Indeed, the absence of EF4 protects E. coli from
being killed by quinolones (wild-type protein would be destructive) [40]. Thus, EF4
and MazEF appear to be bifunctional proteins that protect from moderate stress but
promote death when stress is high. How the action of MazF and EF4 is connected
to the initial lesion is unknown.
626 Y. Hong et al.

Fig. 20.2 Scheme describing bifunctional nature of factors involved in the live-or-die stress-
response pathway. Low to moderate levels of stress result in a protective stress response. The stress-
ors, such as quinolones, generate specific primary lesions that in an unknown way stimulate the
MazF toxin to cleave mRNA, thereby halting translation and allowing cells time to repair damage.
The protein fragments resulting from translation or MazF-mediated mRNA cleavage are tagged for
degradation by tmRNA and EF4, a ribosomal protein that facililtates restart of stalled ribosomes.
Truncated, misfolded peptides that enter cell membranes stimulate the Cpx two-component system
to induce genes involved in membrane protein repair. Safety valves YihE kinase and KatG catalase
negatively regulate MazF and detoxify peroxide, respectively. At high stress levels, mRNA cleavage
by MazF is extensive, and EF4 blocks the tagging of truncated peptides by tmRNA. Thus, high
levels of protein fragments accumulate and enter cell membranes. That causes the Cpx system to
induce the Arc two-component system, which in turn leads to high-level production of superoxide.
Superoxide dismutates to peroxide; the Fenton reaction then converts peroxide to hydroxyl radical.
Hydroxyl radical damages nucleotides and many macromolecule types, causing mutations and cell
death. To assure death, MazF cleaves katG mRNA, which lowers the level of KatG, a protein that
would otherwise reduce peroxide levels. Other protective functions, such as induction of membrane
repair by Cpx/Arc and induction of the SoxRS/OxyRS regulon by superoxide/peroxide, are
overwhelmed

Insertion of truncated, misfolded proteins into the cell membrane activates a two-­
component membrane stress-response system called Cpx [41, 42]. When stress is
high, activation of Cpx stimulates another two-component system (Arc) that then
contributes to the generation of elevated levels of ROS [43] as discussed below in
Sect. 6. The destructive feature of Cpx is revealed by a CpxR deficiency protecting
20 Antimicrobial-Mediated Bacterial Suicide 627

from nalidixic acid-mediated cell death [19]. The Cpx system also serves to repair
membrane protein damage [44] and to mitigate MazF toxicity [19] when stress is
moderate. Thus, Cpx has both destructive and protective functions; it represents a
third bifunctional system involved in lethal antimicrobial action. Arc also exhibits
protective and destructive properties (arc is discussed in more detail in Sect. 6.2).
Superoxide occupies a central position in the lethal stress response. At moderate
stress levels, superoxide is thought to accumulate and stimulate protective gene
responses, such as induction of the SoxRS regulon. During periods of harsh stress,
superoxide may accumulate to high levels and rapidly dismutate, thereby creating
elevated levels of hydrogen peroxide (superoxide dismutases are 1000-fold more
efficient than catalase/peroxidase, enzymes that detoxify hydrogen peroxide by con-
verting it to water). The result is accumulation of hydrogen peroxide, which can
then be converted to hydroxyl radical, a compound whose toxic effects may over-
whelm the protective functions stimulated by superoxide. As pointed out above,
such a bifunctional nature of superoxide would explain why low concentrations of
metabolic generators of superoxide, such as plumbagin and paraquat, reduce the
lethal action of bleomycin and the quinolones [22, 23, 45], even though high con-
centrations of plumbagin kill bacteria.
Several safety valves operate within this scheme of bifunctional factors. One
involves a protein kinase called YihE. The absence of YihE elevates the lethal action
of nalidixic acid by 100-fold but only if the MazEF toxin is present [19]. Thus, YihE
appears to be a negative regulator of MazF. Another safety valve is the katG cata-
lase/peroxidase, which, as mentioned above, detoxifies peroxide by converting it to
water. Deletion of katG increases norfloxacin lethality by 20-fold without affecting
MIC [18]. Similarly, deletion of ahpCF increases lethality of both ampicillin and
kanamycin, although this deletion has no effect on quinolone-mediated killing.
Among the many protective genes (safety valves) induced by oxidative stress are
efflux pumps that remove noxious stressors [23, 45]; presumably the pumps reduce
the signal leading to ROS accumulation.
Variations on the theme developed above are likely to emerge as more is learned
about individual stressors. For example, each stressor generates a unique lesion that
is likely to be recognized in a unique way, and many pathways may connect the
lesion to the accumulation of ROS. Thus, the scheme in Fig. 20.2 should be consid-
ered only as a framework for future testing.

20.4.3 ROS-Mediated Toxicity

The most toxic of the reactive oxygen species is hydroxyl radical. It readily dam-
ages DNA by creating single-strand breaks and by converting single-strand DNA
lesions into double-strand breaks [46]. Hydroxyl radical also carbonylates proteins
[47] and peroxidates membrane lipids [48]. In addition to directly breaking DNA,
hydroxyl radical oxidizes the guanine nucleotide pool, thereby producing 8-oxo-­
guanine. When 8-oxo-guanine is incorporated into DNA, it is expected to lead to
double-strand breaks. The DNA breaks are thought to derive from excision of the
628 Y. Hong et al.

8-oxo-dG, since deficiencies in the excision enzymes protect (10–20 fold) from the
lethal action of norfloxacin, ampicillin, and kanamycin [20, 49]. Overexpression of
mutT, which “sanitizes” 8-oxo-guanine from the nucleotide pool, reduces the rate of
killing for the three antimicrobial classes [49]. Thus, lethal DNA damage is a com-
mon theme associated with killing, even by antimicrobials that do not have DNA as
their primary target. That explains why ampicillin induces the SOS response [7].
DNA damage leads to the interesting possibility that ROS is self-amplifying.
When a threshold concentration of ROS is reached, DNA damage could be suffi-
cient to stimulate the production of even more ROS. Like a nuclear reaction, ROS
accumulation would be unstoppable, and death would be assured. In this sense,
bacterial cells self-destruct when faced with severe lethal stress [19]. To our knowl-
edge, stress-mediated ROS accumulation due to self-amplification has not been
demonstrated. One prediction from the self-amplification idea is that primary stress
of one type, such as DNA damage by a quinolone, will cause ROS-mediated dam-
age of a second type, such as protein carbonylation. Repair of protein damage, if
specifically blocked by a lon or hslV deficiency, would exacerbate damage of a third
type, such as lipid peroxidation. Our unpublished work supports this scenario
(X. Wang & X. Zhao).

20.5  hallenges to the ROS-Mediated Stress-Response

C
Hypothesis

The initial proposal for ROS-mediated antimicrobial lethality [25] required modifi-
cation with respect to sod effects (Sect. 3.1), reconsideration of mutations affecting
iron-sulfur proteins (Sect. 3.1), and complexities with respect to fluoroquinolones
having at least two lethal mechanisms [34, 38]. However, the overarching idea was
independently solidified [18, 19, 23]. At about the same time, two laboratories chal-
lenged the idea that ROS contribute to the lethal action of multiple antimicrobial
classes. Below we discuss the resulting controversy.
One of the challenging studies [24] reported failure to confirm ROS involvement
in the lethal action of norfloxacin, kanamycin, and ampicillin under conditions that
were similar to those reported by Kohanski et al. [25]. Those conditions, which
involved use of a single drug concentration, appear to have been too narrowly
defined to take into account differences in conditions between laboratories. For
example, norfloxacin concentration was known to be important for observing lethal-
ity [38], so it needed to be varied. In the other work, Keren et al. [50] focused on
whether killing depends on ROS (requires ROS). By examining high levels of stress,
they confirmed that fluoroquinolones have an ROS-independent mode of killing [34,
38] that is insensitive to anaerobiosis [38]. The issue of whether ROS contribute to
killing, as proposed by Collins [25] and extended by our work [18], was actually
supported by Keren et al. [50] at low levels of stress. The high levels examined were
outside the range of discrimination: only residual “persisters” survived (such cells
20 Antimicrobial-Mediated Bacterial Suicide 629

are expected to be insensitive to ROS-mediated killing due to lack of respiration).

Moreover, many of the experiments involved long incubation times, a feature that is
problematic, since it was known that ROS effects are largely kinetic [51]. Thus,
ROS are not required for killing, as was previously known, and whether they con-
tribute to lethal action was not experimentally challenged. Subsequent work from
Collins [20] solidified the idea that ROS contribute to antimicrobial killing by
addressing other issues raised by Keren et al. and Liu and Imlay (discussed below).
A follow-up review article by Imlay suggested that in bacterial cells, endogenous
ROS concentrations are not high enough to kill cells [52]. In studies of thymineless
death [21] and quinolone-mediated killing (G. Luan, unpublished observations), we
found that lethal stress creates a substrate that is hypersusceptible to ROS attack,
thereby making generation of extreme levels of endogenous ROS unnecessary to
kill cells. Overall, the challenges have led to reexamination and solidification of the
ROS-lethality hypothesis.
Two other reports also contained objections to the idea that lethal antimicrobials
kill bacteria by a common mechanism involving ROS. One [53] showed that cpx-­
mediated resistance applies to some (aminoglycoside, hydroxyurea) but not all
(fluoroquinolone, β-lactam) antimicrobials. Since resistance is a bacteriostatic phe-
nomenon (see Fig. 20.1), the work failed to focus on the lethal stress response. The
second report [26] showed that for gentamicin, an iron requirement concerns drug
uptake, which would supersede iron effects on ROS formation (similar issues apply
to respiration and drug uptake, as discussed in Sect. 6). Additional work is required
to separate drug uptake from a lethal stress response.
The issues raised above elicited several sets of follow-up experiments [20]. One
set involved additional dyes to detect a variety of ROS. In general, ampicillin and
norfloxacin were very active; gentamicin was less striking, but the results were
clear. Another set addressed the failure [24] to detect extracellular peroxide follow-
ing antimicrobial treatment. When an intracellular assay was introduced, peroxide
was readily detected. A third point involved induction of promoters of genes known
to be sensitive to peroxide (pOxyS) and superoxide (pSoxS). Both were induced by
norfloxacin and ampicillin. These tests, which gave signals equivalent to 10 μM
exogenous hydrogen peroxide, produced gene expression patterns that were similar
for this peroxide concentration and for norfloxacin or ampicillin treatment. In a
fourth point, lethal activity for the three antimicrobial classes was lower under
anaerobic conditions. Thus, the key issues raised by the challenges to the ROS-­
lethality hypothesis are accommodated by experimental considerations [2, 3, 20]
and by reiterating that ROS contribute to, rather than completely account for, lethal
activity [34, 38].
Among the necessary clarifications were experimental definitions of killing and
the lethal stress response (Fig. 20.1). The lethal stress response must occur after
formation of the primary lesion. Consequently, studies that focus on factors acting
at or before primary lesion formation, such as drug uptake, efflux, and target inter-
actions, are largely uninformative. The effects of these factors can be removed from
consideration by normalizing lethal activity to growth inhibition (MIC) [18], a fea-
ture often absent from studies of ROS and killing.
630 Y. Hong et al.

20.6  ole of Respiration in ROS Accumulation and Cell

R
Death

20.6.1 Source of ROS

The respiratory chain is a major source of superoxide and hydrogen peroxide [54,
55], as these ROS form when molecular oxygen oxidizes redox enzymes, such as
fumarate reductase (Frd), succinate dehydrogenase (Sdh), and aspartate oxidase
(NadB) [56]. Indeed, about a quarter of the cytoplasmic H2O2 derives from NadB
[57]. The electrons involved in formation of superoxide and hydrogen peroxide are
thought to derive largely from the tricarboxylic acid (TCA) cycle, which we discuss
below. In principle, dismutation of superoxide can also serve as a source of H2O2,
which can then form highly toxic hydroxyl radical via the Fenton reaction.
TCA cycle. Several lines of evidence connect the TCA cycle with the lethal action
of antibiotics. One is that treatment of E. coli cultures with bactericidal antibiotics
leads to upregulation of genes involved in central metabolism, including the TCA
cycle and respiration [25, 43, 58]. A second line links the efficacy of bactericidal
antibiotic therapy to carbon flux through the TCA cycle [59, 60]. A third line showed
that mutations in genes involved in the TCA cycle reduce stress-mediated lethality.
Surprisingly, antibiotic classes, represented by norfloxacin, ampicillin, and kana-
mycin, differ in the TCA cycle genes involved. For example, norfloxacin-­mediated
lethality is reduced by mutation of icdA and acnB, while killing by ampicillin is
reduced by mutation of these two genes (icdA and acnB) and sucB. Killing by kana-
mycin is reduced by mutation of four genes, icdA, acnB, sucB, and mdh [25].
Additional work is needed to determine whether these differences are characteristic
of drug mechanism or whether they result from differences in the concentration of
the three drugs examined (normalized to MIC).
Respiration. E. coli encodes three cytochrome terminal oxidases: bd-I (CydAB),
bd-II (AppAB), and bo (CyoABCDE). Reduction of stress-mediated lethality by a
deficiency of cydB, which is associated with decreased ROS levels during hydroxy-
urea treatment [61] and thymineless death [21], establishes the importance of the
respiratory chain. Moreover, the rate of oxygen consumption serves as a measure of
respiration that can be compared with antimicrobial-mediated cell death [20, 62].
For example, a deficiency of atpA increases oxygen consumption, ROS level, and
killing during treatment of cells with ampicillin or norfloxacin [62]. Conversely, a
cytochrome oxidase null mutant fails to show an acceleration of respiration or cell
death when treated with norfloxacin, ampicillin, or gentamicin [62]. Treatment of
wild-type cells with the bacteriostatic agent chloramphenicol rapidly attenuates cel-
lular respiration and the lethality associated with lethal doses of norfloxacin or
ampicillin [62]. In yet another example, NADH-coupled electron transport (NADH
dehydrogenase I) is a common upregulated pathway for all three bactericidal drugs
[25]. Such data fit with upregulation of genes involved in central metabolism and
respiration being associated with exposure to lethal antimicrobials [25, 43].
20 Antimicrobial-Mediated Bacterial Suicide 631

Complexities of drug uptake and killing. We have proposed that ROS contributes to
the lethal activity of aminoglycosides (kanamycin), because lethal action, when
normalized to MIC to rule out effects of drug uptake, efflux, and target interactions,
was decreased by mutation of superoxide dismutase genes and increased by a defi-
ciency in a catalase/peroxidase [18]. Other works implicated the TCA cycle in ami-
noglycoside lethality [25, 43], and recently fumarate or glyoxylate, used to activate
or inhibit the TCA cycle, potentiated or suppressed, respectively, the lethal activity
of tobramycin with Pseudomonas aeruginosa [63]. Complexity arises because
proton-­motive force, which derives from respiration and is necessary for aminogly-
coside uptake [26], is a required precursor to lethal activity. Moreover, uptake and
killing are stimulated by the same factor (e.g., respiration). Thus, normalization to
MIC is required to remove uptake from consideration before the effect of respiration
on ROS and killing can be assessed.

20.6.2  ole of the Arc Two-Component System

R
in the Response to Oxidative Stress

The E. coli ArcAB two-component system participates in the regulation of multiple

operons that are involved in central metabolism, such as the enzymes of the TCA
cycle, pyruvate dehydrogenase, cytochrome o ubiquinol oxidase, and NADH-­
quinone oxidoreductase I [64]. In general, Arc represses these genes during anaero-
bic growth. A connection to antimicrobial lethality was made through a gene
expression study in which the effects of a lethal aminoglycoside (gentamicin) were
compared to those of the bacteriostatic derivative spectinomycin – a spike in the
expression of arc-associated TCA cycle genes was specific to gentamicin treatment
[43]. Deletion of arc then reduced gentamicin-mediated accumulation of hydroxyl
radical and increased survival [43]. The ability of an arc deficiency to protect from
the lethal activity of norfloxacin and ampicillin suggested a position for wild-type
arc in the pathway stimulating self-destruction [43] (the connection still requires
tests in which lethal action is normalized to MIC, as pointed out above for other
experiments involving the TCA cycle, respiration, and aminoglycoside-mediated
killing).
The wild-type arc system also appears to protect from oxidative stress, an effect
opposite to the stimulation of killing described above. The DNA-binding activity of
ArcA is controlled by reversible phosphorylation through ArcB, whose kinase
activity is governed by the redox states of quinone pools. Those pools are linked to
the NADH/NAD+ redox couple through respiration [64]. Under some conditions,
disruption of ArcAB leads to reduced survival following oxidative stress generated
by hydrogen peroxide [65, 66].
The arc system also upregulates proteases involved in removal of misfolded
membrane proteins [43], which would be protective. This observation fits with cross
talk between Arc and Cpx, which repairs damage to membrane proteins. As pointed
632 Y. Hong et al.

out in an earlier section, Cpx protects from low levels of lethal stress and contributes
to death when stress is high [19]. Indeed, Cpx may contribute to the accumulation
of ROS by stimulating Arc to accelerate the TCA cycle [43]. We reiterate the gen-
eral theme in which a set of proteins (MazEF, EF-4, Cpx, and ArcAB) protect from
moderate levels of lethal stress, but when stress is severe, the proteins contribute to
the accumulation of ROS and cell death.

20.7 ROS-Mediated Programmed Cell Death

Among the important questions concerning the ROS-lethality hypothesis is whether

ROS accumulation and toxicity cause death or whether dead cells are simply the
source of ROS. One type of support for ROS causality is that pretreatment with a
low dose of peroxide causes a 30-min lag in killing by norfloxacin, ampicillin, and
gentamicin [20], consistent with protective systems being induced.
The question of causality can also be addressed by determining whether bacterial
cells continue along an ROS-dependent death pathway even after removal of the
initial stressor. As a test, we treated E. coli cultures with quinolones (nalidixic acid
or ciprofloxacin) at lethal concentrations for times sufficient to reduce survival
when assayed by colony formation on drug-free agar. When we included 0.5 x MIC
thiourea (an ROS scavenger) in the agar to block ROS action occurring after removal
of quinolone, thiourea increased survival by 5–10- and ~100-fold for wild-type and
ΔyihE cells, respectively [19]. These data show that the toxic action of ROS contin-
ues even after removal of the initial stressor in cells that are not yet dead (they still
form colonies if treated with thiourea).
In follow-up work, we took advantage of the finding that when a dnaB-TS mutant
is shifted to nonpermissive conditions, cells die if the medium is rich in nutrients
(they live if the medium contains only minimal salts and glucose) [37]. We reasoned
that cell death would occur in rich medium because higher levels of ROS would be
produced by the stress of replication inhibition, as we observed in a study of thy-
mineless death [21]. Thus, temperature shifts with a reversible dnaB-TS mutant
provided another way to rapidly remove the primary stressor. We found that killing
was blocked by the presence of catalase or an ROS scavenger (thiourea) in the agar
plates used for determination of survival (Y. Hong & X. Zhao, unpublished observa-
tion). Thus, two types of stressor appear to stimulate bacterial programmed cell
death (PCD), which we define as ROS-dependent death arising after removal of the
primary stressor.
PCD, which is known to require metabolic energy, has previously been associ-
ated with physiological or developmental signals (reviewed in [4]). For example,
with eukaryotic cells, PCD includes apoptosis, autophagy, programmed necrosis,
tissue homeostasis, immune function, stress responses, and processes that are criti-
cal for embryogenesis. With bacteria, the concept of PCD has been controversial,
because the molecular events in bacterial death have not been identified [67, 68] and
because it has not been obvious how suicide by some members of a bacterial popu-
20 Antimicrobial-Mediated Bacterial Suicide 633

lation would benefit other members [69]. Nevertheless, the existence of bacterial
PCD is supported by studies of sporulation and by detection of apoptosis biomark-
ers previously reported for eukaryotic systems (with antibiotic stress, E. coli cells
exhibit DNA fragmentation, chromosome condensation, and membrane depolariza-
tion [70]). Among the examples of bacterial development are mother cell lysis dur-
ing spore formation, which is seen with Bacillus subtilis and Myxococcus xanthus
(as reviewed in [4]). With Xanthomonas campestris, apoptosis-like phenotypes, as
mentioned above, are seen during death caused by incubation in rich Luria-Bertani
(LB) medium [71, 72]. However, in none of these examples have cells been shown
to continue along a documented death pathway after withdrawal of the primary
stressor – in each case, the primary stressor was present throughout the experiment.
Thus, data showing that lethal stress propels bacteria along an ROS-requiring death
pathway, described above, add key support to the idea that bacteria undergo a self-­
destructive process in response to stress.

20.8  aradoxical Tolerance at High Quinolone

P
Concentration Involves ROS

A different type of ROS-related phenomenon is observed with the paradoxical loss

of killing that occurs at very high concentrations of quinolone. At moderate concen-
trations, quinolones are very lethal, but with some derivatives and bacterial species,
100% of cultured cells survive when quinolone concentrations are very high. The
phenomenon is observed with a variety of bacterial species and with many different
quinolones, most notably with nalidixic acid, a first-generation member of the
quinolone-­type compounds (nalidixic acid is formally a naphthyridine rather than a
quinolone, as are some other clinically used agents, such as enoxacin and gemi-
floxacin). We reasoned that if ROS contribute to quinolone lethality, a drop in ROS
accumulation could explain the loss of killing seen at very high nalidixic acid
concentrations.
When E. coli cultures are treated with various concentrations of nalidixic acid,
followed by measurement of ROS using an ROS-sensitive dye, high, nonlethal con-
centrations of the drug induce lower levels of ROS than moderate, lethal concentra-
tions when measured by fluorescence microscopy (individual cells) and flow
cytometry (batch cultures) (G. Luan et al., unpublished observations). At the high,
nonlethal quinolone concentrations, sublethal doses of exogenous hydrogen
­peroxide become lethal and eliminate nalidixic acid-associated paradoxical sur-
vival. Thus, the quinolone-mediated lesions needed for ROS toxicity persist at high,
nonlethal quinolone concentrations. That leaves a drop in ROS as the most likely
explanation for loss of killing.
Nalidixic acid-induced accumulation of ROS and death are blocked by inhibitors
of protein synthesis, such as chloramphenicol ([73] and G. Luan et al., unpublished
observations). Among the effects of chloramphenicol is inhibition of respiration, the
source of ROS [62]. We found that a deficiency of catalase (ΔkatG) raised nalidixic
634 Y. Hong et al.

acid-induced ROS levels and overcame the inhibitory effect of chloramphenicol on

quinolone-mediated killing. Thus, the inhibitory effects of chloramphenicol must
not be absolute. As expected, stimulation of nalidixic acid lethality by ΔkatG was
blocked by additional treatment with a combination of thiourea and bipyridyl.
Nalidixic acid also inhibits protein synthesis [73], a finding that leads to the fol-
lowing explanation for high-concentration tolerance. At high concentrations, nali-
dixic acid causes chromosome breakage [36, 74] that leads to loss of DNA
supercoiling. Initiation of transcription and therefore translation are expected to
decline when supercoiling is lost, which will lead to a reduction in respiration and
ROS. Since DNA damage can be repaired during the long incubation period required
to measure survival (X. Wang & X. Zhao, unpublished observations), a reduction of
ROS would allow bacterial survival. Thus, the ROS-lethality hypothesis provides an
explanation for a decades-old quinolone mystery.

20.9 Thymineless Death and Related Antimicrobials

A study of thymineless death provides insight into one way in which ROS kill bac-
teria. Thymineless death refers to the rapid loss of bacterial viability that occurs
during starvation for thymine or thymidine. The phenomenon is of considerable
interest, because the underlying molecular events also apply to the action of several
antibacterial (trimethoprim, sulfamethoxazole), antimalarial (pyrimethamine, sul-
fonamide), anticancer (methotrexate, fluorouracil), and immune-modulating (meth-
otrexate) agents. Many explanations have been proposed to explain thymineless
death. Among these are unbalanced growth, toxin-antitoxin module action, nucleo-
tide misincorporation, induction of the SOS regulon, destruction of replication
forks, and partial degradation of DNA at oriC [75, 76]. None of these explanations
have satisfied all of the experimental observations. Recent attention has focused on
proteins involved in recombinational repair, as some appear to contribute to death
while others facilitate survival [77–80].
Since DNA replication is severely slowed by withdrawal of thymidine and since
other means of inhibiting DNA replication (fluoroquinolone or hydroxyurea treat-
ment) cause the accumulation of ROS, we examined the possibility that ROS con-
tribute to thymineless death. Two processes appear to be required for rapid death:
generation of persistent single-strand DNA regions and accumulation of ROS [21].
The attack of single-strand DNA regions by ROS then leads to lethal d­ ouble-­stranded
DNA breaks. Interference with either production of single-strand DNA or accumu-
lation of ROS inhibits thymineless death. Our current view is that proteins involved
in recombinational DNA repair, such as RecF and RecQ, expand the single-­strand
DNA substrate for ROS attack (the absence of these proteins reduces thymineless
death). Other DNA repair proteins, such as RecBC, are involved in surviving the
damage, and the SOS response increases expression of SulA, which serves as a
checkpoint that halts cell division until the damage to DNA is repaired in cells that
do not die [77].
20 Antimicrobial-Mediated Bacterial Suicide 635

Although both quinolone treatment and thymine starvation block replication,

which is presumably a signal for initiating ROS accumulation, subsequent events
must differ, because different genes are involved. For example, ROS accumulation
and thymineless death are unaffected by the absence of toxin-antitoxin modules,
and disruption of arcA/B, lon, clpA/P, cpxA/R, or ssrA has little effect on thymine-
less death [21]. Nevertheless, surges in ROS and involvement of the respiratory
chain are common features [25, 61, 62]

20.10 Potential Consequences of Antioxidant Consumption

Part of the evidence supporting a role for ROS in antimicrobial lethality stems from
the protective effects of antioxidants. For example, thiourea, glutathione, and vita-
min C reduce fluoroquinolone lethality by orders of magnitude [18, 25, 51, 81];
thiourea and glutathione reduce lethality of daptomycin and oxacillin by 10–100-­
fold [51]. Since human consumption of antioxidants is large [82], the potential for
interference with antimicrobial action exists. One aspect may involve food prod-
ucts. In a recent study, Marathe et al. [83] examined effects of the antioxidant cur-
cumin on ciprofloxacin-mediated lethality with Salmonella. Curcumin is a common
food ingredient in Southeast Asia, and it is often used medicinally. Curcumin
reduces the lethal activity of ciprofloxacin with cultured bacteria, with bacteria
infecting macrophages, and with bacteria infecting mice. Since curcumin also sup-
presses the antibacterial activity of the immune response, it may act in two ways to
increase bacterial survival [83]. Thus, a cautionary note has been raised concerning
antibiotic therapy and consumption of foods having antioxidant activity [83].
Antioxidants are also consumed as nutritional supplements. One of the popular
agents is resveratrol, a natural polyphenol antioxidant [81]. Resveratrol is thought
to have beneficial effects for ailments such as cardiovascular disease [84], neurode-
generative disorders [84, 85], and some forms of cancer [84, 86]. To address whether
antioxidant nutritional supplements are likely to interfere with the ability of antimi-
crobials to kill bacteria, we added resveratrol to cultures of E. coli and S. aureus,
treated them with antimicrobial, and assayed for bacterial survival [87]. Resveratrol,
at concentrations likely to be present during human consumption, reduced killing by
two- to threefold during a 2-h exposure to moxifloxacin or kanamycin. At higher but
still subinhibitory concentrations, resveratrol lowered antimicrobial lethality by
more than 1000-fold. Resveratrol also reduced the accumulation of ROS
characteristic of treatment with oxolinic acid, a first-generation quinolone.
­
Collectively these observations support the general idea that the lethal activity of
some antimicrobials involves ROS.
Subinhibitory concentrations of resveratrol also promoted (two- to sixfold) the
recovery of rifampicin-resistant mutants arising from the action of ciprofloxacin,
kanamycin, or daptomycin. This finding can be explained by resveratrol lowering
ROS to sublethal levels that are still mutagenic, while the absence of resveratrol
allows ROS levels to be high enough to kill mutagenized cells. Suppression of kill-
636 Y. Hong et al.

ing by antimicrobials and promotion of mutant recovery by curcumin and resvera-

trol suggest that antioxidant consumption may contribute to the emergence of
antimicrobial resistance, especially if new derivatives and/or formulations of resve-
ratrol markedly raise bioavailability. One approach to preserving antibiotic activity
would be to suspend antioxidant consumption during treatment with lethal
antimicrobials.

20.11 Unresolved Issues

20.11.1 In Vitro Systems

Several aspects of the ROS-lethality hypothesis require additional work. A central

issue is to understand how a bacterium communicates information about an antimi-
crobial lesion to the apparatus that produces ROS. How the initial lesion is detected
is not known for any system. A second issue concerns events that occur during
anaerobic growth: some antimicrobials are lethal in the absence of oxygen, which
would preclude involvement of ROS. Although killing may proceed through the
direct action of primary lesions, we postulate that under anaerobic conditions, non-­
oxygen centered radicals perform the same function as ROS during aerobic growth.
That hypothesis is currently being tested.
Chemical probes present a different type of challenge. While both thiourea and
bipyridyl protect from antimicrobial lethality, off-target effects are difficult to elimi-
nate. In the case of iron chelators, such as 2,2′-bipyridyl, action could occur in many
ways, since bacterial cells contain a variety of iron-containing proteins that might
affect a crucial activity unrelated to ROS accumulation. Excluding off-target effects
using mutants is a challenge for events, such as the Fenton reaction, that are not
mediated by protein or RNA, because resistant mutants are not available as controls.
Nevertheless, ROS do not accumulate when nalidixic acid-treated cells are admin-
istered bipyridyl (G. Luan et al., unpublished observation).

20.11.2 Contribution of Infection

During infection, bacteria experience stress from host defense systems. One conse-
quence may be preparation of the pathogen for antimicrobial-mediated stress. For
example, host-derived reactive oxygen or reactive nitrogen species may induce bac-
terial protective systems that would counter bacterial production of ROS induced by
antimicrobials. According to the scheme presented in Fig. 20.1, suppression of
ROS-mediated killing would constitute a form of bacterial tolerance.
Heritable, genetic changes that increase antimicrobial tolerance can be selected
both in vitro and in vivo. Indeed, genome-wide maps of tolerance-associated genes,
the “tolerome,” have been described [88] that include numerous genes and pathways
20 Antimicrobial-Mediated Bacterial Suicide 637

that slow bacterial growth. Particularly interesting are mutations in global transcrip-
tional regulators, because they can produce a rapid, pleiotropic phenotypic effect:
they may constitute a prominent mechanism underlying tolerance. A clinically rel-
evant example of a global regulator of tolerance is seen with S. aureus mutants that
are deficient in the quorum-sensing agr regulon (see also Chap. 14). Paradoxically,
defects in this virulence factor are selected during serious hospital infection and are
associated with worse outcome [89, 90]. Recent work indicates that Δagr mutants
are less readily killed by gentamicin, ciprofloxacin, and several other lethal stressors
without affecting MIC [91], thereby conforming to a tolerance phenotype. Wild-
type agr normally downregulates a peroxidase that would otherwise protect from
ROS-mediated quinolone lethality. In the absence of agr and downregulation, cip-
rofloxacin loses some lethality. With gentamicin and daptomycin, modulation of S.
aureus tolerance by Δagr involves leakage of bacterial components that affect anti-
microbial lethality [91, 92]. Thus, mutation of global regulators during clinical
infection can result in multiple phenotypes associated with tolerance that are not
detected as resistance.

20.12 Concluding Remarks

The initial assertion that ROS contribute to the lethal action of multiple antimicro-
bial classes [25] has expanded to include a contribution of ROS to several severe
stress conditions. The idea that bacteria produce toxic ROS supports the more gen-
eral concept that bacteria respond to lethal stress by self-destructing. Existing data,
both direct and indirect, lead to the following scenario: when bacteria experience
severe stress, they make a live-or-die decision based on the response of several
bifunctional genes that are protective at low levels of stress but destructive at high
levels (Fig. 20.2). Destruction is achieved by ROS accumulation exceeding a thresh-
old beyond which genes designed to protect from oxidative stress fail to halt the
self-amplification of ROS. How bacterial suicide may confer a selective advantage
upon bacterial populations is currently unknown (see [4]).
The bifunctional nature of the lethal stress-response system makes application of
the principles challenging, because lethal treatment may be either enhanced or
diminished by increasing oxidative stress. For example, elevation of intracellular
superoxide levels prior to drug exposure may be protective, because the cells are less
susceptible to antimicrobials due to induction of protective genes. However, if stress
is severe, ROS accumulation will lead to cell death. Collectively, these observations
suggest that superoxide concentration, localization (intracellular, extracellular), and
regulation are complex. They may even change during the course of drug exposure
and infection. Nevertheless, several practical applications emerge from ROS contrib-
uting to antibacterial lethality. One is to avoid consuming antioxidants during ther-
apy with lethal antimicrobials, as antioxidants quench ROS. Another is that ROS
accumulation may represent a new form of interfering cross-reaction for tolerance
638 Y. Hong et al.

among seemingly unrelated drugs: drugs that reduce metabolism could inhibit the
production of ROS needed for full lethal activity of other antimicrobials.
Overall, consideration of ROS enriches our understanding of how bacteria
respond to severe stress and opens new ways to control bacterial populations.
Major Points
• Although lethal activity of current antimicrobials may be sufficient to clear most
infections, antimicrobial lethality needs to be even more rapid and extensive to
restrict the induction of resistance.
• Bacteria exhibit a lethal, ROS-mediated stress response that contributes to the
lethal activity of multiple antimicrobial classes; bacteriostatic assays of antimi-
crobial activity are often insensitive to perturbation of lethal stress responses.
• The effects of ROS are transient, which makes measurements of MBC uninfor-
mative with respect to rapid killing and suppression of induction of new resistant
mutants.
• Suppression of the ROS-mediated lethal stress response with antioxidants may
contribute to the emergence of resistance.
• Modulation of the ROS-mediated lethal stress response offers a new way to con-
trol bacterial populations.

Acknowledgments We thank the following for critical comments on the manuscript: Marila
Gennaro and Bo Shopsin.

References

1. Dwyer D, Collins J, Walker G. Unraveling the physiological complexities of antibiotic lethal-

ity. Annu Rev Pharmacol Toxicol. 2015;55:9.1–9.20.
2. Zhao X, Drlica K. Reactive oxygen species and the bacterial response to lethal stress. Curr
Opin Microbiol. 2014;21:1–6.
3. Zhao X, Hong Y, Drlica K. Moving forward with ROS involvement in antimicrobial lethality.
J Antimicrob Chemother. 2015;70:639–42.
4. Lewis K. Programmed death in bacteria. Microbiol Molecular Biol Rev. 2000;64:503–14.
5. Craig WA. Pharmacodynamics of antimicrobials: general concepts and applications. In:
Nightingale C, Murakawa T, Ambrose P, editors. Antimicrobial pharmacodynamics in theory
and clinical practice. New York: Marcel Dekker; 2002. p. 1–22.
6. Craig W. Pharmacokinetic/pharmacodynamic parameters: rationale for antibacterial dosing of
mice and men. Clin Infect Dis. 1998;26:1–12.
7. Miller C, Thomsen L, Gaggero C, Mosseri R, Ingmer H, Cohen S. SOS response induction by
β-lactams and bacterial defense against antibiotic lethality. Science. 2004;305:1629–31.
8. Malik M, Hoatam G, Chavda K, Kerns R, Drlica K. Novel approach for comparing quinolones
for emergence of resistant mutants during quinolone exposure. Antimicrob Agents Chemother.
2010;54:149–56.
9. Cirz R, Romesberg F. Induction and inhibition of ciprofloxacin resistance-conferring muta-
tions in hypermutator bacteria. Antimicrob Agents Chemother. 2006;50:220–5.
10. Zhao X, Drlica K. Restricting the selection of antibiotic-resistant mutants: a general strategy
derived from fluoroquinolone studies. Clin Infect Dis. 2001;33(Suppl 3):S147–S56.
20 Antimicrobial-Mediated Bacterial Suicide 639

11. Drlica K, Zhao X. Mutant selection window hypothesis updated. Clin Infect Dis. 2007;44:681–8.
12. Baym M, Lieberman T, Kelsic E, Chait R, Gross R, Yelin I, et al. Spatiotemporal microbial
evolution on antibiotic landscapes. Science. 2016;353:1147–51.
13. Cui J, Liu Y, Wang R, Tong W, Drlica K, Zhao X. The mutant selection window demonstrated
in rabbits infected with Staphylococcus aureus. J Infect Dis. 2006;194:1601–8.
14. Ni W, Song X, Cui J. Testing the mutant selection window hypothesis with Escherichia coli
exposed to levofloxacin in a rabbit tissue cage infection model. Eur J Clin Microbiol Infect
Dis. 2014;33:385–9.
15. Zhang B, Gu X, Li Y, Li X, Gu M, Zhang N, et al. In vivo evaluation of mutant selection
window of cefquinome against Escherichia coli in piglet tissue-cage model. BMC Vet Res.
2014;10:297.
16. Xiong M, Wu X, Ye X, Zhang L, Zeng S, Huang Z, et al. Relationship between cefquinome
PK/PD parameters and emergence of resistance of Staphylococcus aureus in rabbit tissue-cage
infection model. Front Microbiol. 2016;7:874.
17. Levin-Reisman I, Ronin I, Gefen O, Braniss I, Shoresh N, Balaban N. Antibiotic tolerance
facilitates the evolution of resistance. Science. 2017;355:826–30.
18. Wang X, Zhao X. Contribution of oxidative damage to antimicrobial lethality. Antimicrob
Agents Chemother. 2009;53:1395–402.
19. Dorsey-Oresto A, Lu T, Mosel M, Wang X, Salz T, Drlica K, et al. YihE kinase is a central
regulator of programmed cell death in bacteria. Cell Rep. 2013;3:528–37.
20. Dwyer D, Belenky P, Yang J, MacDonald I, Martell J, Takahashi N, et al. Antibiotics induce
redox-related physiological alterations as part of their lethality. Proc Natl Acad Sci U S A.
2014;111:E2100–E9.
21. Hong Y, Li L, Luan G, Drlica K, Zhao X. Contribution of reactive oxygen species to thymine-
less death in Escherichia coli. Nat Microbiol. 2017;(in press).
22. Burger R, Drlica K. Superoxide protects Escherichia coli from bleomycin mediated lethality.
J Inorg Biochem. 2009;109:1273–7.
23. Mosel M, Li L, Drlica K, Zhao X. Superoxide-mediated protection of Escherichia coli from
antimicrobials. Antimicrob Agents Chemother. 2013;57:5755–9.
24. Liu Y, Imlay J. Cell death from antibiotics without the involvement of reactive oxygen species.
Science. 2013;339:1210–3.
25. Kohanski M, Dwyer D, Hayete B, Lawrence C, Collins J. A common mechanism of cellular
death induced by bactericidal antibiotics. Cell. 2007;130:797–810.
26. Ezraty B, Vergnes A, Banzhaf M, Duverger Y, Huguenot A, Brochado A, et al. Fe-S clus-
ter biosynthesis controls uptake of aminoglycosides in a ROS-less death pathway. Science.
2013;340:1583–7.
27. Greenberg JT, Monach P, Chou JH, Josephy PD, Demple B. Positive control of a global anti-
oxidant defense regulon activated by superoxide-generating agents in Escherichia coli. Proc
Natl Acad Sci U S A. 1990;87(16):6181–5.
28. Oethinger M, Podglajen I, Kern WV, Levy SB. Overexpression of the marA or soxS regulatory
gene in clinical topoisomerase mutants of Escherichia coli. Antimicrob Agents Chemother.
1998;42(8):2089–94.
29. Koutsolioutsou A, Martins EA, White DG, Levy SB, Demple B. A soxRS-constitutive muta-
tion contributing to antibiotic resistance in a clinical isolate of Salmonella enterica (Serovar
typhimurium). Antimicrob Agents Chemother. 2001;45(1):38–43.
30. Goswami M, Mangoli SH, Jawali N. Involvement of reactive oxygen species in the action of
ciprofloxacin against Escherichia coli. Antimicrob Agents Chemother. 2006;50(3):949–54.
31. Goswami M, Mangoli SH, Jawali N. Effects of glutathione and ascorbic acid on streptomycin
sensitivity of Escherichia coli. Antimicrob Agents Chemother. 2007;51(3):1119–22.
32. Albesa I, Becerra MC, Battan PC, Paez PL. Oxidative stress involved in the antibacterial action
of different antibiotics. Biochem Biophys Res Commun. 2004;317(2):605–9.
33. Becerra MC, Albesa I. Oxidative stress induced by ciprofloxacin in Staphylococcus aureus.
Biochem Biophys Res Commun. 2002;297(4):1003–7.
640 Y. Hong et al.

34. Wang X, Zhao X, Malik M, Drlica K. Contribution of reactive oxygen species to pathways of
quinolone-mediated bacterial cell death. J Antimicrob Chemother. 2010;65:520–4.
35. Howard BM, Pinney RJ, Smith JT. 4-quinolone bactericidal mechanisms. Arzneimittelforschung/
Drug Res. 1993;43:1125–9.
36. Chen C-R, Malik M, Snyder M, Drlica K. DNA gyrase and topoisomerase IV on the bacterial
chromosome: quinolone-induced DNA cleavage. J Mol Biol. 1996;258:627–37.
37. Zhao X, Malik M, Chan N, Drlica-Wagner A, Wang J-Y, Li X, et al. Lethal action of quino-
lones with a temperature-sensitive dnaB replication mutant of Escherichia coli. Antimicrob
Agents Chemother. 2006;50:362–4.
38. Malik M, Hussain S, Drlica K. Effect of anaerobic growth on quinolone lethality with
Escherichia coli. Antimicrob Agents Chemother. 2007;51:28–34.
39. Wu X, Wang X, Drlica K, Zhao X. A toxin-antitoxin module in Bacillus subtiltis can both
mitigate and amplify effects of lethal stress. PLoS One. 2011;6:e23909.
40. Li L, Hong Y, Luan G, Mosel M, Malik M, Drlica K, et al. Ribosomal elongation factor 4
promotes cell death associated with lethal stress. MBio. 2014;5:e01708.
41. Pogliano J, Lynch A, Belin D, Lin E, Beckwith J. Regulation of Escherichia coli cell envelope
proteins involved in protein folding anddegradation by the Cpx two-component system. Genes
Dev. 1997;11:1169–82.
42. Raivio T, Silhavy T. Transduction of envelope stress in Escherichia coli by the Cpx two-­
component system. J Bacteriol. 1997;179:7724–33.
43. Kohanski M, Dwyer D, Wierzbowski J, Cottarel G, Collins J. Mistranslation of membrane
proteins and two-component system activation trigger antibiotic-mediated cell death. Cell.
2008;135:679–90.
44. Raivio T, Silhavy T. Periplasmic stress and ECF sigma factors. Annu Rev Microbiol.
2001;55:591–624.
45. Wu Y, Vulic M, Keren I, Lewis K. Role of oxidative stress in persister tolerance. Antimicrob
Agents Chemother. 2012;56:4922–6.
46. Kobayashi S, Ueda K, Komano T. The effects of metal ions on the DNA damage induced by
hydrogen peroxide. Agric Biol Chem. 1990;54:69–76.
47. Grimsrud P, Xie H, Griffin T, Bernlohr D. Oxidative stress and covalent modification of protein
with bioactive aldehydes. J Biol Chem. 2008;283:21837–41.
48. Girotti A. Lipid hydroperoxide generation, turnover, and effector action in biological systems.
J Lipid Res. 1998;39:1529–42.
49. Foti J, Devadoss B, Winkler J, Collins J, Walker G. Oxidation of the guanine nucleotide pool
underlies cell death by bactericidal antibiotics. Science. 2012;336:315–9.
50. Keren I, Wu Y, Inocencio J, Mulcahy L, Lewis K. Killing by bactericidal antibiotics does not
depend on reactive oxygen species. Science. 2013;339:1213–6.
51. Liu Y, Liu X, Qu Y, Wang X, Li L, Zhao X. Inhibitors of reactive oxygen accumulation delay
and/or reduce the lethality of several antistaphylococcal agents. Antimicrob Agents Chemother.
2012;56:6048–58.
52. Imlay J. Diagnosing oxidative stress in bacteria: not as easy as you might think. Curr Opin
Microbiol. 2015;24:124–31.
53. Mahoney TF, Silhavy TJ. The Cpx stress response confers resistance to some, but not all, bac-
tericidal antibiotics. J Bacteriol. 2013;195:1869–74.
54. Korshunov S, Imlay J. Detection and quantification of superoxide formed within the periplasm
of Escherichia coli. J Bacteriol. 2006;188:6326–34.
55. Gonzalez-Flecha B, Demple B. Metabolic sources of hydrogen peroxide in aerobically grow-
ing Escherichia coli. J Biol Chem. 1995;270(23):13681–7.
56. Messner K, Imlay J. Mechanism of superoxide and hydrogen peroxide formation by fuma-
rate reductase, succinate dehydrogenase, and aspartate oxidase. J Biol Chem. 2002;277:
42563–71.
57. Korshunov S, Imlay J. Two sources of endogenous hydrogen peroxide in Escherichia coli. Mol
Microbiol. 2010;75:1389–401.
20 Antimicrobial-Mediated Bacterial Suicide 641

58. Dwyer D, Kohanski M, Hayete B, Collins J. Gyrase inhibitors induce an oxidative damage
cellular death pathway in Escherichia coli. Mol Syst Biol. 2007;3:91. Epub.
59. Baek S, Li A, Sassetti C. Metabolic regulation of mycobacterial growth and antibiotic sensitiv-
ity. PLoS Biol. 2011;(5):e1001065.
60. Thomas V, Kinkead L, Janssen A, Schaeffer C, Woods K, Lindgren J, et al. A dysfunctional tri-
carboxylic acid cycle enhances fitness of Staphylococcus epidermidis during β-lactam stress.
MBio. 2013;4:e00437–13.
61. Davies B, Kohanski M, Simmons L, Winkler J, Collins J, Walker G. Hydroxyurea induces
hydroxyl radical-mediated cell death in Escherichia coli. Mol Cell. 2009;36:845–60.
62. Lobritz M, Belenky P, Porter C, Gutierrez A, Yang J, Schwarz E, et al. Antibiotic efficacy is
linked to bacterial cellular respiration. Proc Natl Acad Sci U S A. 2015;112:8173–80.
63. Meylan S, Porter C, Yang J, Belenky P, Gutierrez A, Lobritz M, et al. Carbon sources tune
antibiotic susceptibility in Pseudomonas aeruginosa via tricarboxylic acid cycle control. Cell
Chem Biol. 2017;24:195–206.
64. Park D, Akhtar M, Ansari A, Landick R, Kiley P. The bacterial response regulator ArcA
uses a diverse binding site architecture to regulate carbon oxidation globally. PLoS Genet.
2013;9:e1003839.
65. Loui C, Chang A, Lu S. Role of the ArcAB two-component system in the resistance of
Escherichia coli to reactive oxygen stress. BMC Microbiol. 2009;9:183.
66. Lu S, Killoran P, Fang F, Riley L. The global regulator ArcA controls resistance to reactive
nitrogen and oxygen intermediates in Salmonella enterica serovar Enteritidis. Infect Immun.
2002;70:451–61.
67. Deponte M. Programmed cell death in protists. Biochim Biophys Acta. 2008;1783(7):1396–405.
68. Jimenez C, Capasso JM, Edelstein CL, Rivard CJ, Lucia S, Breusegem S, et al. Different
ways to die: cell death modes of the unicellular chlorophyte Dunaliella viridis exposed to
various environmental stresses are mediated by the caspase-like activity DEVDase. J Exp Bot.
2009;60(3):815–28.
69. Bayles KW. Bacterial programmed cell death: making sense of a paradox. Nat Rev Microbiol.
2014;12(1):63–9.
70. Dwyer DJ, Camacho DM, Kohanski MA, Callura JM, Collins JJ. Antibiotic-induced bac-
terial cell death exhibits physiological and biochemical hallmarks of apoptosis. Mol Cell.
2012;46(5):561–72.
71. Gautam S, Sharma A. Involvement of caspase-3-like protein in rapid cell death of Xanthomonas.
Mol Microbiol. 2002;44(2):393–401.
72. Raju KK, Gautam S, Sharma A. Molecules involved in the modulation of rapid cell death in
Xanthomonas. J Bacteriol. 2006;188(15):5408–16.
73. Crumplin GC, Smith JT. Nalidixic acid: an antibacterial paradox. Antimicrob Agents
Chemother. 1975;8:251–61.
74. Malik M, Zhao X, Drlica K. Lethal fragmentation of bacterial chromosomes mediated by
DNA gyrase and quinolones. Mol Microbiol. 2006;61:810–25.
75. Hanawalt P. A balanced perspective on unbalanced growth and thymineless death. Front
Microbiol. 2015;6:504.
76. Khodursky AGE, Hanawalt PC. Thymineless death lives on: new insights into a classic phe-
nomenon. Annu Rev Microbiol. 2015;69:247–63.
77. Fonville N, Bates D, Hastings P, Hanawalt P, Rosenberg S. Role of RecA and the SOS response
in thymineless death in Escherichia coli. PLoS Genet. 2010;6:e1000865.
78. Fonville N, Vaksman Z, DeNapoli J, Hastings P, Rosenberg S. Pathways of resistance to thy-
mineless death in Escherichia coli and the function of UvrD. Genetics. 2011;189:23–36.
79. Sangurdekar D, Hamann B, Smirnov D, Srienc F, Hanawalt P, Khodursky A. Thymineless
death is associated with loss of essential genetic information from the replication origin. Mol
Microbiol. 2010;75:1455–67.
642 Y. Hong et al.

80. Kuong K, Kuzminov A. Disintegration of nascent replication bubbles during thymine star-
vation triggers RecA- and RecBCD-dependent replication origin destruction. J Biol Chem.
2012;287:23958–70.
81. Stivala L, Savio M, Carafoli F, Perucca P, Bianchi L, Maga G, et al. Specific structural determi-
nants are responsible for the antioxidant activity and the cell cycle effects of resveratrol. J Biol
Chem. 2001;276:22586–94.
82. Radimer K, Bindewald B, Hughes J, Ervin B, Swanson C, Picciano M. Dietary supplement use
by US adults: data from the National Health and nutrition examination survey, 1999–2000. Am
J Epidemiol. 2004;160:339–49.
83. Marathe S, Kumar R, Ajitkumar P, Nagaraja V. DC. Curcumin reduces the antimicrobial
activity of ciprofloxacin against Salmonella typhimurium and Salmonella typhi. J Antimicrob
Chemother. 2013;68:139–52.
84. Aires V, Delmas D. Common pathways in health benefit properties of RSV in cardiovascular
diseases, cancers and degenerative pathologies. Curr Pharmaceut Biotech. 2015;16:219–44.
85. Granzotto A, Zatta P. Resveratrol and Alzheimer's disease: message in a bottle on red wine and
cognition. Front Aging Neurosci. 2014;6:95.
86. Yang X, Li X, Ren J. From French paradox to cancer treatment: anti-cancer activities and
mechanisms of resveratrol. Anti Cancer Agents Med Chem. 2014;14:806–25.
87. Liu Y, Zhou J, Qu Y, Yang X, Shi G, Wang X, et al. Resveratrol antagonizes antimicrobial
lethality and stimulates ecovery of bacterial mutants. PLoS One. 2016;11:e0153023.
88. Brauner A, Fridman O, Gefen O, Balaban N. Distinguishing between resistance, tolerance and
persistence to antibiotic treatment. Nat Rev Microbiol. 2016;14:320–30.
89. Fowler V, Sakoulas G, McIntyre L, Meka V, Arbeit R, Cabell C, et al. Persistent bacteremia due
to methicillin-resistant Staphylococcus aureus infection is associated with agr dysfunction and
low-level in vitro resistance to thrombin-induced platelet microbicidal protein. J Infect Dis.
2004;190:1140–9.
90. Schweizer M, Furuno J, Sakoulas G, Johnson J, Harris A, Shardell M, et al. Increased mortality
with accessory gene regulator (agr) dysfunction in Staphylococcus aureus among bacteremic
patients. Antimicrob Agents Chemother. 2011;55:1082–7.
91. Kumar K, J Chen, Drlica K, Shopsin B. Dysfunction of the agr virulence regulator modulates
antimicrobial-mediated killing of Staphylococcus aureus. MBio. 2017;in press.
92. Pader V, Hakim S, Painter K, Wigneshweraraj S, Clarke T, Edwards A. Staphylococcus aureus
Inactivates daptomycin by releasing membrane phospholipids. Nat Microbiol. 2016;2:16194.
Chapter 21
PK/PD-Based Prediction of “Anti-­Mutant”
Antibiotic Exposures Using In Vitro
Dynamic Models

Alexander A. Firsov, Yury A. Portnoy, and Stephen H. Zinner

21.1 Introduction

Antibiotic resistance is emerging among a wide variety of pathogens responsible for

both common and infrequent infections. This has stimulated a search for so-called
“anti-mutant” antibiotic dosing regimens, i.e., regimens that might prevent or
severely restrict the enrichment of resistant mutant subpopulations that can be
designed from drug concentration-resistance relationships. Clinical reports con-
cerning the selection of antibiotic-resistant mutants and/or loss in susceptibility of
bacterial pathogens during treatment are too sparse to allow delineation of these
concentration-resistance relationships. For this reason, PK/PD (pharmacokinetic/
pharmacodynamic) relationships with the emergence of bacterial resistance have
been studied in vitro using dynamic models and, to a lesser extent, in vivo, using
animal models. Comprehensive reviews of these studies published recently [1, 2]
cover most of the in vitro and in vivo studies on this topic. Both reviews indirectly
reflect discrepancies and even contradictions among results reported from various
study groups. Given this situation and, especially, controversial interpretations of
resistance data by some authors, we address whether the inconsistencies among the
reported findings are real or apparent. We also highlight shortcomings and/or limita-
tions in study design and analysis that might have led to unjustified conclusions.
Since the contribution of in vitro studies to the current knowledge of concentration-­
resistance relationships seems precise, data obtained in dynamic models are ana-
lyzed in this chapter, with special emphasis on the mutant selection window (MSW)

A. A. Firsov (*) · Y. A. Portnoy

Gause Institute of New Antibiotics, Moscow, Russia
e-mail: [email protected]
S. H. Zinner
Mount Auburn Hospital, Harvard Medical School, Cambridge, MA, USA

© Springer International Publishing AG, part of Springer Nature 2018 643

I. W. Fong et al. (eds.), Antimicrobial Resistance in the 21st Century, Emerging Infectious
Diseases of the 21st Century, https://doi.org/10.1007/978-3-319-78538-7_21
644 A. A. Firsov et al.

hypothesis [3, 4]. In this regard, the present chapter is an extension of an earlier
publication [5].

21.2  ntibiotic Concentration-Bacterial Resistance

A
Relationships

Targeted bacterial resistance studies using in vitro dynamic models [6–16] were
launched at the turn of the new millennium. In most of the resistance studies, loss in
susceptibility of antibiotic-exposed pathogens and/or the enrichment of resistant
mutant subpopulations was related to the ratio of 24-h area under the concentration-­
time curve (AUC24) to the MIC. The fact that AUC24/MIC ratio rather than AUC24
per se related to susceptibility and/or population analysis data implies that AUC24/
MIC is an inter-strain resistance predictor that allows generalization of findings
obtained with different bacterial strains.
Contrary to expectations, most of these early works did not reveal clear relation-
ships between antibiotic exposure, expressed by the AUC24/MIC ratio, and the loss
in susceptibility of antibiotic-exposed pathogen cultures and/or the enrichment of
resistant mutants. For example, similar low resistance frequencies were reported in
an in vitro staphylococcal study at AUC24/MICs for ciprofloxacin that varied by a
factor of 16 [7]. Moreover, in the same study, frequency of resistance in
Staphylococcus aureus exposed to norfloxacin was paradoxically more pronounced
at a relatively high AUC24/MIC ratio (55 h) than at a low value of AUC24/MIC (3 h).
Obviously, in the absence of relationships between AUC24/MIC and loss in suscep-
tibility of antibiotic-exposed pathogens and/or the enrichment of resistant mutant
subpopulations, the AUC24/MIC thresholds reported to protect against the enrich-
ment of resistant mutants [8–15] are questionable.
These failures result from shortcomings in study design summarized elsewhere
[5], among which are the absence of resistant mutants in the starting inoculum,
insufficient duration of simulated treatments, simulation of AUC24/MIC ratios that
provide either sub-optimal or super-optimal effects of antibiotics on resistant
subpopulation(s) but not intermediate AUC24/MICs ratios, and inappropriate data
analysis. In particular, simple AUC24/MIC relationships with the emergence of
resistance – the greater the AUC24/MIC ratio, the less pronounced the enrichment of
resistant mutants or loss in susceptibility of antibiotic-exposed bacteria – were
reported in some studies, although more complex relationships between antibiotic
concentration and bacterial resistance are expected based on the “mutant selection
window” (MSW) hypothesis [3, 4]. According to this idea, resistant mutants are
selectively enriched at antibiotic concentrations between the MIC and the mutant
prevention concentration (MPC) but not at concentrations below the MIC or above
the MPC.
Strong support for the window hypothesis was first provided by an in vitro
dynamic model study in which S. aureus was exposed to 3-day dosing with
21 PK/PD-Based Prediction of “Anti-Mutant” Antibiotic Exposures Using In Vitro… 645

c­ iprofloxacin, gatifloxacin, levofloxacin, or moxifloxacin such that peak concentra-

tions were equal to the MIC, between the MIC and MPC and above the MPC [17].
Loss in fluoroquinolone susceptibility, expressed by the ratio of postexposure MIC
(MICfinal) to the pre-exposure MIC (MICinitial), was observed at concentrations that
fell inside the MSW but not at concentrations below the MIC or above the MPC. By
using the combined data for the four fluoroquinolones administered over a wide
range of the AUC24/MIC ratio, the AUC24/MIC relationship to MICfinal/MICinitial was
seen as a bell-shaped curve fitted by a Gaussian-type function. The loss in suscepti-
bility of S. aureus occurred at AUC24/MIC ratios of 25–100 h but not at AUC24/MICs
<15 h or > 200 h.
Given the bell-shaped pattern of the AUC24/MIC relationship with MICfinal/
MICinitial, the previous failures to correlate resistance with AUC24/MIC using linear
or log-linear regression are understandable. Moreover, the bell-shaped AUC24/MIC-­
resistance relationships established with fluoroquinolone-exposed S. aureus [17]
allowed us to take a fresh look at earlier studies that failed to link bacterial resis-
tance manifestations with antibiotic exposures. For example, despite the use of dif-
ferent endpoints of resistance (resistance frequency in a staphylococcal study with
norfloxacin and ciprofloxacin [7] and susceptibility measurements in a study with
four fluoroquinolones [17]), data were consistent with a bell-shaped curve when
MICfinal/MICinitial was plotted against AUC24/MIC. Indeed, the higher resistance fre-
quency seen at a relatively high AUC24/MIC ratio for norfloxacin (55 h) compared
to the lower frequency at a low AUC24/MIC (3 h) can be explained by the former
AUC24/MIC value corresponding to the maximum on the bell-shaped curve with the
latter AUC24/MIC value corresponding to the beginning of the ascending portion of
the MICfinal/MICinitial-AUC24/MIC curve [17]. Low ciprofloxacin resistance frequen-
cies at AUC24/MIC of 11 and 178 h are also explainable: the former value corre-
sponds to the beginning of the ascending portion of the curve and the latter to the
end of descending portion of the bell-shaped MICfinal/MICinitial-AUC24/MIC curve.
Thus, apparently paradoxical data reported with ciprofloxacin and norfloxacin [7]
are explained by the specific shape of the AUC24/MIC-resistance curves.
Bell-shaped curves that reflect AUC24/MIC-MICfinal/MICinitial relationships also
have been observed with S. aureus exposed to levofloxacin [18, 19] and the investi-
gational fluoroquinolone ABT 492 [18]. Unlike simulations of the pharmacokinet-
ics of the four fluoroquinolones described above [17], when AUC24/MIC-MICfinal/
MICinitial curves were superimposed, similar curves for S. aureus exposed to levo-
floxacin and ABT 492 did not superimpose. Bell-shaped AUC24/MIC relationships
have also been observed with 48-hour colony counts of resistant S. aureus in a
garenoxacin study [20]. Fragments of a bell-shaped curve of the AUC24/MIC-­
dependent population analysis profile (PAP) index were observed in a study that
exposed Streptococcus pneumoniae and Pseudomonas aeruginosa to moxifloxacin
[21], and a bell-shaped curve has been reported with moxifloxacin-exposed S. pneu-
moniae [22]. As with S. aureus [17], the most pronounced loss in susceptibility of
S. pneumoniae to the fluoroquinolone was observed at the intermediate AUC24/MIC
ratios of 40–50 h. Using population analysis data, a similar bell-shaped relationship
was established between AUC24/MIC and the ratio of mutation frequency (f)
646 A. A. Firsov et al.

observed with moxifloxacin-exposed S. pneumoniae (ffinal) to the frequency of muta-

tion before antibiotic dosing (finitial). The ffinal/finitial ratio (mutants resistant to 4 × MIC)
was maximal at the intermediate value of AUC24/MIC of 40 and 60 h.
Qualitatively similar AUC24/MIC-resistance relationships have been established
with fluoroquinolone-exposed Gram-negative bacteria [20, 23–28]. Unlike resis-
tance studies with Gram-positive bacteria, population analysis data were used in
most Gram-negative studies. With population data, not only posttreatment sizes of
resistant subpopulations [20], which are insufficiently informative [29, 30], but inte-
gral evaluation of the time courses of resistant mutant enrichment over the entire
simulated treatment [24–28] was used. The area under the mutant concentration-­
time curve (AUBCM) [29] served as an endpoint of bacterial resistance and was
related to simulated AUC24/MIC ratios. With AUBCM-based analysis, bell-shaped
relationships were observed with ciprofloxacin-exposed E. coli (four strains: [24,
25]), K. pneumoniae (three strains [27]), and P. aeruginosa (four strains [26, 28]). A
similar pattern was observed with the AUC24/MIC relationship using 48-hour col-
ony counts for resistant P. aeruginosa in the above-mentioned garenoxacin study
(one strain) [20] and with the MICfinal/MICinitial ratio as an endpoint in a study that
exposed E. coli to two veterinary fluoroquinolones – enrofloxacin and marbofloxa-
cin (data presented for one strain [23]). Also, bell-shaped relationships between
MICfinal/MICinitial and AUC24/MIC could be reconstructed [17] based on data reported
in a study with levofloxacin- and trovafloxacin-exposed Bacteroides fragilis [15].
Thus, fluoroquinolones clearly exhibit a bell-shaped relationship between drug
exposure and mutant enrichment or loss in bacterial susceptibility for both Gram-­
positive and Gram-negative bacteria.
Although most resistance studies have been performed with fluoroquinolones,
bell-shaped curves showing AUC24/MIC-dependent changes in MICfinal/MICinitial
and in the ratio of maximal-to-initial numbers of resistant mutants (Nmax/Ninitial)
have been reported with daptomycin- and vancomycin-exposed S. aureus (two
strains) [31]. Similar curves were observed when plotting AUBCM against AUC24/
MIC for daptomycin [32]. A bell-shaped relationship was also observed with resis-
tant E. coli subpopulation densities after a 24-h exposure to q.i.d fosfomycin [33].
Moreover, a bell-shaped relationship between AUBCM and AUC24/MIC was
recently established with linezolid-exposed S. aureus (three strains) [34]. It is
worth noting that for linezolid studies, it was necessary to include a small number
of resistant cells in the starting inoculum [35], since multiple copies of the line-
zolid target are likely to exist and therefore require spontaneous mutants to have
multiple mutations, a rare event.
Overall, the findings mentioned above indicate that the relationship of AUC24/
MIC with resistance is bell shaped for antibiotics of several classes and with both
Gram-positive and Gram-negative pathogens. Thus, AUC24/MIC can be used to pre-
dict the emergence of resistance.
21 PK/PD-Based Prediction of “Anti-Mutant” Antibiotic Exposures Using In Vitro… 647

21.3  lternative Predictors of Resistant Bacteria

A
Enrichment

Although the AUC24/MIC ratio has been related to the enrichment of resistant bacte-
rial mutants, it is not the only predictor for the emergence of antibiotic resistance.
Other predictors that emerge from the mutant window hypothesis are the ratio of
AUC24 to the MPC, the time during which antibiotic concentration is inside the
MSW (TMSW), and the time when drug concentration is above the MPC (T>MPC).
Each is briefly discussed below.

21.3.1 AUC24/MPC

From the standpoint of the MSW hypothesis, the AUC24/MPC ratio rather than the
AUC24/MIC might better predict the enrichment of resistant mutants and/or loss in
susceptibility of antibiotic-exposed bacteria [36–38] because MPC directly mea-
sures mutant susceptibility, which in many cases does not correlate well with bulk
culture MIC. Nevertheless, AUC24/MIC- and AUC24/MPC-resistance relationships
for a given organism can differ only in quantitative, not in qualitative terms (bell-­
shaped curves shifted along the x-axis). Consequently, the predictive potentials of
AUC24/MIC and AUC24/MPC ratios can be distinguished most clearly by their abil-
ity to serve as inter-strain predictors of resistance.
Obviously, bacterial resistance studies that utilize at least two bacterial strains
with different MPC/MIC ratios are needed to distinguish between AUC24/MPC and
AUC24/MIC as potential predictors of resistance. To compare the abilities of AUC24/
MPC and AUC24/MIC as inter-strain predictors of resistance, two strains of S.
aureus with distinctly different MPC/MIC (4 versus 16) were used in a study that
simulated twice-daily dosing of ciprofloxacin for 3 days [29]. When comparing the
descending portions of the AUC24/MPC and AUC24/MIC relationships with AUBCM
over a wide range of drug exposure, the AUC24/MIC plots were more stratified than
the respective AUC24/MPC plots. For example, with mutants resistant to 4 × MIC
ciprofloxacin, the square correlation coefficient for the AUBCM against log AUC24/
MPC relationship was 1.6-fold greater (r2 0.70) than for the AUBCM against log
AUC24/MIC relationship (r2 0.43). Even greater differences between AUC24/MPC
and AUC24/MIC relationships were reported with mutants resistant to 8 × MIC of
antibiotic (r2 0.72 versus 0.35). Figure 21.1 shows a systematic increase in the pre-
dictive power of AUC24/MPC and a concomitant decrease in the predictive power of
AUC24/MIC with an increase in culture MIC. These findings suggest that the AUC24/
MPC ratio is a more potent inter-strain predictor for staphylococcal resistance to the
fluoroquinolone than the AUC24/MIC ratio. This implies lower strain-to-strain vari-
ability in AUC24/MPC thresholds that prevent mutant enrichment. Less variation in
the “anti-mutant” thresholds is desired because clinical recommendations need to
be suitable for many strains.
648 A. A. Firsov et al.

Fig. 21.1 MPC- and MIC-related pharmacokinetic variables as predictors of the enrichment of
ciprofloxacin-resistant mutants of S. aureus. (Reconstructed from Ref. [29])

Fig. 21.2 AUC24/MIC- and AUC24/MPC-dependent resistance of levofloxacin-exposed S. aureus.

(Reconstructed from Ref. [19])

The distinct advantages of the AUC24/MPC over the AUC24/MIC ratio were dem-
onstrated subsequently in a similarly designed study with levofloxacin-exposed S.
aureus [19]. In this work three strains with the same MIC for levofloxacin but with
distinctly different MPCs (MPC/MIC from 8 to 64) were treated with once-daily
fluoroquinolone for 3 days. According to our analysis, plotting MICfinal/MICinitial
against either AUC24/MPC or AUC24/MIC did not allow combination of data
obtained with individual S. aureus strains: both AUC24/MPC and AUC24/MIC rela-
tionships with MICfinal/MICinitial were too stratified to be combined. However, quali-
tative characteristics of resistance, i.e., the loss in susceptibility (posttreatment MIC
elevation) or the absence of such a loss, were better correlated to AUC24/MPC than
to AUC24/MIC in a strain-independent manner. Reconstructed from reported data
[19], Fig. 21.2 demonstrates bacterial strain specificity of the AUC24/MIC-resistance
relationships in contrast to the strain-independent AUC24/MPC-resistance
21 PK/PD-Based Prediction of “Anti-Mutant” Antibiotic Exposures Using In Vitro… 649

Fig. 21.3 Strain-to-strain

variability in the “anti-­
mutant” thresholds of
AUC24/MIC (■) and
AUC24/MPC ( ).
(Reconstructed from Ref.
[24, 27–29])

r­ elationship. As seen in the figure, unlike the stratified AUC24/MIC plots observed
with individual strains, the respective AUC24/MPC are virtually superimposed.
In contrast to fluoroquinolone-exposed S. aureus, with Gram-negative bacteria,
correlations between AUBCM, which reflects the enrichment of ciprofloxacin-­
resistant mutants, and simulated AUC24/MPC were not as strong as between AUBCM
and AUC24/MIC. The respective r2s with ciprofloxacin-exposed Escherichia coli
were 0.69 versus 0.86 [24], with Klebsiella pneumoniae they were 0.72 versus 0.76
[27], and with P. aeruginosa they were 0.65 versus 0.75 for the AUC24/MPC and
AUC24/MIC ratios [28]. This difference between Gram-negative and Gram-positive
bacteria could reflect strain-to-strain variability in the “anti-mutant” thresholds. As
seen in Fig. 21.3, with each Gram-negative situation, the scattering of the “anti-­
mutant” AUC24/MPC was more pronounced than with AUC24/MIC: with E. coli it
was 4-fold versus 2-fold, with K. pneumoniae it was 25-fold versus 2-fold, and with
P. aeruginosa it was 26-fold versus 5-fold differences. Unlike the Gram-negative
bacteria tested, S. aureus strains exhibit less variable “anti-mutant” AUC24/MPC
ratios than the respective AUC24/MIC ratios (2.4-fold versus 3.8-fold differences).
In a recent study that exposed S. aureus strains to 5-day treatment with linezolid
(MPC/MIC from 2.5 to 5), a lower r2 (0.79) was also reported for the AUBCM rela-
tionship with AUC24/MPC than with AUC24/MIC (r2 0.91) [34]. However, strain-to-­
strain variability for the “anti-mutant” AUC24/MPC and AUC24/MIC ratios was
identical: twofold differences (from 48 to 96 h and from 120 to 240 h, respectively).
In another study that simulated 5-day treatments of two strains of S. aureus with
daptomycin (MPC/MIC from 3 to 5) and vancomycin (MPC/MIC from 3 to 8) [31],
the AUC24/MPC ratio was less predictive for S. aureus resistance than the AUC24/
MIC ratio. In contrast to reasonable AUC24/MIC relationships with Nmax/Ninitial (r2
0.68 for mutants resistant to 2 × MIC and r2 0.66 for mutants resistant to 4 × MIC
of the tested antibiotics) and MICfinal/MICinitial (r2 0.64), there was no correlation
between AUC24/MPC and either population analysis or susceptibility data. Based on
650 A. A. Firsov et al.

Fig. 21.4 AUC24/MIC- and AUC24/MPC-dependent resistance of S. aureus to five fluoroquino-

lones: ciprofloxacin (◊), gatifloxacin (▽), gemifloxacin (△), levofloxacin (□), moxifloxacin
(◯). (Reconstructed from Ref. [39])

data obtained with two S. aureus strains exposed to daptomycin and vancomycin,
MICfinal/MICinitial plots against AUC24/MPC were characterized by widely scattered
points. Thus, preference for a particular parameter may vary according to the
antibiotic-­pathogen pairs studied.
In this section we confined ourselves to studies containing clear evidence on the
advantages or disadvantages of AUC24/MPC over AUC24/MIC. There are other stud-
ies in which the conclusion that one of the predictors of bacterial resistance (usually
AUC24/MPC ratio) is preferable was not supported by the experimental findings.
One example is a resistance study with two strains of S. aureus exposed to five fluo-
roquinolones (three AUC24/MPC ratios per one antibiotic-pathogen pair, MPC/MIC
ratios from 2 to 15) [39]. According to the author’s statement, AUC24/MPC was
recognized as “the only parameter to correlate with the development of resistance”
although this correlation was extremely weak (r2 0.2). As seen in Fig. 21.4, when
plotting MICfinal/MICinitial against simulated AUC24/MPC or AUC24/MIC ratios, a
cloud of scattered points is observed using both potential predictors for emergence
of staphylococcal resistance. It is possible that this scatter could be avoided by using
a wider range of simulated AUC24/MPC or AUC24/MIC ratios and treatments longer
than 48 h. At least 72-hour treatments were recommended in our resistance studies
with fluoroquinolones [32]. Short observation times (24 h) could have affected the
results even more for a single-dose study with ciprofloxacin-exposed E. coli (three
strains with MPC/MIC from 4 to 16) [40]. For this reason, the authors’ conclusion
that “AUC/MPC ratio was the single pharmacodynamic index that predicted preven-
tion of resistant mutant development” should be taken with caution.
Thus, contrary to expectations, use of AUC24/MPC as an inter-strain predictor of
resistance has an advantage over AUC24/MIC only with some antibiotic-pathogen
pairs (fluoroquinolone-exposed S. aureus) but not with other pairs (fluoroquinolone-­
exposed E. coli, K. pneumoniae, P. aeruginosa, daptomycin- and vancomycin-­
exposed S. aureus). Further studies are needed to understand apparent differences
among bacterial species and antimicrobials.
21 PK/PD-Based Prediction of “Anti-Mutant” Antibiotic Exposures Using In Vitro… 651

21.3.2 TMSW

Unlike AUC24/MIC, which is commonly accepted as a predictor for the emergence

of bacterial resistance, TMSW (time in the mutant selection window) has been used
only infrequently, even though this parameter is closely linked to the MSW hypoth-
esis. A sigmoid TMSW relationship with resistance was discovered in the study,
described above, that exposed S. aureus to four fluoroquinolones [17]. The MICfinal/
MICinitial increased systematically with increases in TMSW when expressed as a per-
centage of the dosing interval. Subsequently, similar relationships were observed
with gatifloxacin-exposed S. aureus when simulating normal and impaired elimina-
tion pharmacokinetics (half-lives 7 and 31 h, respectively) [41]. At both half-lives,
the TMSW plots of the MICfinal/MICinitial ratio were sigmoid in shape, but they were
different for normal and impaired elimination of gatifloxacin: at a given TMSW, the
loss in susceptibility of antibiotic-exposed S. aureus was more pronounced in the
normal than in the impaired case, showing pharmacokinetic profile-dependent
emergence of resistance. Similar relationships were also observed with daptomy-
cin- and vancomycin-exposed S. aureus [31] using population analysis and suscep-
tibility data: a systematic increase in the MICfinal/MICinitial and Nmax/Ninitial was
associated with longer TMSWs. These findings suggest that the TMSW may be an addi-
tional predictor of bacterial resistance.
Against this background, other studies [40, 42–44] called into question the rele-
vance of TMSW for predicting loss in susceptibility and/or the enrichment of resistant
mutants. Our analysis, described below, indicates that conclusions drawn in these
studies are incorrect, and/or they are unsupported by the reported data. For example,
the false impression that TMSW does not predict S. aureus resistance [42] resulted
from the unjustified combination of data obtained in simulations of conventional
dosing regimens in which ciprofloxacin concentrations exceeded the MIC plus con-
stant rate infusions in which antibiotic concentrations were close to the MIC
(1.2 × MIC), a situation that may be described as providing TMSW of either 100% or
0% of the dosing interval. Meanwhile, by plotting MICfinal/MICinitial against TMSW
achieved in simulations of only conventional, intermittent ciprofloxacin dosing, a
reasonable TMSW-resistance relationship could be established (Fig. 21.5). Therefore,
the authors’ conclusion for the lack of “a clear relationship between TMSW and the
degree of resistance” [42] contradicts their own data. A similar relationship can be
seen with ciprofloxacin-exposed E. coli [40], at least for one of three strains for
which quantitative resistance data were presented (Fig. 21.6). Consequently, the
conclusion drawn by the authors about the lack of “a simple relationship between
TMSW and the prevention of the emergence of resistance” is not supported by the
experimental findings. We conclude that the data reported in these two studies [40,
42] are more in support of, rather than against, using TMSW as a predictor of bacterial
resistance.
In another study, measurements with isoniazid-exposed Mycobacterium tubercu-
losis led to the conclusion that “TMSW does not predict the emergence of resistance”
[43]. Given the antibiotic half-life-dependent relationships of TMSW with resistance
652 A. A. Firsov et al.

Fig. 21.5 TMSW-dependent

loss in susceptibility of S.
aureus 8043 (◯) and 8282
(△) exposed to twice-­
daily (white symbols) and
thrice-daily (black
symbols) ciprofloxacin.
(Reconstructed from Ref.
[42]) fitted by equation:
Y = Y0 + a/{1 + exp
[−(x – x0)/b]} (Eq. 1).
Y0 = 1, a = 20.13,
b = 3.000, x0 = 55.26

Fig. 21.6 TMSW-dependent

resistance of E. coli Nu14
to ciprofloxacin (half-life
4 h). (Reconstructed from
Ref. [40]) fitted by Eq. 1:
Y0 = 0, x0 = 29.52,
a = 7.134, b = 1.727

[41], this conclusion was the result of inappropriate combination of data obtained in
simulations of isoniazid pharmacokinetics in fast and slow acetylators (half-lives
1.8 and 4.2 h, respectively). Quite possibly fast and slow acetylator data might relate
to different TMSW-resistance relationships. On the other hand, there were insufficient
data to establish a specific relationship for each type of pharmacokinetic profile
(two points per one profile only). Moreover, in both cases, TMSW varied over very
small ranges: from 30% to 54% (fast acetylator simulations) and from 80% to 100%
(slow accelerator simulations) of the dosing interval. Based on these limited data,
delineation of a relationship or lack of a relationship is questionable.
A less clear situation is found with a study [44] in which TMSW failed to be predic-
tive for moxifloxacin and levofloxacin resistance with S. pneumoniae (four strains
exposed to 3-day treatments with the fluoroquinolones). In this case the unsuccess-
ful attempts to relate susceptibility of antibiotic-exposed bacteria with TMSW might
21 PK/PD-Based Prediction of “Anti-Mutant” Antibiotic Exposures Using In Vitro… 653

have resulted from the overestimation of MPCs, at least for a S. pneumoniae strain
that exhibited a biphasic pattern in the frequency-concentration curve. As shown in
our study with ciprofloxacin-exposed E. coli [24], the higher MPC derived from the
second phase of such curves describes the second-step mutations that might or
might not be present in pharmacokinetic simulations. As a result, the true value of
TMSW could be overestimated, and the value of T>MPC could be underestimated.
Moreover, in most simulations (e.g., in five out of six experiments with moxifloxa-
cin), there was no loss in susceptibility of S. pneumoniae. Because of the unbal-
anced study design, TMSW was used to explain the lack of resistance rather than the
emergence of resistance. We conclude that arguments against the predictive value of
TMSW are at best weak.
However, even in studies in which relationships between TMSW and emergence of
resistance were demonstrated, the predictive power of TMSW was always lower than
the AUC24/MIC ratio. For example, in the studies described above, with S. aureus
exposed to four fluoroquinolones, the respective r2s were 0.72 versus 0.90 [17], with
ciprofloxacin-exposed E. coli the r2s were 0.61 versus 0.84 [25], and with
ciprofloxacin-­exposed P. aeruginosa they were 0.56 versus 0.80 [45]. With
doripenem-­exposed P. aeruginosa r2s were 0.69 versus 0.80 [45], and with
glycopeptide-­exposed S. aureus they were 0.60 versus 0.68 (MICfinal/MICinitial data)
or 0.50 versus 0.64 (Nmax/Nmin data) [31].
The differences in predictive power described above may be due to relating
AUC24/MIC to TMSW while ignoring information about the position of simulated
antibiotic concentrations inside the MSW, a feature that is likely to be very impor-
tant with respect to mutant amplification. To test this hypothesis, the enrichment of
ciprofloxacin-resistant S. aureus was examined at drug concentrations that oscil-
lated near the MPC, i.e., close to the top of the MSW (“upper case”), or close to the
MIC, i.e., at the lower limit of the MSW (“lower case”). In both cases the TMSW was
the same [46]. In this study, two methicillin-resistant strains of S. aureus (MPC/
MIC 4 and 16) were exposed to twice-daily ciprofloxacin for 3 consecutive days.
The simulated AUC24/MIC were 50 h (“lower case”) and 260 h (“upper case”) to
provide TMSW of 75% of the dosing interval with one strain and 30 h (“lower case”)
and 100 h (“upper case”) to provide TMSW of 56% with another strain. With each
strain, AUBCM (a measure of mutant enrichment) observed in the “lower case” was
much greater than in the “upper case,” thereby showing less pronounced enrichment
of ciprofloxacin-resistant staphylococci at antibiotic concentrations oscillating near
the MPC than near the MIC, even though for each strain TMSW was the same.
Heterogeneity of the MSW was further examined in a study that exposed four
Escherichia coli strains to twice-daily ciprofloxacin dosing for 3 days [25]. To
explore the different predictive powers of TMSW and AUC24/MIC, the enrichment of
ciprofloxacin-resistant E. coli mutants was studied at wide ranges of TMSWs and
AUC24/MICs (up to eight points per strain). Peak antibiotic concentrations were
simulated to be close to the MIC, between the MIC and MPC, and above the MPC;
TMSW varied from 0% to 100% of the dosing interval. The amplification (enrich-
ment) of resistant mutants was monitored by plating on media with 8 × MIC of the
antibiotic. With each organism, TMSW plots of the AUBCM split into two portions,
654 A. A. Firsov et al.

Fig. 21.7 TMSW-dependent changes in the AUBCM reflecting the enrichment of E. coli mutants
resistant to 4 × MIC of ciprofloxacin fitted by Eq. 1, separately for points that belong to the ascend-
ing portion (Y0 = 0, x0 = 20.18, a = 459.1, b = 15.86) and descending portion (Y0 = 0, x0 = 83.40,
a = 500.1, b = 6.969) of the AUBCM-AUC24/MIC curve. AUC24/MIC values are shown in callouts.
(Reconstructed from Ref. [25])

one for antibiotic concentrations below the MPC (T>MPC = 0) and the other for
c­ oncentrations consistently above the MPC (T>MPC > 0). The result was a hysteresis
loop. Figure 21.7 illustrates a TMSW relationship with AUBCM observed with one of
the E. coli strains examined. As seen in the figure, when antibiotic concentrations
were below the MPC (points corresponding to the ascending portion of the bell-­
shaped AUBCM-AUC24/MIC curve – AUC24/MIC ratios of 15, 30 and 60 h), the
AUBCM at a given TMSW was greater than at the same TMSW relevant to the descend-
ing portion of the AUBCM-AUC24/MIC curve (AUC24/MIC ratios of 360 and 720 h
gave the same TMSW).
The distinct T>MPC-dependent splitting of the AUBCM-TMSW curves (Fig. 21.7)
prevents consideration of data obtained at T>MPC = 0 and at T>MPC > 0 as a single data
set. When the data with the four E. coli strains were combined, a sigmoid function
fits well with AUBCM versus TMSW data sets taken separately at T>MPC = 0 and
T>MPC > 0 (r2s 0.81 and 0.92, respectively). In both cases, correlation of TMSW with
resistance appeared to be of the same power as observed with the AUC24/MIC ratio
(r2 0.84). In contrast to the separated analysis of the TMSW data referring to the condi-
tions of T>MPC = 0 or T>MPC > 0, fitting the whole data pool while ignoring T>MPC
exhibited a weaker correlation between TMSW and mutant enrichment (r2 0.61).
Hysteresis loops have also been reported for TMSW relationships with S. aureus
resistance to linezolid [47]. Using inocula of three methicillin-resistant S. aureus
strains (MIC of linezolid = 2 mg/L), spiked with low concentrations of previously
selected resistant mutants (MIC, 8 mg/L), AUC24/MIC- and TMSW-dependent mutant
enrichment was observed in 5-day treatments with twice-daily linezolid. With each
strain, TMSW relationships with the AUBCM (for mutants resistant to 4 × MIC) exhib-
21 PK/PD-Based Prediction of “Anti-Mutant” Antibiotic Exposures Using In Vitro… 655

ited a hysteresis loop, with the upper sigmoid corresponding to T>MPC = 0, and the
lower one to the T>MPC > 0. Based on combined data obtained with the three bacterial
strains, AUBCM correlated better with TMSW data taken separately when T>MPC was
zero or exceeded zero (r2 0.99) than with pooled data ignoring T>MPC (r2 0.24).
We conclude that a hysteresis loop is inherent in the TMSW relationships with
mutant enrichment. It is very likely that the incorrect combination of data obtained
at T>MPC = 0 and at T>MPC > 0 is among the reasons for an underestimation of the true
role of TMSW as a predictor of the emergence of bacterial resistance. For example, in
a resistance study with meropenem-exposed Acinetobacter baumannii [48], the
conclusion that TMSW is not a suitable parameter relating to mutant enrichment might
result from inappropriate combining TMSWs belonging to the upper (T>MPC = 0) and
lower (T>MPC > 0) portions of the hysteresis loop. With each A. baumannii strain, the
TMSWs observed at the minimal antibiotic exposure met the condition T>MPC = 0,
whereas the TMSWs at the maximal exposure met the situation in which T>MPC > 0.
Overall, although TMSW is mutually related to the MSW, the appropriate use of
this parameter requires consideration of the T>MPC data.

21.3.3 T>MPC

The available reports on the use of T>MPC as a predictor of bacterial resistance are
much less frequent than those that report AUC24/MIC, AUC24/MPC, and TMSW, in
part because antibiotic concentrations simulated in these studies exceeded the MPCs
for only a short time or did not reach the MPCs. Even in cases in which T>MPCs were
positive, the reported data [39, 40] are too limited to delineate quantitative T>MPC
relationships with the enrichment of resistant mutants. However, unlike the staphy-
lococcal study with five fluoroquinolones [39], a reasonable link between the emer-
gence of bacterial resistance (qualitative characteristics only) and T>MPC can be seen
from the E. coli study using ciprofloxacin [40]. With each of three E. coli strains, at
least in simulations of ciprofloxacin pharmacokinetics having a half-life of 4 h, the
emergence of resistance was consistently associated with lower T>MPC. Apparently,
the authors’ conclusion that the emergence of bacterial resistance cannot be pre-
dicted by the T>MPC reflects the inability to combine data obtained with different E.
coli strains: at the same T>MPC, resistance to ciprofloxacin developed with one strain
but not with another.
In another study, suppression of A. baumannii resistance to meropenem (again,
qualitative characteristics – the presence or absence of resistant mutants of antibiotic-­
exposed bacteria) was achieved for two strains with MPC/MIC ratios of approxi-
mately 15 and 60 [48] at similar T>MPCs. It is noteworthy that strain-independent
T>MPC-resistance relationships could be established for each mode of antibiotic
administration (0.5- and 3-h infusions). These relationships are specific for the type
of simulated pharmacokinetic profile: the protective T>MPC was lower in the longer
656 A. A. Firsov et al.

Fig. 21.8 T>MPC-dependent

resistance of A. baumannii
CSRA24 (◯) and
CSRA91 (△) exposed
with meropenem (0.5-h
infusion – open symbols,
solid lines 3-h infusion –
filled symbols, dotted
lines). (Reconstructed from
Ref. [48])

than in shorter meropenem infusions (Fig. 21.8). Such data are further evidence for
pharmacokinetic profile-dependent emergence of bacterial resistance.
A quantitative resistance index, AUBCM, was first related to T>MPC in the above-
mentioned ciprofloxacin study with E. coli [25]. When AUBCM versus T>MPC data
sets for four strains of E. coli were combined, a mono-exponential decay function
fits these data with a relatively high r2 (0.71). Using the points that met the condition
of T>MPC > 0, similar correlations between AUBCM with AUC24/MPC (r2 0.74) and
with AUC24/MIC (r2 0.81) were observed. Thus, the predictive power of T>MPC was
not inferior to AUC24/MPC or to AUC24/MIC ratios.
Even stronger correlations were reported recently between AUBCM and T>MPC
with linezolid-exposed S. aureus [47]. A sigmoid function fits combined data for
three S. aureus strains with a high r2 (0.99). For the points that meet the condition
T>MPC > 0, the sum of TMSW and T>MPC equals 100% of the dosing interval, and the
T>MPC plot of AUBCM is a mirror image of the TMSW plot at T>MPC > 0 with the same
r2. In this study, both T>MPC and TMSW at T>MPC > 0 exhibited stronger correlations
with AUBCM than did AUC24/MPC (r2 0.80) and AUC24/MIC (r2 0.85).
Thus, together with AUC24/MIC, AUC24/MPC, and TMSW, T>MPC can be consid-
ered as a strain-independent predictor for the emergence of bacterial resistance.

21.4  linical Relevance of In Vitro Resistance Studies:

C
Predicted “Anti-Mutant” AUC24/MIC Ratios
Versus Clinically Attainable AUC24/MICs

Predicting the “anti-mutant” AUC24/MIC ratios relative to clinically attainable

AUC24/MICs is a primary goal of bacterial resistance studies with dynamic models.
Obviously, such predictions can be ensured only when reasonable AUC24/MIC rela-
tionships were established with mutant enrichment and/or changing susceptibility
21 PK/PD-Based Prediction of “Anti-Mutant” Antibiotic Exposures Using In Vitro… 657

of antibiotic-exposed bacteria. However, an “anti-mutant” AUC24/MIC ratio pre-

dicted from in vitro studies always represents a conservative target for dosing
adjustment because dynamic models do not consider host defense factors. Moreover,
unlike AUC24/MIC breakpoints used to determine the potential for an antibacterial
to kill susceptible subpopulations [49], the “anti-mutant” AUC24/MIC ratios pre-
dicted with in vitro studies cannot be referred to clinically established protective
AUC24/MICs because they have not been reported. Therefore, these predictions are
more conditional than are those with antibiotic effects on susceptible bacterial
subpopulations.

21.4.1 Monotherapy

The “anti-mutant” AUC24/MIC ratios were established when S. aureus was exposed
to fluoroquinolones [17]. Based on the bell-shaped AUC24/MIC relationships with
MICfinal/MICinitial, predicted “protective” AUC24/MICs appeared to be similar for
levofloxacin (201 h), moxifloxacin (222 h), gatifloxacin (241 h), and ciprofloxacin
(244 h). However, these thresholds are clinically attainable only with moxifloxacin.
With a 400 mg dose of moxifloxacin, the “anti-mutant” AUC24/MIC ratio is 66% of
the clinically attainable value, whereas with two 500 mg doses of ciprofloxacin, a
500 mg dose of levofloxacin, or a 400 mg dose of gatifloxacin, the respective anti-­
mutant AUC24/MICs are 420%, 220%, and 190% of the clinically attainable values.
Thus, at least against S. aureus, moxifloxacin is expected to protect against resis-
tance development in a clinical setting, whereas the three other fluoroquinolones
will likely enrich mutant subpopulations.
Resistance thresholds reported in vitro studies from different research groups
exhibit considerable variability. For example, with grepafloxacin-exposed S. pneu-
moniae, “protective” AUC24/MIC ratios varied from 32 h [14] to 80 h [10] while
those of levofloxacin were from 9 h [14] to 26 h [9] and 35 h [11]. Furthermore,
although moxifloxacin-resistant S. pneumoniae were not found at AUC24/MIC ratios
of 60 h [11] and 107 h [10], significant losses in susceptibility were seen at AUC24/
MICs as high as 43,500 h [14]. Analysis of these findings [5] indicates that different
estimates of the “anti-mutant” AUC24/MIC ratio can be attributed to differences in
study design and data processing. For this reason, it is of particular interest to com-
pare “anti-mutant” AUC24/MICs obtained under the same experimental conditions.
Based on data reported in ciprofloxacin resistance studies that determine “anti-­
mutant” AUC24/MIC ratios using the descending portion of the MICfinal/MICinitial or
AUBCM versus AUC24/MIC curve [17, 29, 50], lower resistance thresholds were
established with Gram-positive than with Gram-negative bacteria. The “anti-­
mutant” AUC24/MIC ratios were 125–244 h with S. aureus (three strains) [17, 29],
700–1100 h with E. coli (four strains), 1300–2600 h with K. pneumoniae (three
strains), and 300–1400 h with P. aeruginosa (four strains) [50]. However, when
related to clinically attainable AUC24/MIC ratios for each individual strain, different
distributions emerge for ciprofloxacin “anti-mutant” potentials for different species.
658 A. A. Firsov et al.

Fig. 21.9 “Anti-mutant” thresholds of AUC24/MIC related to the clinically attainable AUC24/MIC
ratios: ciprofloxacin against Gram-positive and Gram-negative bacteria. (Reconstructed from Ref.
[17, 29, 50])

As shown in Fig. 21.9, with E. coli but not with S. aureus, P. aeruginosa, and K.
pneumoniae, the predicted “anti-mutant” AUC24/MICs are achieved in a clinical
setting (ratio of the resistance threshold to the clinically achievable AUC24/MIC <1).
Our resistance study with daptomycin- and vancomycin-exposed S. aureus [31]
provides an example of a more favorable situation, at least for daptomycin. Based
on combined data obtained with two S. aureus strains, the predicted “anti-mutant”
AUC24/MIC ratio (200 h) was smaller than the clinically attainable AUC24/MIC90s:
380 h for a 4 mg/kg dose of daptomycin and 570 h for a 6 mg/kg dose of the antibi-
otic. Unlike daptomycin, the “anti-mutant” target for vancomycin virtually coin-
cided with the clinically attainable AUC24/MIC90 ratio (200 h for 1 gm twice-daily
dosing). Reasonably optimistic predictions were made in a linezolid study with S.
aureus [34]: with two of three studied strains, the “anti-mutant” AUC24/MIC ratios
were equal to the clinically attainable value (120 h for a 600 mg dose twice a day).
However, with the third S. aureus strain, the “anti-mutant” threshold was twofold
greater.
Overall, these examples show that clinically achievable AUC24/MIC ratios may
not always overlap the “anti-mutant” AUC24/MIC thresholds. In conjunction with
extensive MPC testing of various antibiotic-pathogen pairs [51], these findings pre-
dict emergence of bacterial resistance with the use of many existing antibiotics as
usually prescribed. As dose escalation is rarely possible due to limited patient toler-
ability, combined antibiotic therapy provides an alternative to the replacement of a
less “protective” agent by a more “protective” agent.
21 PK/PD-Based Prediction of “Anti-Mutant” Antibiotic Exposures Using In Vitro… 659

21.4.2 Combined Therapy

Resistance studies that expose bacteria to antibiotic combinations using in vitro

dynamic models have been relatively infrequent. In one example, to determine
whether doxycycline can minimize or prevent the emergence of staphylococcal
resistance to moxifloxacin, 5-day treatments with once-daily moxifloxacin or doxy-
cycline given alone or in several combinations were simulated over a threefold
AUC24/MIC range for each antibiotic [52]. Combined use of moxifloxacin and dox-
ycycline delayed the loss in susceptibility of S. aureus to both antibiotics at moder-
ate AUC24/MIC ratios or completely prevented the loss at relatively high AUC24/
MIC. In another study [53], a combination of twice-daily linezolid with once-daily
doxycycline against Enterococcus faecium was tested in three-day treatments.
Unlike linezolid alone, neither growth on linezolid-containing media (4×, 8× and
16 × MIC) nor changes in susceptibility occurred when combined use of these anti-
biotics was simulated. The presence of linezolid decreased the numbers of
doxycycline-­resistant enterococci present without any loss in susceptibility.
More recently, due to renewed interest in colistin, a series of papers on colistin
combinations with other agents have been published. For example, colistin combi-
nations with doripenem are protected against the emergence of colistin resistance in
colistin-susceptible and hetero-resistant but not in multidrug-resistant K. pneu-
moniae [54]. Also, combinations of colistin with fosfomycin reduced the probabil-
ity of development of Enterobacteriaceae resistant to both antibiotics [55].
Moreover, colistin prevented the enrichment of meropenem-resistant A. baumannii
in simulations of combined treatments with colistin + carbapenem [56] and the loss
of tigecycline susceptibility in A. baumannii treated with a tigecycline-colistin com-
bination [57]. With an endocardial vegetation model, ceftaroline and ceftriaxone but
not gentamicin prevented the enrichment of daptomycin-resistant Streptococcus
mitis [58]. With M. tuberculosis, linezolid-rifampicin combinations exhibited
greater “anti-mutant” power than the single agents in simulations of antibiotic phar-
macokinetics over wide dose ranges [59].
These combination studies may be useful for designing regimens to prevent and/
or restrict the emergence of bacterial resistance, but none provides a way to predict
these effects from independent tests that do not take into account antibiotic pharma-
cokinetics. One such way has been recently proposed in a resistance study with S.
aureus exposed to linezolid-rifampicin combinations [60]. Using antibiotic concen-
tration ratios that correspond to linezolid/rifampicin AUC24 ratios to be simulated in
a dynamic model, the MPCs of each agent in combination were determined. For
example, to determine the MPCs necessary for the prediction of the “anti-mutant”
effects of simulated antibiotic pharmacokinetics using a combination of clinical
daily doses of linezolid (1200 mg) and rifampicin (600 mg) providing AUC24 ratio
of 240/60 = 4, the respective “pharmacokinetically derived” antibiotic concentra-
tion ratio was 4 to 1. MPC testing at pharmacokinetically derived concentration
ratios showed a 2.5–3.1-fold decrease of linezolid MPC when in combination with
660 A. A. Firsov et al.

rifampicin and a 760–6400-fold decrease of rifampicin MPC in the presence of

linezolid.
Using a mixed inoculum of linezolid-susceptible and -resistant cells [35] for
pharmacokinetic simulations, combinations of linezolid with rifampicin completely
suppressed the enrichment of linezolid-resistant S. aureus mutants and restricted the
development of rifampicin resistance in S. aureus [60]. In contrast to simulated
combined treatments, mutants resistant to both antibiotics were enriched when line-
zolid and rifampicin were administered separately. These effects were likely due to
lowering the MPC of linezolid and rifampicin in the combinations relative to MPCs
of the single antibiotics. In essence, linezolid-rifampicin combinations provided
much longer times above the antibiotic MPCs (73–100% of the dosing interval for
linezolid and 42–58% for rifampicin) compared to the T>MPCs in mono-treatments
(0–44% for linezolid and 0% for rifampicin). Thus, the T>MPCs for antibiotic combi-
nations provided a quantitative description of how combined use of linezolid and
rifampicin restricts the enrichment of linezolid-resistant relative to rifampicin-­
resistant mutants.
It is possible that an MPC-based prediction of the “anti-mutant” potential for
linezolid and rifampicin combinations was successful due to pharmacokinetically
derived concentration ratios used to determine MPC of the antibiotics given in com-
bination. For each simulated dosing regimen, including clinically relevant dosing,
the MPC of each antibiotic was determined at the concentration ratio that strictly
corresponded to the ratio of AUC24s provided by a given linezolid-rifampicin com-
bination. As seen in Fig. 21.10, the MPC of linezolid, combined with rifampicin,
was independent of the antibiotic concentration ratio, whereas the MPC of rifampi-
cin, combined with linezolid, decreased systematically with increases in linezolid
concentrations in the combination. Therefore the MPCs reported in studies with
other antibiotic combinations at arbitrarily chosen concentration ratios [61–65]
might be insufficiently predictive for the “anti-mutant” effects.
Overall, the linezolid-rifampicin study [60] suggests that “anti-mutant” antibi-
otic combinations can be predicted by the MPCs determined at pharmacokinetically-­
based antibiotic concentration ratios. This approach avoids uncertainties about the
optimal choice of antibiotic concentration ratios, as occurs with checkerboard tech-
niques for susceptibility testing when the optimal concentration ratio may or may
not have any relationship to human antibiotic pharmacokinetics.

21.5 Conclusions

Analysis of the enrichment of resistant bacterial subpopulations using in vitro

dynamic models shows the usefulness of this approach to better understand PK/
PD-mediated enrichment of resistant mutants with concomitant loss in pathogen
susceptibility. These studies have contributed to the delineation of AUC24/MIC,
AUC24/MPC, TMSW, and T>MPC relationships with resistance and to the prediction of
“anti-mutant” thresholds and dosing regimens. However, current knowledge of
21 PK/PD-Based Prediction of “Anti-Mutant” Antibiotic Exposures Using In Vitro… 661

Fig. 21.10 MPC values of

linezolid and rifampicin
alone and in combinations.
(Reconstructed from Ref.
[60])

these relationships and their clinical relevance remain limited because of the scar-
city of dynamic resistance studies with many antibiotic classes and diverse patho-
gens. Indeed, only a few bacterial species have been examined; quantitative findings
reported with a limited number of pathogens will remain applicable only to those
antimicrobial-pathogen pairs and not to other strains of the same species until the
data are generalized. Moreover, further studies that compare inter-strain predictions
of mutant enrichment using AUC24/MIC and AUC24/MPC are particularly needed,
due to apparent differences between fluoroquinolone-exposed Gram-positive and
Gram-negative bacteria. Nevertheless, bacterial resistance studies using dynamic
models provide notable progress in understanding of the mutant selection window
as a framework for predicting the selective enrichment of resistant mutants.

Major Points
• Relationships between PK/PD (pharmacokinetic/pharmacodynamic) indices and
the emergence of bacterial resistance are a basis for designing “anti-mutant”
antibiotic dosing regimens, i.e., regimens that are expected to prevent or restrict
the enrichment of resistant mutant subpopulations.
662 A. A. Firsov et al.

• In vitro dynamic models provide a way to study the enrichment of resistant

mutants while simulating human antibiotic pharmacokinetics.
• Using these models, bell-shaped relationships between the ratios of the area
under the concentration-time curve (AUC) to the MIC or MPC (mutant preven-
tion concentration) and the enrichment of resistant mutants and/or loss in suscep-
tibility of antibiotic-exposed bacteria are established.
• The general pattern of these relationships is consistent with the mutant selection
window (MSW) hypothesis that predicts that the selection of resistant mutants
occurs largely at antibiotic concentrations between the MIC and MPC.
• Together with AUC/MIC and AUC/MPC ratios, times inside the MSW and above
the MPC can be predictive for the emergence of bacterial resistance.
• Based on the AUC/MIC-resistance relationships, the “anti-mutant” thresholds
were predicted for various “antibiotic-pathogen” pairs.
• For most cases examined, doses used clinically expose bacterial pathogens to
concentrations inside the MSW for much of the dosing interval, a feature that
reveals a fundamental dosing flaw with respect to the emergence of resistance.

References

1. Linder KE, Nicolau DP, Nailor MD. Predicting and preventing antimicrobial resistance utiliz-
ing pharmacodynamics: part I gram positive bacteria. Expert Opin Drug Metabol Toxicol.
2016;12:267–80.
2. Abdelraouf K, Linder KE, Nailor MD, Nicolau DP. Predicting and preventing antimicrobial
resistance utilizing pharmacodynamics: part II gram-negative bacteria. Expert Opin Drug
Metabol Toxicol, 2017; doi: https://doi.org/10.1080/17425255.2017.1329417.
3. Zhao X, Drlica K. Restricting the selection of antibiotic-resistant mutants: a general strategy
derived from fluoroquinolone studies. Clin Infect Dis. 2001;33:147–56.
4. Blondeau JM, Zhao X, Hansen G, Drlica K. Mutant prevention concentrations of fluoroqui-
nolones for clinical isolates of Streptococcus pneumoniae. Antimicrob Agents Chemother.
2001;45:433–8.
5. Firsov AA, Zinner SH. Lubenko IY. In: Nightingale CH, Ambrose PG, Drusano GL,
Murakawa T, editors. vitro dynamic models as tools to predict antibiotic pharmacodynamics,
Antimicrobial pharmacodynamics in theory and clinical practice. 2nd ed. New York: Informa
Healthcare USA, Inc.; 2007. p. 45–78.
6. Lacy MK, Lu W, Xu X, Tessier PR, Nicolau DP, Quintiliani R, Nightingale
CH. Pharmacodynamic comparisons of levofloxacin, ciprofloxacin, and ampicillin against
Streptococcus pneumoniae in an in vitro model of infection. Antimicrob Agents Chemother.
1999;43:672–7.
7. Aeschlimann JR, Kaatz GW, Rybak MJ. The effects of NorA inhibition on the activities of levo-
floxacin, ciprofloxacin and norfloxacin against two genetically related strains of Staphylococcus
aureus in an in-vitro infection model. J Antimicrob Chemother. 1999;44:343–9.
8. Peterson ML, Hovde LB, Wright DH, Hoang AD, Raddatz JK, Boysen PJ, Rotschafer
JC. Fluoroquinolone resistance in Bacteroides fragilis following sparfloxacin exposure.
Antimicrob Agents Chemother. 1999;43:2251–5.
9. Madaras-Kelly KJ, Demasters TA. In vitro characterization of fluoroquinolone concentra-
tion/MIC antimicrobial activity and resistance while simulating clinical pharmacokinetics of
levofloxacin, ofloxacin, or ciprofloxacin against Streptococcus pneumoniae. Diagn Microbiol
Infect Dis. 2000;37:253–60.
21 PK/PD-Based Prediction of “Anti-Mutant” Antibiotic Exposures Using In Vitro… 663

10. Coyle EA, Kaatz GW, Rybak MJ. Activities of newer fluoroquinolones against ciprofloxacin-­
resistant Streptococcus pneumoniae. Antimicrob Agents Chemother. 2001;45:1654–9.
11. Zhanel GG, Walters M, Laing N, Hoban DJ. In vitro pharmacodynamic modelling simulat-
ing free serum concentrations of fluoroquinolones against multidrug-resistant Streptococcus
pneumoniae. J Antimicrob Chemother. 2001;47:435–40.
12. Thorburn CE, Edwards DI. The effect of pharmacokinetics on the bactericidal activity of cip-
rofloxacin and sparfloxacin against Streptococcus pneumoniae and the emergence of resis-
tance. J Antimicrob Chemother. 2001;48:15–22.
13. Ross GH, Wright DH, Hovde LB, Peterson ML, Rotschafer JC. Fluoroquinolone resistance in
anaerobic bacteria following exposure to levofloxacin, trovafloxacin, and sparfloxacin in an in
vitro pharmacodynamic model. Antimicrob Agents Chemother. 2001;45:2136–40.
14. Klepser ME, Ernst EJ, Petzold CR, Rhomberg P, Doern GV. Comparative bactericidal activi-
ties of ciprofloxacin, clinafloxacin, grepafloxacin, levofloxacin, moxifloxacin, and trova-
floxacin against Streptococcus pneumoniae in a dynamic in vitro model. Antimicrob Agents
Chemother. 2001;45:673–8.
15. Peterson ML, Hovde LB, Wright DH, Brown GH, Hoang AD, Rotschafer
JC. Pharmacodynamics of trovafloxacin and levofloxacin against Bacteroides fragilis in an in
vitro pharmacodynamic model. Antimicrob Agents Chemother. 2002;46:203–10.
16. Wright DH, Gunderson SM, Hovde LB, Ross GH, Ibrahim AS, Rotschafer JC. Comparative
pharmacodynamics of three newer fluoroquinolones versus six strains of staphylococci in
an in vitro model under aerobic and anaerobic conditions. Antimicrob Agents Chemother.
2002;46:1561–3.
17. Firsov AA, Vostrov SN, Lubenko IY, Drlica K, Portnoy YA, Zinner SH. In vitro pharmaco-
dynamic evaluation of the mutant selection window hypothesis using four fluoroquinolones
against Staphylococcus aureus. Antimicrob Agents Chemother. 2003;47:1604–13.
18. Firsov AA, Vostrov SN, Lubenko IY, Arzamastsev AP, Portnoy YA, Zinner SH. ABT492 and
levofloxacin: comparison of their pharmacodynamics and their abilities to prevent the selec-
tion of resistant Staphylococcus aureus in an in vitro dynamic model. J Antimicrob Chemother.
2004;54:178–86.
19. Liang B, Bai N, Cai Y, Wang R, Drlica K, Zhao X. Mutant prevention concentration-based
pharmacokinetic/pharmacodynamic indices as dosing targets for suppressing the enrich-
ment of levofloxacin-resistant subpopulations of Staphylococcus aureus. Antimicrob Agents
Chemother. 2011;55:2409–12.
20. Tam VH, Louie A, Deziel MR, Liu W, Leary R, Drusano GL. The relationship between quino-
lone exposures and resistance amplification is characterized by an inverted U: a new paradigm
for optimizing pharmacodynamics to counter select resistance. Antimicrob Agents Chemother.
2007;51:744–7.
21. MacGowan AP, Rogers CA, Holt HA, Bowker KE. Activities of moxifloxacin against, and
emergence of resistance in, Streptococcus pneumoniae and Pseudomonas aeruginosa in an in
vitro pharmacokinetic model. Antimicrob Agents Chemother. 2003;47:1088–95.
22. Zinner SH, Lubenko IY, Gilbert D, Simmons K, Zhao X, Drlica K, Firsov AA. Emergence of
resistant Streptococcus pneumoniae in an in vitro dynamic model that simulates moxifloxacin
concentrations inside and outside the mutant selection window: related changes in susceptibil-
ity, resistance frequency and bacterial killing. J Antimicrob Chemother. 2003;52:616–22.
23. Gebru E, Choi M-J, Lee S-J, Damte D, Park SC. Mutant prevention concentration and mech-
anism of resistance in clinical isolates and enrofloxacin/marbofloxacin-selected mutants of
Escherichia coli of canine origin. J Med Microbiol. 2011;60:1512–22.
24. Firsov AA, Strukova EN, Shlykova DS, Portnoy YA, Kozyreva VK, Edelstein MV, Dovzhenko
SA, Kobrin MB, Zinner SH. Bacterial resistance studies using in vitro dynamic models: the
predictive power of the mutant prevention and minimum inhibitory antibiotic concentrations.
Antimicrob Agents Chemother. 2013;57:4956–62.
25. Firsov AA, Portnoy YA, Strukova EN, Shlykova DS, Zinner SH. Predicting bacterial resis-
tance using the time inside the mutant selection window: possibilities and limitations. Int
J Antimicrob Agents. 2014;44:301–5.
664 A. A. Firsov et al.

26. Firsov AA, Portnoy YA, Strukova EN, Shlykova DS, Zinner SH. Bacterial antibiotic resistance
studies using in vitro dynamic models: population analysis vs. susceptibility testing as end-
points of mutant enrichment. Int J Antimicrob Agents. 2015;46:313–8.
27. Strukova EN, Portnoy YA, Romanov AV, Edelstein MV, Zinner SH Firsov AA. Searching for
the optimal predictor of ciprofloxacin resistance in Klebsiella pneumoniae by using in vitro
dynamic models. Antimicrob Agents Chemother. 2016;60:1208–15.
28. Strukova EN, Portnoy YA, Zinner SH, Firsov AA. Predictors of bacterial resistance using
in vitro dynamic models: area under the concentration-time curve related to either the mini-
mum inhibitory or mutant prevention antibiotic concentration. J Antimicrob Chemother.
2016;71:678–84.
29. Firsov AA, Smirnova MV, Strukova EN, Vostrov SN, Portnoy YA, Zinner SH. Enrichment of
resistant Staphylococcus aureus at ciprofloxacin concentrations simulated within the mutant
selection window: bolus versus continuous infusion. Int J Antimicrob Agents. 2008;32:488–93.
30. Golikova MV, Strukova EN, Portnoy YA, Firsov AA. PK/PD modeling as a tool to predict
bacterial resistance to antibiotics: alternative analyses of the experimental data. Antibiot
Chimiother. 2015;60:11–6. rus
31. Firsov AA, Smirnova MV, Lubenko IY, Vostrov SN, Portnoy YA, Zinner SH. Testing the
mutant selection window hypothesis with Staphylococcus aureus exposed to daptomycin and
vancomycin in an in vitro dynamic model. J Antimicrob Chemother. 2006;58:1185–92.
32. Smirnova MV, Vostrov SN, Strukova EN, Dovzhenko SA, Kobrin MB, Portnoy YA, Zinner
SH, Firsov AA. The impact of duration of antibiotic exposure on bacterial resistance predic-
tions using in vitro dynamic models. J Antimicrob Chemother. 2009;64:815–20.
33. VanScoy BD, McCauley J, Ellis-Grosse EJ, Okusanya OO, Bhavnani SM, Forrest A, Ambrose
PG. Exploration of the pharmacokinetic-pharmacodynamic relationships for fosfomycin effi-
cacy using an in vitro infection model. Antimicrob Agents Chemother. 2015;59:7170–7.
34. Firsov AA, Alieva KN, Strukova EN, Golikova MV, Portnoy YA, Dovzhenko SA, Kobrin MB,
Romanov AV, Edelstein MV, Zinner SH. Testing the mutant selection window hypothesis
with Staphylococcus aureus exposed to linezolid in an in vitro dynamic model. J Antimicrob
Chemother. 2017;72:3100–7.
35. Firsov AA, Golikova MV, Strukova EN, Portnoy YA, Romanov AV, Edelstein MV, Zinner SH.
In vitro resistance studies with bacteria that exhibit low mutation frequencies: the prediction
of “antimutant” linezolid concentrations using a mixed inoculum containing both susceptible
and resistant Staphylococcus aureus. Antimicrob Agents Chemother. 2015;59:1014–9.
36. Drlica K, Zhao X. Mutant selection window hypothesis updated. Clin Infect Dis.
2007;44:681–68.
37. Zhao X, Drlica K. A unified anti-mutant dosing strategy. J Antimicrob Chemother.
2008;62:434–6.
38. Blondeau JM. New concepts in antimicrobial susceptibility testing: the mutant prevention con-
centration and mutant selection window approach. Vet Dermatol. 2009;20:383–96.
39. Allen GP, Kaatz GW, Rybak MJ. In vitro activities of mutant prevention concentration-­targeted
concentrations of fluoroquinolones against Staphylococcus aureus in a pharmacodynamic
model. Int J Antimicrob Agents. 2004;24:150–60.
40. Olofsson SK, Marcusson LL, Lindgren PK, Hughes D, Cars O. Selection of ciprofloxacin
resistance in Escherichia coli in an in vitro kinetic model: relation between drug exposure and
mutant prevention concentration. J Antimicrob Chemother. 2006;57:1116–21.
41. Firsov AA, Vostrov SN, Lubenko IY, Zinner SH, Portnoy YA. Concentration-dependent
changes in the susceptibility and killing of Staphylococcus aureus in an in vitro dynamic
model that simulates normal and impaired gatifloxacin elimination. Int J Antimicrob Agents.
2004;23:60–6.
42. Campion JJ, McNamara PJ, Evans ME. Evolution of ciprofloxacin-resistant Staphylococcus
aureus in in vitro pharmacokinetic environments. Antimicrob Agents Chemother.
2004;48:4733–44.
43. Gumbo T, Louie A, Liu W, Brown D, Ambrose PG, Bhavnani SM, Drusano GL. Isoniazid
bactericidal activity and resistance emergence: integrating pharmacodynamics and pharma-
21 PK/PD-Based Prediction of “Anti-Mutant” Antibiotic Exposures Using In Vitro… 665

cogenomics to predict efficacy in different ethnic populations. Antimicrob Agents Chemother.

2007;51:2329–36.
44. Homma T, Hori T, Sugimori G, Yamano Y. Pharmacodynamic assessment based on mutant
prevention concentrations of fluoroquinolones to prevent the emergence of resistant mutants
of Streptococcus pneumoniae. Antimicrob Agents Chemother. 2007;51:3810–5.
45. Zinner SH, Gilbert D, Greer K, Portnoy YA, Firsov AA. Сoncentration-resistance relationships
with Pseudomonas aeruginosa exposed to doripenem and ciprofloxacin in an in vitro model.
J Antimicrob Chemother. 2013;68:881–7.
46. Firsov AA, Lubenko IY, Smirnova MV, Strukova EN, Zinner SH. Enrichment of
fluoroquinolone-­resistant Staphylococcus aureus: oscillating ciprofloxacin concentrations
simulated at the upper and lower portions of the mutant selection window. Antimicrob Agents
Chemother. 2008;52:1924–8.
47. Alieva KN, Strukova EN, Golikova MV, Portnoy YA, Zinner SH, Firsov AA. Time inside the
mutant selection window as a predictor of staphylococcal resistance to linezolid. J Antibiot.
2017; (in press).
48. Li X, Wang L, Zhang XJ, Yang Y, Gong WT, Xu B, Zhu YQ, Liu W. Evaluation of meropenem
regimens suppressing emergence of resistance in Acinetobacter baumannii with human simu-
lated exposure in an in vitro intravenous-infusion hollow-fiber infection model. Antimicrob
Agents Chemother. 2014;58:6773–81.
49. Firsov AA, Lubenko IY, Vostrov SN, Portnoy YA, Zinner SH. Antistaphylococcal effect related
to the area under the curve/MIC ratio in an in vitro dynamic model: predicted breakpoints ver-
sus clinically achievable values for seven fluoroquinolones. Antimicrob Agents Chemother.
2005;49:2642–7.
50. Strukova EN, Portnoy YA, Zinner SH, Firsov AA. Species differences in ciprofloxacin resis-
tance among gram-negative bacteria: can “anti-mutant” ratios of the area under the concentra-
tion–time curve to the MIC be achieved clinically?. J Chemother. 2017; doi:https://doi.org/10
.1080/1120009X.2017.1335980.
51. Drlica K. The mutant selection window and antimicrobial resistance. J Antimicrob Chemother.
2003;52:11–7.
52. Firsov AA, Vostrov SN, Lubenko IY, Portnoy YA, Zinner SH. Prevention of the selection of
resistant Staphylococcus aureus by moxifloxacin plus doxycycline in an in vitro dynamic
model: an additive effect of the combination. Int J Antimicrob Agents. 2004;23:451–6.
53. Zinner SH, Gilbert D, Lubenko IY, Greer K, Firsov AA. Selection of linezolid-resistant
Enterococcus faecium in an in vitro dynamic model: protective effect of doxycycline.
J Antimicrob Chemother. 2008;61:629–6357.
54. Deris ZZ, Yu HH, Davis K, Soon RL, Jacob J, Ku CK, Poudyal A, Bergen PJ, Tsuji BT, Bulitta
JB, Forrest A, Paterson DL, Velkov T, Li J, Nation RL. The combination of colistin and doripe-
nem is synergistic against Klebsiella pneumoniae at multiple inocula and suppresses colis-
tin resistance in an in vitro pharmacokinetic/pharmacodynamic model. Antimicrob Agents
Chemother. 2012;56:5103–12.
55. Albur MS, Noel A, Bowker K, MacGowan A. The combination of colistin and fosfomycin is
synergistic against NDM-1-producing Enterobacteriaceae in in vitro pharmacokinetic/phar-
macodynamic model experiments. Int J Antimicrob Agents. 2015;46:560–7.
56. Liu X, Zhao M, Chen Y, Bian X, Li Y, Shi J, Zhang J. Synergistic killing by meropenem and
colistin combination of carbapenem-resistant Acinetobacter baumannii isolates from Chinese
patients in an in vitro pharmacokinetic/pharmacodynamic model. Int J Antimicrob Agents.
2016;48:559–63.
57. Cai X, Yang Z, Dai J, Chen K, Liu H, Ni W, Wei C, Cui J. Pharmacodynamics of tigecy-
cline alone and in combination with colistin against clinical isolates of multidrug-resistant
Acinetobacter baumannii in an in vitro pharmacodynamic model. Int J Antimicrob Agents.
2017;49:609–16.
58. Yim J, Smith JR, Singh NB, Rice S, Stamper K, Garcia de la Maria C, Bayer AS, Mishra NN,
Miró JM, Tran TT, Arias CA, Sullam P, Rybak MJ. Evaluation of daptomycin combinations
666 A. A. Firsov et al.

with cephalosporins or gentamicin against Streptococcus mitis group strains in an in vitro model
of simulated endocardial vegetations (SEVs). J Antimicrob Chemother. 2017;72:2290–6.
59. Drusano GL, Neely M, Van Guilder M, Schumitzky A, Brown D, Fikes S, Peloquin C, Louie
A. Analysis of combination drug therapy to develop regimens with shortened duration of treat-
ment for tuberculosis. PLoS One. 2014;9:1–10.
60. Firsov AA, Golikova MV, Strukova EN, Portnoy YA, Dovzhenko SA, Kobrin MB, Zinner
SH. Pharmacokinetically based prediction of the effects of antibiotic combinations on resis-
tant Staphylococcus aureus mutants: in vitro model studies with linezolid and rifampicin.
J Chemother. 2017;29:220–6.
61. Zhanel GG, Mayer M, Laing N, Adam HJ. Mutant prevention concentrations of levofloxacin
alone and in combination with azithromycin, ceftazidime, colistin (polymyxin E), merope-
nem, piperacillin-tazobactam, and tobramycin against Pseudomonas aeruginosa. Antimicrob
Agents Chemother. 2006;50:2228–30.
62. Cai Y, Yang J, Kan Q, Nie X, Wang R, Liang B, Bai N. Mutant prevention concentration of
colistin alone and in combination with levofloxacin or tobramycin against multidrug-resistant
Acinetobacter baumannii. Int J Antimicrob Agents. 2012;40:477–8.
63. Liu LG, Zhu YL, Hu LF, Cheng J, Ye Y, Li JB. Comparative study of the mutant prevention
concentrations of vancomycin alone and in combination with levofloxacin, rifampicin and fos-
fomycin against methicillin-resistant Staphylococcus epidermidis. J Antibiot. 2013;66:709–12.
64. Wei W, Yang H, Hu L, Ye Y, Li J. Activity of levofloxacin in combination with colistin against
Acinetobacter baumannii: in vitro and in a Galleria mellonella model. J Microbiol Immunol
Infect. 2015; doi: https://doi.org/10.1016/j.jmii.2015.10.010.
65. Wu J, Jiang TT, Su JR, Li L. Antimicrobial activity of linezolid combined with minocycline
against vancomycin-resistant Enterococci. Chin Med J. 2013;126:2670–5.
 Part IV
Bringing Compounds to Market
Chapter 22
The Role of Pharmacometrics
in the Development of Antimicrobial
Agents

Justin C. Bader, Elizabeth A. Lakota, Brian VanScoy,

Sujata M. Bhavnani, and Paul G. Ambrose

We live in a world teeming with antimicrobial-resistant pathogens. For a number of

pan-resistant pathogens, our once plentiful antimicrobial armamentarium is now
quite limited. There is a critical need for new antimicrobial agents to treat patients
with infections due to these highly resistant organisms such as Gram-negative
bacilli [1]. The need for new agents is especially great for the treatment of patient
populations at great risk for morbidity and mortality, such as those with hospital-­
acquired and ventilator-associated bacterial pneumonia (HABP and VABP, respec-
tively) arising from resistant pathogens.
Pharmacokinetic-pharmacodynamic (PK-PD) principles have recently become
an important cornerstone for antimicrobial agent assessment. The use of these prin-
ciples together with the broader science of pharmacometrics, a branch of science
that includes population pharmacokinetic (PK) and PK-PD analysis, has enabled
both early- and late-stage analyses supporting antimicrobial dose selection. These
data have served to greatly de-risk antimicrobial drug development and increase the
likelihood of regulatory success [2]. Our confidence in pharmacometric data stems,
in large measure, from the general concordance that exists between the results from
PK-PD analyses based on data from preclinical models of infection and those from
randomized clinical trials [3, 4]. Recent US Food and Drug Administration (US
FDA) and the European Medicines Agency guidance documents recommending the
use of PK and PK-PD analyses throughout the drug development process for a num-
ber of indications [5–11] demonstrate the reliance on pharmacometric data for regu-
latory pathways to develop antimicrobial agents.
The benefits of a pharmacometric approach are even more relevant when devel-
oping antimicrobial agents for the treatment of patients with multiple or extensively
drug-resistant (MDR and XDR, respectively) pathogens. Given the rarity of such

J. C. Bader · E. A. Lakota · B. VanScoy · S. M. Bhavnani (*) · P. G. Ambrose

Institute for Clinical Pharmacodynamics, Schenectady, NY, USA
e-mail: [email protected]

© Springer International Publishing AG, part of Springer Nature 2018 669

I. W. Fong et al. (eds.), Antimicrobial Resistance in the 21st Century, Emerging Infectious
Diseases of the 21st Century, https://doi.org/10.1007/978-3-319-78538-7_22
670 J. C. Bader et al.

pathogens, enrollment of patients in clinical trials for these agents can be slow and
often requires several years to accrue a modest level of enrollment. This severely
limits the amount of information available to conduct traditional statistical analyses
of clinical data. Moreover, it is unethical to enroll patients into a randomized clini-
cal trial of any design for which one treatment arm is not reliably active against
MDR or XDR pathogens. Consequently, if we are to develop antimicrobial agents
for the treatment of seriously ill patients infected with MDR and XDR pathogens,
the normal paradigm of basing antimicrobial approval on the results of multiple
randomized clinical studies is difficult to impossible. This is ultimately due to the
lack of comparators with suitable efficacy and low numbers of patients with these
infections. In this context, we must therefore consider other data to supply the evi-
dence necessary for antimicrobial drug approval. The focus of this chapter is not
only on basic pharmacometric concepts in the setting of pathogens with usual drug
resistance (UDR) but also on how pharmacometric analyses can be used to leverage
limited clinical data packages in order to support antimicrobial drug approval for
the treatment of patients with infections due to specified MDR or XDR pathogens.

22.1 The Bottom Line Upfront

The certainty that pharmacometric data provides to support antimicrobial agent

drug development begins in the laboratory. The answers to three critical preclinical
questions can be used to forecast the clinical efficacy and durability of an antimicro-
bial drug regimen. The questions posed, which can be answered using in vivo and
in vitro PK-PD infection models, include the following:
1. What is the PK-PD index that is most associated with efficacy?
2. What is the magnitude of the PK-PD index necessary for efficacy?
3. What is the relationship between antimicrobial drug exposure and time to emer-
gence of drug resistance?
An important goal for the drug development scientist is to leverage the relevant
preclinical PK-PD infection models to answer each question using an appropriate
model in the most time and cost-efficient manner. As discussed in Sect. 22.2, there
are a number of standard preclinical PK-PD infection models, including the neutro-
penic murine-thigh and murine-lung infection models and the one-compartment
in vitro infection model, which have been used to characterize the PK-PD of antimi-
crobial agents. The model chosen should be most appropriate to answer the scien-
tific question posed (Fig. 22.1). While a number of in vivo and in vitro PK-PD
infection models can be used to identify the PK-PD index most associated with
efficacy, infection models that can be used to characterize the relationship between
drug exposure and time to emergence of resistance are limited. The evaluation of
PK-PD relationships for emergence of resistance is best accomplished using the
in vitro hollow-fiber infection model.
In Sect. 22.3, the importance of developing and refining a population pharmaco-
kinetic (PK) model in order to inform and support programmatic decisions is
22 The Role of Pharmacometrics in the Development of Antimicrobial Agents 671

Fig. 22.1 The preclinical toolbox for antimicrobial drug development

reviewed. The value of these models and breadth of questions they are able to
answer will depend largely upon the richness of the data upon which they are built
and refined. This section will present the consideration needed for ensuring that one
is collecting data which are relevant for answering pivotal questions.
Section 22.4 describes the iterative process of dosing regimen selection. Analyses
to support early-stage dosing regimen selection integrate the aforementioned pre-
clinical information with healthy volunteer PK data using Monte Carlo simulation
in the context of the minimum inhibitory concentration (MIC) distribution(s) for the
target pathogen(s). These analyses should explicitly account for between-species
differences in PK, protein binding, and effect site exposures. The underlying popu-
lation PK model used for the simulations should be developed using a robust clini-
cal dataset, which initially includes data from healthy subjects after receiving single
and multiple doses, and is ultimately refined using data from special populations
and target patient populations treated with dosing regimens intended for labeling.
Finally, Sect. 22.5 will discuss the value of PK-PD analyses based on clinical data in
the context of clinical data packages in the setting of UDR or MDR and/or XDR. The
use of these data to confirm that adequate drug exposures relative to nonclinical PK-PD
targets for efficacy are achieved in the context of both robust and limited clinical data
packages is reviewed. The opportunities for evaluating PK-PD relationships for safety
endpoints and use these data to guide labeling and/or clinical practice guidelines will be
addressed. Finally, the concept of Bayesian analyses, which integrate preclinical and
clinical PK-PD information to inform clinical trial design questions, will be discussed.
672 J. C. Bader et al.

22.2 Assembling a Robust Preclinical PK-PD Data Package

The preclinical PK-PD package for a new drug application (NDA) serves as the
foundation for selecting and supporting dosing regimens for clinical study. These
data are vital to ensuring the success of any drug development program but should
be held in higher regard when developing antimicrobial agents to treat patients with
infections due to MDR and XDR organisms. As will be discussed in greater detail
in Sect. 22.5, clinical data are likely to be limited in such programs; thus, we must
put greater weight on preclinical data to increase regulatory certainty. Consequently,
as described herein, the selection, design, and execution of preclinical studies must
be thoughtfully planned to ensure a robust PK-PD data package is obtained. Such
data will then allow for more informative preclinical inputs for dose selection analy-
ses as described in Sect. 22.3.

22.2.1  etermining the PK-PD Index Most Associated

D
with Efficacy

To begin formulating a preclinical PK-PD data package, the first question which
must be asked and answered is in regard to the PK-PD index which is most associ-
ated with efficacy for a given antimicrobial. Antimicrobials are typically said to
exhibit concentration- or time-dependent patterns of killing activity [12]. In the case
of antimicrobials with concentration-dependent activity, the rate and extent of kill-
ing increase in tandem with drug concentrations. This pattern of activity is best
described using the ratios of the area under the concentration-time curve (AUC) or
maximum concentrations (Cmax) over the MIC (AUC:MIC and Cmax:MIC ratios,
respectively). The objective when dosing concentration-dependent antimicrobials is
to achieve exposures in patients which maximize the killing of pathogens while
minimizing the likelihood of witnessing drug-induced toxicities.
Alternatively, the objective when administering antimicrobials which exhibit
time-dependent killing is not to maximize drug exposures but rather to optimize
dosing to maintain drug concentrations above a target threshold such as an
MIC. Accordingly, this pattern of activity can be characterized by the percentage
of time drug concentrations remain above an MIC or other threshold (%T > MIC
and %T > threshold, respectively). Jointly, the AUC:MIC ratio, Cmax:AUC ratio,
and %T > MIC comprise the three most commonly utilized PK-PD indices to
describe antimicrobial activity (Fig. 22.2). Determining the PK-PD index which
best describes efficacy for a given antimicrobial can be challenging if one does
not take extra precautions. Given that the magnitude of drug introduced into a
system impacts all of the aforementioned PK-PD indices, significant collinearity
is observed when attempting to differentiate these indices on the basis of dose
alone. That is to say AUC:MIC ratio, Cmax:AUC ratio, and %T > MIC all increase
with dose. Dose-­fractionation studies are used to mitigate the impact of this col-
22 The Role of Pharmacometrics in the Development of Antimicrobial Agents 673

Fig. 22.2 PK-PD indices

depicted utilizing a plasma
concentration-time curve

Fig. 22.3 Relationships between change in log10 CFU from baseline at 24 h and total-drug
AUC:MIC ratio, Cmax:MIC ratio, and %T > MIC for anidulafungin against C. glabrata based on
data from a neutropenic murine candidiasis model [13]. (Reproduced from Ref. [14] with permis-
sion from J Antimicrob Chemother. Copyright © 2017 British Society for Antimicrobial
Chemotherapy, [Journal of Antimicrobial Chemotherapy, 2018; 73 (suppl 1):i44-50.])

linearity through the administration of dosing regimens which utilize the same
total dose of an antimicrobial agent but which are differentiated in their frequency
of dosing (e.g., 600 mg once daily, 300 mg twice daily, 150 mg four times daily,
etc.). A range of exposures is obtained by administering regimens in a similar
manner over multiple dose levels.
Using data obtained from a dose-fractionation study that was conducted using
a neutropenic murine candidiasis model [13], Fig. 22.3 shows relationships
between change in log10 CFU from baseline at 24 h and total-drug AUC:MIC
ratio, Cmax:MIC ratio, and %T > MIC for anidulafungin against C. glabrata [14].
674 J. C. Bader et al.

In this study, neutropenic mice were infected with Candida glabrata and admin-
istered 1 of 20 ­anidulafungin dosing regimens. Total doses of 1.25, 5, 20, 80, or
320 mg/kg were administered over a 96-h period in the form of one, two, four, or
six divided doses (i.e., doses were given every 96, 48, 24, or 16 h, respectively).
Hill models were used to characterize the relationships between changes in fungal
density (i.e., colony-­forming units [CFU]) in homogenized kidney tissue at 96 h
relative to baseline and the three aforementioned PK-PD indices. The data pre-
sented demonstrated that changes in fungal density were most closely associated
with anidulafungin AUC:MIC and Cmax:MIC ratios, indicating that this agent
exhibits a concentration-­dependent pattern of fungicidal activity. However, unlike
AUC values, the Cmax is achieved at a transient time point, making it difficult to
accurately capture in studies and apply to support dose selection. Therefore, in
such situations, AUC:MIC ratio serves as a more reliable and predictable PK-PD
index than Cmax:MIC ratio.

22.2.2 Identifying PK-PD Targets for Efficacy

Once the PK-PD index most associated with efficacy is known, the next step is to
determine the magnitudes of this index which are associated with various levels of
pathogen killing. These thresholds are commonly known as PK-PD targets for effi-
cacy, and they provide crucial information to assist in estimating the likelihood of
achieving efficacious drug concentrations in patients following the administration
of a given antimicrobial dosing regimen. Dose-ranging studies are used to derive
these PK-PD targets, wherein changes in microbial density are evaluated across a
wide range of antimicrobial doses. Given that the PK-PD index most associated
with efficacy is known by this point in time, there is no longer a need to account for
potential collinearities. Consequently, all doses are administered over identical dos-
ing intervals (e.g., every 24 h). The interval evaluated in these studies will be that
which best describes the relationship between change in bacterial burden and the
PK-PD index most associated with efficacy, as established by the results obtained
from prior dose-fractionation studies. Common thresholds assessed in these studies
include net stasis (i.e., no change in the density of bacteria or fungi from that
observed at baseline) and 1- and 2-log10 reductions in the counts of CFUs relative to
baseline observations.
Figure 22.4, which shows data from VanScoy et al. [15], illustrates the type
of data that can be derived from an in vitro dose-ranging study. In these studies,
a one-­compartment in vitro infection model was used to simulate total-drug epi-
thelial lining fluid (ELF) AUC values ranging from 33.3 to 7942 mg•h/L follow-
ing administration of arbekacin, an investigational aminoglycoside. These drug
exposures were evaluated against four Pseudomonas aeruginosa isolates, the
MIC values for which ranged from 2 to 8 mg/L. The relationship between
change in log10 CFU from baseline at 24 h and total-drug ELF AUC:MIC ratio,
22 The Role of Pharmacometrics in the Development of Antimicrobial Agents 675

Fig. 22.4 Arbekacin total-drug ELF AUC:MIC0–24 ratio targets associated with net bacterial stasis
and 1- and 2-log10 CFU reductions from baseline for P. aeruginosa based on data from a one-­
compartment in vitro infection model [15]

the PK-PD index associated with arbekacin efficacy, was evaluated using a Hill
model. Using this model, the magnitudes of total-drug ELF AUC:MIC ratio
associated with net b­ acterial stasis and 1- and 2-log10 CFU reductions from
baseline, which were 56.9, 142, and 393, respectively, were identified as shown
in Fig. 22.4.
These data exhibit several characteristics which indicate the robustness of the
above-described AUC:MIC ratio targets. We can be assured that the PK-PD rela-
tionship was well captured as evidenced by the nearly complete sigmoidal curve
obtained by the fitted Hill model. This is a product of designing the dose-ranging
study to obtain a wide range of AUC:MIC ratios by evaluating a large range of doses
and various isolates with differing MIC values. Moreover, the coefficient of deter-
mination (r2) of 0.856 for this relationship was high, which tells us that the relation-
ship between change in log10 CFU and the AUC:MIC ratio is strong. Finally, the
data pertaining to each of the various isolates evaluated are well dispersed along the
fitted relationship with no apparent trends, indicating that no substantial differences
in efficacy were observed across these isolates.
676 J. C. Bader et al.

22.2.3 Accounting for PK-PD Variability

Devoting time and resources to the design of studies that account for variability in
PK-PD relationships for efficacy is crucial to the development of a robust preclini-
cal PK-PD package. Bacteria and fungi are extremely complex and adaptive organ-
isms that can develop a myriad of antimicrobial resistance mechanisms and undergo
changes in their inherent fitness. Consequently, the variability among isolates for a
given pathogen needs to be considered when designing preclinical studies. The con-
sideration of such variability provides an opportunity to better characterize the
PK-PD of a given antimicrobial agent. The following will detail best practices for
designing PK-PD studies in order to maximize the information that can be gained in
light of the above-­described variability.
To begin, let us review the design of dose-fractionation studies and consider how
best to select an isolate for evaluation. Given that the primary objective when con-
ducting these studies is to discriminate among the various PK-PD indices and deter-
mine which is most associated with efficacy for an antimicrobial, the intention
should be to minimize the potential of generating variable and inexplicable results.
Therefore, when evaluating an antimicrobial agent, it is best to select a well-defined
isolate that is known to grow well in the in vitro or in vivo system intended for study
and for which consistent and predictable PK-PD data have been generated previ-
ously (e.g., as observed in prior time-kill studies).
Regarding dose-ranging studies, the objective when selecting isolates should be
to study a diverse collection of isolates such that inter-isolate variability can be
adequately characterized. The challenge panel of isolates should have MIC values
that encompass a clinically relevant range and that express applicable resistant
determinants. Given that these studies are employed to derive PK-PD targets associ-
ated with efficacy which are then used to forecast dosing regimens for patients in the
UDR setting or even the setting of MDR and XRD, it is important to account for the
population of pathogens expected in either of these clinical settings. Examples of
isolate collections used for PK-PD analyses that meet the above-described criteria
are described below.
Figure 22.5 shows the relationship between change in log10 CFU from baseline
and free-drug plasma AUC0–24:MIC ratio for an investigational anti-staphylococcal
agent, afabicin, against seven Staphylococcus aureus isolates with MIC values rang-
ing from 0.004 to 0.06 mg/L, the data for which was obtained from studies utilizing
a murine-thigh infection model [16]. When assessed relative to the range of MIC
values (<0.001–0.25 mg/L, MIC90 = 0.008 mg/L) evaluated in a recent in vitro sur-
veillance study of 660 S. aureus isolates collected from European, North American,
Latin American, and Asian hospitals from 2013 to 2014 [17], the range of MIC
values for the isolate collection studied was considered robust. In addition, the latter
consideration is presented in Fig. 22.6 which shows data obtained from one-­
compartment in vitro infection model studies of eravacycline, an investigational
tetracycline, against five Escherichia coli isolates stratified by those that were tetra-
cycline susceptible and non-susceptible based on either the Clinical and Laboratory
22 The Role of Pharmacometrics in the Development of Antimicrobial Agents 677

Fig. 22.5 Relationship between the change in log10 CFU from baseline at 24 h and free-drug
plasma AUC0–24:MIC ratio for afabicin based on data for seven S. aureus isolates studied in a neu-
tropenic murine-thigh infection model [16]

Fig. 22.6 Relationships

between change in log10
CFU from baseline at 24 h
and free-drug AUC0–24:MIC
ratio for eravacycline based
on data for tetracycline
susceptible and non-­
susceptible E. coli isolates
studied in a one-­
compartment in vitro
infection model [18]
678 J. C. Bader et al.

Standards Institute or US FDA susceptibility breakpoints [18–20]. In both ­examples,

the pooled data co-modeled well, producing robust sigmoidal relationships.
While there is no specific number of isolates from a family or genus of pathogens
which must be utilized to effectively characterize variability in the relationship
between the PK-PD index and reduction in bacterial burden, a sufficient number of
isolates needs to be initially studied. Assuming the isolates selected account for all
clinically relevant resistance determinants and embody a suitable range of MIC
values, investigators should evaluate the mean and median values for the PK-PD
targets to ensure they are relatively similar [21]. However, the ultimate decision
regarding how many isolates should be evaluated must be made on a case-by-­case
basis and may best be determined iteratively after examination of results based on a
reasonably robust number of isolates.

22.2.4 The Importance of Evaluating Pharmacokinetics

The importance of collecting samples to evaluate the PK within in vitro systems is

often underestimated. Many investigators may choose to forgo collecting PK data
when conducting studies, opting instead to simply assume the targeted PK profile
was obtained within the system. However, this step is critical to ensuring accurate
and reliable information is obtained when modeling PK-PD data. Despite the many
variables which can be held constant (e.g., antimicrobial and media flow rates
throughout the system), observed concentration-time profiles within in vitro sys-
tems can vary from those expected for a multitude of reasons such as drug degrada-
tion (e.g., hydrolysis or photodegradation), binding of drug to the infection model
components, or even simple methodologic errors. Failing to account for these fac-
tors by neglecting to collect PK samples can result in errors of varying magnitude
of impact when interpreting the study data. For instance, consider an in vitro dose-
ranging study in which no PK samples are collected but rather for which it is
assumed that the targeted concentration-­time profile will be obtained. However, the
concentrations actually achieved within the system are lower than those expected.
Consequently, the exposures evaluated for PK-PD target determination will be
falsely elevated, causing a rightward shift in the PK-PD relationship. Therefore, any
PK-PD targets derived from these data will also be falsely elevated. Figure 22.7
provides an illustration of this concept, while Fig. 22.8 shows data from a study
conducted by Louie et al. using a hollow-fiber in vitro infection model [22] for
which lower concentration-time profiles for ceftaroline than expected were achieved.
In the above-described study [22], ceftaroline was administered with and without
avibactam (also referred to as NXL104), a non-β-lactam-β-lactamase inhibitor, over
a 10-day study period. The activity of this combination was evaluated against three
Klebsiella pneumoniae isolates, one of which expressed several β-lactamase
enzymes, including KPC-2, SHV-27, and TEM-1. As shown in Fig. 22.8, all study
arms which included avibactam, achieved ceftaroline concentration-time profiles
similar to those targeted. However, in the study arm which did not contain avibac-
22 The Role of Pharmacometrics in the Development of Antimicrobial Agents 679

Fig. 22.7 Discrepancy

between AUC0–24:MIC
ratio targets associated a
1-log10 CFU reduction
from baseline based on
evaluations performed
using actual and expected
PK

tam, ceftaroline concentrations were much lower than those expected. This is due to
ceftaroline being hydrolyzed by the endogenous β-lactamase enzymes secreted by
the above-described K. pneumoniae isolate. These data underscore the importance
of collecting PK as part of all in vitro and in vivo infection model studies. Had these
investigators not collected PK samples during their study, they would have greatly
overestimated ceftaroline exposures achieved in the in vitro infection model.
The evaluation of in vivo PK is critical to conducting PK-PD analyses based on
data derived from in vivo infection models. Thus, careful consideration needs to be
given to the design of such studies to ensure that useful data are generated. To this
end, measures can be taken to mitigate sources of variability in PK of the antimicro-
bial agent by controlling for factors such as route of administration, species, sex,
and weight. The use of genetically modified animals of the same species, sex, and
weight has allowed for control of the latter set of factors. Along these same lines, the
route of drug administration is another important consideration. While intraperito-
neal injections are routinely utilized due to the relatively large potential space for
injection and the ease and rapidity with which drug administration can be carried
out, training of research staff would be required to ensure proper technique in order
to prevent drug administration into abdominal organs or adipose tissue. Intravenous
(IV) injections are also desirable for the rapid delivery of drug directly into the
bloodstream. However IV injections can be extremely difficult to administer to
smaller species such as BalbC as compared to CD-1 mice. Subcutaneous injections
are often the most preferred and relatively straightforward injection site, resulting in
minimal variability among injections and technicians. Other factors to consider that
can affect PK include the potential for drug interactions between the antimicrobial
agent and analgesic or anesthetic agent.
680 J. C. Bader et al.

Fig. 22.8 Targeted (solid line) concentration-time profiles with observed concentrations (sym-
bols) overlaid for ceftaroline (triangles) and avibactam (NXL104, squares) among in vitro treat-
ment arms including both agents (C to G) or ceftaroline or avibactam alone (B and H, respectively).
(Reproduced from Ref. [22] with permission from Antimicrob Agents Chemother. Copyright ©
American Society for Microbiology)
22 The Role of Pharmacometrics in the Development of Antimicrobial Agents 681

Fig. 22.9 Non-compartmental characterization of PK samples collected both intensively and

sparsely following multiple dose drug administration

Other study design elements to consider include infection status and PK sam-
pling. Although in vivo PK studies are often carried out in noninfected animals, it
can be advantageous to evaluate PK in infected animals. Moreover, when appli-
cable, PK samples from the effect site of interest should be collected in addition
to blood samples (e.g., ELF in the case of murine-lung infection models). Intensive
PK sampling enables non-compartmental analyses (commonly referred to as
“SHAM analyses”) of the PK data. Non-compartmental analyses allow for lines
across the individual or pooled concentrations obtained from animals to be con-
nected, thus allowing for the shape of the concentration-time profile(s) to be char-
acterized. Based on these data, exposure measures such as the AUC and Cmax can
be calculated. While PK sampling is typically intensive, it is also possible to
design less intensive PK sampling strategies and use compartmental analyses to
derive drug exposures. Compartmental analyses offer the benefit of allowing for
more precise estimates of drug exposure with less reliance on intensively sampled
PK from study animals. If non-compartmental analyses are used to evaluate PK
data, intensive sampling around multiple doses may be required to ensure reliable
estimates of drug exposure. Consider the case presented in Fig. 22.9 in which
animals were dosed every 12 h over a 48-h period, but PK were only intensively
sampled following the first and third doses. If a non-compartmental analysis were
used to evaluate these data, exposures would be greatly underestimated given that
the concentrations following the second and fourth doses would be largely
ignored. Thus, it is important to ensure that the PK sampling strategy is adequate,
682 J. C. Bader et al.

especially if ­non-­compartmental analyses are planned. Given that PK sampling is

often terminal for study animals, designing a study with less intensively sampled
PK and using compartmental analyses is also beneficial from an ethical perspec-
tive. Compartmental analyses are also more desirable when greater variability in
the PK of the antimicrobial agent is anticipated as would be the case when PK is
studied in infected animals. Compartmental models and their use for evaluating
clinical PK data in conjunction with Monte Carlo simulation to evaluate potential
antimicrobial dosing regimens for clinical studies are discussed in Sects. 22.3 and
22.4, respectively.

22.2.5 Resistance Prevention

In order to preserve the efficacy of an antimicrobial agent, we must determine not

only the doses required for microbial killing but also those needed to slow or pre-
vent the emergence of on-therapy resistance. Such dosing regimens hold the prom-
ise of durability. That is, antimicrobial dosing regimens which are selected on the
basis of resistance prevention are likely to maintain their antimicrobial activity well
after commercialization. Key to this effort is characterizing the time course and
relationship between drug exposures and resistance emergence. The gold standard
for obtaining such information is the hollow-fiber in vitro infection model, but many
other less sophisticated in vitro tools such as mutation frequencies and one-­
compartment (chemostat) models may be utilized to obtain a basic understanding of
this relationship.
In the case of one-compartment in vitro infection models, drug-free media
and antimicrobial doses are pumped into a central compartment which contains
a microorganism of interest. Waste and drug degradation by-products are pumped
out of the system and enter a peripheral waste reservoir. This system allows
investigators to precisely simulate targeted concentration-time profiles and is
more cost-effective than a hollow-fiber model but is limited in its ability to deter-
mine exposures that prevent amplification of resistant subpopulations. This limi-
tation is due to the inadvertent loss of microorganisms from the central
compartment when waste is removed from the system, which ultimately leads to
artificial decreases in susceptible and resistant organism density. Conversely, the
hollow-fiber in vitro infection model is a two-compartment model which utilizes
a cartridge comprised of thousands of small tubular filters (fibers). These filters
allow the free diffusion of media, drug, and waste products while trapping the
microorganisms within the cartridge. Consequently, the system enables the eval-
uation of resistance emergence over long periods of time (weeks to months)
within the hollow-fiber cartridge.
Figure 22.10 presents data obtained from a hollow-fiber in vitro infection model
evaluating the time course of resistance emergence across a wide range of doses for
the cephalosporin-β-lactamase inhibitor combination, ceftolozane-tazobactam [23].
This system was used to simulate free-drug serum c­eftolozane-tazobactam
 22 The Role of Pharmacometrics in the Development of Antimicrobial Agents

Fig. 22.10 Changes in the total and resistant bacterial populations over time for single ceftolozane, piperacillin-tazobactam, and control dosing regimens (a)
683

and a range of ceftolozane-tazobactam dosing regimens (b). (Reproduced from Ref. [23] with permission from Antimicrob Agents Chemother. Copyright ©
American Society for Microbiology)
684 J. C. Bader et al.

c­ oncentration-time profiles and evaluate the activity of these agents administered

together against a laboratory-derived Escherichia coli strain in order to determine
the combination of doses required to prevent the amplification of a resistant bacte-
rial subpopulation. In order to evaluate this endpoint, changes in bacterial density
for the total and resistant bacterial populations were determined. Samples for bacte-
rial enumeration were collected at 0 and 5 h after the start of the experiment and on
Days 1, 2, 3, 4, 6, 8, and 10 and plated on drug-free agar and agar supplemented
with piperacillin-tazobactam or ceftolozane-tazobactam at concentrations that were
three times the baseline MIC of the E. coli strain utilized. The total bacterial popula-
tion and resistant subpopulation were represented by the CFUs which grew on drug-
free agar plates and plates containing agar supplemented with drug concentrations
that were three times the baseline MIC, respectively.
As shown in Fig. 22.10, the range of ceftolozane-tazobactam doses utilized
yielded a wide range of effects. For the lowest ceftolozane-tazobactam regimen
(125/62.5 mg), the growth of the resistant bacterial subpopulation was stabilized
with no effect on the magnitude of the total bacterial population. The intermediate
dosing regimens (250/125 and 500/250 mg) resulted in the amplification of resis-
tance. Within these treatment arms, the resistant bacterial subpopulation was shown
to steadily increase, representing a greater proportion of the total bacterial popula-
tion over time. Finally, for the highest dosing regimens (750/375, 1000/500, and
1500/750 mg), resistance emergence was prevented as evidenced by the eradication
of the resistant subpopulation and eventual sterilization of the in vitro system. Data
such as these are highly informative and can aid in the selection of optimal dosing
regimens which minimize the risk of on-therapy amplification of the growth of
drug-resistant subpopulations.
The strength of the findings for the above-described example is that the E. coli
strain evaluated was genetically constructed in order to isolate a specific mechanism
of resistance for study, CTX-M-15 in this case. However, strains that are produced
in this manner are often less biologically fit and less virulent than those encountered
clinically. Thus, using clinical isolates that have been genotyped is a preferable
approach when designing PK-PD studies to evaluate pathogens with selected resis-
tance mechanisms.
Another important consideration for designing studies to evaluate the emer-
gence of resistance is the target pathogens of interest and the likely resistance
determinants that are expected to be observed. In this case, the evaluation of an
Enterobacter isolate which can overexpress AmpC β-lactamases as a result of
mutations in the AmpC transcription protein (AmpR) in response to β-lactam
exposure [24, 25] would have provided the opportunity to pressure test the dos-
ing regimen for ceftolozane-­tazobactam (1000 and 500 mg) that was chosen for
clinical study and ultimately approved for use for the treatment of patients with
complicated intra-­abdominal infections (cIAI) and complicated urinary tract
infections (cUTI) [26]. The conduct of such studies in early-stage development
provides the opportunity to evaluate decisions about both the choice of the part-
ner β-lactam and β-lactamase inhibitor and the dose of each agent in
combination.
22 The Role of Pharmacometrics in the Development of Antimicrobial Agents 685

22.3 Population Pharmacokinetic Analyses

Selection of a dosing regimen for Phase 2 and 3 trials is a critical decision in the
development of new antimicrobials. The dose selected needs to be low enough to
prevent severe drug-related toxicities, yet high enough to achieve efficacy in the
majority of patients. Ideally, this dose should also be of sufficient magnitude to
prevent the emergence of on-therapy resistance. The first step in selecting a clinical
dosing regimen is to understand the PK of the drug in humans, including PK vari-
ability and factors contributing to this variability. Once this is known and thor-
oughly understood, simulations of various proposed dosing regimens can be
conducted as will be discussed in Sect. 22.4.
Population PK modeling is the current gold standard for performing PK analyses
in this context. Although simpler tools are available (such as non-compartmental,
naïve pooling, and standard two-stage approaches), a population PK model offers a
few key advantages. The first of these is that the population approach allows for the
quantification of between-subject variability. This is particularly useful when simu-
lating various dosing regimens as one can estimate the width and shape of the
expected exposure distribution. The second advantage is the ability to quantify and
explain variability in PK through factors such as body size, age, gender, or clearing
organ function (i.e., patient-specific covariates). Again, this is useful when perform-
ing simulations as one can determine if the same dose can be administered to all
patients or if dose should be adjusted based on patient characteristics (e.g., weight-­
based dosing or adjustments based on patient renal function). The final key advan-
tage of population PK models is that they can be informed by sparsely collected PK
data based on an optimized sampling scheme. This is especially important when
utilizing data from Phase 2 and 3 trials as it is rarely feasible to collect rich PK data
in all patients in these studies due to ethical and logistical constraints.
In essence, population PK models use differential equations to describe the time
course of drug concentrations across a population of subjects. Traditionally, com-
partmental models are utilized, wherein the model compartments represent hypo-
thetical spaces to which drug may distribute (e.g., vasculature, tissue, macrophages,
etc.). The central compartment most often represents the vascular space and all
areas of the body where drug rapidly equilibrates. Various peripheral compart-
ments may be added to the base model to represent areas of the body with slower
distribution characteristics. In addition, absorption compartments can be added to
characterize the time course of drug disposition following various routes of admin-
istration such as absorption through the gastrointestinal tract after oral administra-
tion. Each compartment is represented with a differential equation using various
PK parameters (e.g., clearance or volume of distribution) to describe the move-
ment of drug into and out of the compartment. The values of these parameters can
vary across subjects. For instance, take an oral antibiotic with a concentration-time
profile which can be described with a one-compartment model. The model can be
described with two differential equations – one for the gut/gastrointestinal tract
and one for the plasma and other places in the body where the drug rapidly distrib-
686 J. C. Bader et al.

Fig. 22.11 Distributions of individual PK parameter estimates using a population PK model

utes. The ­differential equations often use a “ka” parameter to describe absorption
from the gut/gastrointestinal tract, a volume of distribution term to relate drug
amounts to measured drug concentrations, and a clearance term to describe removal
of the drug from the body. Using a population PK model, one can determine a
mean parameter value across all subjects (known as the “population mean value,”
as shown in Fig. 22.11). In addition, the model can also estimate an individual
parameter estimate for each subject, as shown in Fig. 22.11. The distribution
around the population mean value is termed “interindividual variability.”
Quantification of this variability is crucial for dose selection analyses, which are
discussed further in Sect. 22.4.
The development of the population PK model should be an iterative process,
beginning during Phase 1 development. If the studies are designed appropriately,
analysis of Phase 1 data should allow for the identification of anomalies in PK, such
as nonlinearity or nonstationary, early in drug development. Nonlinearity refers to a
change in exposure which is not proportional to the change in dose. For example, a
doubling of dose would typically be expected to cause a doubling of the AUC, but
in the case of a drug which exhibits nonlinear PK, the AUC might only increase
slightly following this doubling of dose. Oftentimes, this is due to saturation of an
absorption process in the gastrointestinal tract following oral administration. More
troubling are instances in which the dose doubles, but the resulting AUC is more
than doubled. This is often due to the saturation of an elimination pathway.
Nonstationary, on the other hand, refers to PK parameters that change with time
irrespective of dose. For example, a drug with auto-induction of clearance can result
in lower exposures on Day 5 of therapy relative to Day 1, even if the same dose is
administered in both cases. Other common PK issues which can be observed in
Phase 1 studies are food and diurnal effects. All of the abovementioned PK com-
plexities can be built into the population PK model structure.
Throughout the iterative process of developing and refining a structural popula-
tion model, covariate analyses can be conducted. Covariate analyses allow for iden-
tification and evaluation of factors that explain variability in the PK parameters.
Typical covariates evaluated include body size measures [e.g., weight, height, body
surface area, body mass index (BMI), lean body weight], sex, race, age, clearing
22 The Role of Pharmacometrics in the Development of Antimicrobial Agents 687

organ function (e.g., creatinine clearance or hepatic function tests), and genotypes.
Trends between covariates and individual PK parameters are explored. Any
covariate-­parameter pairs with a considerable relationship are tested in the model,
and those which are found to have statistically significant relationships remain in
the model. Oftentimes, the value of these analyses may be limited during early-­
stage development when data stem largely from Phase 1 studies comprised of
largely homogenous healthy volunteers. However, covariate analyses become more
informative as they are enriched with data obtained from special populations such
as patients with renal impairment or obesity and Phase 2 and 3 studies, which enroll
large numbers of patients with greater covariate variability.
As evidenced by results of analyses that have demonstrated differences in clear-
ance, volume of distribution, and tissue penetration in patients compared to healthy
volunteers [27–29], differences in PK for patients relative to healthy subjects may be
observed. This is not surprising given the physiologic changes observed in patients,
particularly those who are critically ill. In infected and acutely ill patients, greater PK
variability has been observed [30–32]. Regardless of the scenario, these differences
can be quantified by refining the population PK model. Thus, after development of
the Phase 1 population PK model, it is important that the model is refined using data
collected from Phase 2 and/or Phase 3 studies. Phase 2 is typically the first time the
investigational agent is administered to patients rather than healthy volunteers.
However, in the case of the development of antimicrobial agents for MDR and XDR
pathogens, the conduct of Phase 1b studies conducted in patients prior to small Phase
3 clinical studies is increasingly becoming commonplace.
Lastly, it is important to update the results of the covariate analysis after refining
the population PK model as there is typically a wider range of covariates observed
in Phase 2 and 3 studies. For example, the protocol for Phase 1 studies usually
excludes patients with a BMI above 30 kg/m2. However, in Phase 2 and 3, often-
times there is no exclusion criteria based on BMI. Important relationships between
BMI and PK parameters such as clearance may not be evident until a wider range of
BMI data are included in the analysis datasets. In addition to the wider range of
covariates for patients in Phase 2 and 3 studies relative to subjects enrolled in Phase
1 studies, additional covariates which could not be assessed in healthy volunteers
(e.g., renal function, APACHE II score, and/or infection type) can be evaluated.

22.4 Monte Carlo Simulation and Dose Selection

Monte Carlo simulation is a mathematical technique that uses repeated random

sampling to determine the impact of uncertainty when characterizing the probability
of an event. Such an approach is useful to determine the probability of achieving a
PK-PD target associated with efficacy among simulated patients with drug expo-
sures that would be expected in the target patient population [33, 34]. These analy-
ses, which are commonly referred to as PK-PD target attainment analyses, are
widely used in both early- and late-stage drug development to support the selection
688 J. C. Bader et al.

Fig. 22.12 Representative

example of concentration-­
time profiles for a
simulated patient
population after
administration of a given
dosing regimen. The solid
line represents the median
profile and the lower and
upper dotted lines
represent profiles with
concentrations at the 5th
and 95th percentiles,
respectively

of dosing regimens and develop in vitro susceptibility testing criteria (i.e., suscepti-
bility breakpoints) [35]. In this section, we will briefly outline the use of Monte
Carlo simulation to carry out PK-PD target attainment analyses and discuss how the
results of these analyses may be used to evaluate dosing regimens and inform the
selection of interpretative criteria for in vitro susceptibility testing.
The optimal application of Monte Carlo simulation for dose selection evalua-
tions requires the use of compartmental models as previously discussed in Sect.
22.3. A population PK model for the antimicrobial under investigation can be used
to generate drug exposures of interest for simulated patients. As discussed previ-
ously, interindividual variability can exist on many of the parameters within popula-
tion PK models. As illustrated in Fig. 22.11, by randomly assigning parameter
estimates for simulated patients based on the distributions for these parameters,
interindividual variability can be considered. As shown in Fig. 22.12, PK parameter
estimates assigned to simulated patients can be used to generate concentration-time
profiles after administration of the dosing regimen of interest, which in this example
was an intravenously administered antibiotic infused over 2 h twice daily. The dis-
tribution of these concentration-time profiles, as represented by the median and 5th
and 95th percentiles, allows for an understanding of the variability expected in the
actual patient population.
These above-described concentration-time profiles can be used in conjunction
with fixed MIC values to calculate the PK-PD index associated with efficacy for a
given pathogen by MIC for individual dosing regimens administered to each sim-
ulated patient. The percentage of simulated patients achieving a given PK-PD
target by MIC is then determined. As shown by data assessing the percent proba-
bilities of PK-PD target attainment by MIC for ceftaroline relative to nonclinical
%T > MIC targets for S. aureus shown in Fig. 22.13 [35], these data are com-
monly interpreted in the context of observed MIC values for isolates based on
22 The Role of Pharmacometrics in the Development of Antimicrobial Agents 689

Fig. 22.13 Percentage of simulated patients with normal renal function (80≤ creatinine clearance
≤170 mL/min/1.73 m2) achieving free-drug (f) %T > MIC targets by MIC following administra-
tion of ceftaroline fosamil 600 mg q12h, overlaid on a histogram showing the MIC distributions
for MRSA and MSSA. (Reproduced from Ref. [35] with permission from Antimicrob Agents
Chemother. Copyright © American Society for Microbiology)

in vitro surveillance data. In this example, a collection of 3965 S. aureus isolates

collected from medical centers in the United States, stratified by the 2254 and
1711 isolates which were methicillin-resistant and methicillin-susceptible, respec-
tively (MRSA and MSSA, respectively), were evaluated. Percent probabilities of
PK-PD target attainment of ≥90% up to MIC values that represent the upper mar-
gins of the MIC distribution (i.e., the MIC90 which represents the MIC value at
which ≥90% of isolates are inhibited) would be considered a favorable set of
results for a given dosing regimen. While the evaluation of the probability of
PK-PD target attainment by MIC is useful to support recommendations for dosing
regimens, evaluations weighted over the MIC distribution also provide support for
a given dosing regimen. The latter, which is commonly referred to as the overall
probability of PK-PD target attainment, can be determined by multiplying the
probability of PK-PD target attainment for a specific PK-PD target at a given MIC
value with the probability of occurrence of that MIC value and then taking the
sum of these percentages. When based on robust in vitro surveillance data for a
given pathogen [36], the overall probability of PK-PD target attainment is a met-
ric that provides an expectation of PK-PD target attainment in a simulated popula-
tion based on the MIC distribution for that pathogen likely to be observed in
clinical practice.
The choice of the PK-PD target used to support dose selection and susceptibility
breakpoint recommendations is an important consideration for assessments of
690 J. C. Bader et al.

PK-PD target attainment. Endpoints for such PK-PD targets range from net bacte-
rial stasis to a 2-log10 CFU reduction from baseline. Typically, these data are derived
from neutropenic murine-thigh or murine-lung infection models or when warranted,
in vitro infection models. Net bacterial stasis has been suggested to be an appropri-
ate endpoint for a PK-PD target when selecting dosing regimens to treat patients
with infections associated with lower bacterial inoculums and/or for which source
control, including surgical intervention, is an option. This endpoint may also be
reasonable to assess for inferences about patient populations that are expected to be
immunocompetent and for whom the response rate associated with no treatment is
expected to be relatively high (e.g., ≥60%). Examples of indications that meet these
criteria include acute bacterial skin and skin structure infections (ABSSSI), cIAI,
and cUTI. Reduction of 1-log10 CFU from baseline has been suggested to be an
appropriate endpoint for a PK-PD target when selecting dosing regimens to treat
patients with infections associated with higher bacterial inoculums such as pneumo-
nia, endocarditis or bacteremia, and/or for infected patients who are
­immunocompromised. In such populations, the response rate associated with no
treatment may be low (e.g., ≤40%) [11, 37, 38].
Support for each of the above-described endpoints is based on successful
translations between the results of previous PK-PD analyses based on nonclini-
cal and clinical data [3, 39–42]. Results of analyses of these data have demon-
strated that the same magnitude of the PK-PD target associated with net bacterial
stasis from neutropenic murine-thigh infection models for a given antimicrobial
agent was associated with a high percentage of successful outcomes among
patients with cIAI or ABSSSI [3, 39–42]. The choice of a 1-­log10 CFU reduction
from baseline for the treatment of patients with infections with a higher no-
treatment response is based on an assessment of PK-PD target attainment analy-
sis results for antimicrobial agents that were evaluated for pneumonia. As the
percent probability of achieving a PK-PD target associated with a 1-­log10 CFU
reduction from baseline increased, so too did the probability of a successful
regulatory outcome [2]. The latter was considered an indicator of meeting non-
inferiority in pivotal clinical trials.
While a 2-log10 CFU reduction from baseline has been suggested as an endpoint
for indications such as HABP/VABP [11], attainment of a PK-PD target associated
with such a level of bacterial reduction may not be possible for many antimicrobial
agents, including those currently available and commonly used for these indica-
tions. As previously shown for a meropenem dosing regimen of 2 g q8h infused over
1 h [4], while it is possible to achieve the %T > MIC target associated with a 2-­log10
CFU reduction from baseline, large interpatient variability can hinder the likelihood
of achieving this PK-PD target in many patients. However, one strategy to overcome
this is to administer the same dose as a prolonged infusion over 3 h [43]. Such a
strategy was employed for development of meropenem-vaborbactam, a β-lactam/β-­-
lactamase inhibitor combination recently approved by the US FDA [44]. From a
drug development perspective, the margin of safety should be weighed against goals
for efficacy when considering an endpoint of a 1- vs 2­-log10 CFU reduction from
baseline for indications such as HABP/VABP [38]. If an antimicrobial agent has a
22 The Role of Pharmacometrics in the Development of Antimicrobial Agents 691

wide safety margin, developers have a greater opportunity to utilize a 2-log10 CFU
reduction endpoint.
In addition to the PK-PD target, it is important to consider exposures at the effect
site when applicable. For example, if an antimicrobial agent is being developed to
treat patients with pneumonia, it is important to evaluate the likelihood of achieving
efficacious drug concentrations in ELF. To enable the consideration of ELF expo-
sures, PK data from healthy volunteers and if available, from patients, should be
considered when developing the population PK model. This model would then be
used together with ELF PK-PD targets associated with efficacy derived from a
murine-lung infection model and Monte Carlo simulation to assess PK-PD target
attainment for dosing regimens.
As described above, the assessment of dosing regimens to evaluate in Phase 2
or 3 studies requires the use of a population PK model constructed using PK data
from healthy volunteers enrolled in Phase 1 studies. As such, the interindividual
variability in PK will be limited and may not be reflective of the target patient
population. In such cases, inflating variance in PK parameters (e.g., increasing
the ­interindividual variability terms on PK parameters such as clearance and vol-
ume) as a part of a sensitivity analysis may be a useful approach to further dis-
criminate among candidate dosing regimens [38]. Another limitation of a
population PK model developed using Phase 1 PK data is that covariate distribu-
tions are relatively narrow. Since studies for special populations, including sub-
jects with renal or hepatic impairment, are typically not completed early in a
clinical development program, the evaluation of covariates is not usually avail-
able until late-stage development. Thus, early-stage development decisions for
dose selection should be confirmed after a population PK model has been refined
using data from the target patient population and special populations. Additionally,
an understanding of covariates that are highly influential on PK allows for the
assessment of dosing regimens in simulated patients stratified by ranges of such
covariates to support dosing recommendations for special populations. Data
from simulated patients can be used to support dosing recommendations even if
such dosing regimens were not assessed in clinical trials. This strategy was used
for delafloxacin, a fluoroquinolone that was recently approved by the US FDA
for the treatment of patients with ABSSSI. The delafloxacin dosing regimen
approved for patients with severe renal impairment [45], 450 mg by mouth twice
daily, was not studied in clinical trials [46, 47] but was supported by the results
of population PK and PK-PD target attainment analyses [48].
The above-described strategy to use preclinical PK-PD data, population PK
models, and Monte Carlo simulation both for early- and late-stage development
decisions about dose selection allows developers to mitigate risk and increase the
likelihood of regulatory success. The results of such analyses can also be used to
inform recommendations for interpretative criteria for in vitro susceptibility testing
criteria for the antimicrobial agent of interest against target pathogens. The data
obtained from these simulations can be used in conjunction with clinical outcome
data by MIC and pathogen susceptibility distributions to support susceptibility
breakpoint decisions. Results of PK-PD target attainment analyses to support sus-
692 J. C. Bader et al.

ceptibility breakpoint recommendations is critical information for drug developers

not only when seeking regulatory approval but also early in clinical development.
Given the lengthy process of incorporating an antimicrobial into automated suscep-
tibility testing systems, it behooves developers to perform preliminary susceptibility
breakpoint evaluations in order to ensure informed decisions are made.

22.5 Clinical Data for PK-PD Analyses

As described above, an understanding of the PK-PD characteristics of an antimicro-

bial agent early in drug development increases the likelihood of regulatory success.
However, the evaluation of PK-PD relationships for both efficacy and safety based
on clinical data collected in Phase 2 and 3 can be used to provide valuable informa-
tion to confirm early-stage dose selection decisions and further improve the likeli-
hood of regulatory success. Depending on the indication and whether the
antimicrobial agent is being developed for a setting of UDR versus MDR or XDR
pathogens, the robustness of the clinical data package required can vary. For indica-
tions involving relatively susceptible pathogens and for which a suitable comparator
agent can be studied, the clinical data package includes data from clinical studies
that are powered to demonstrate non-inferiority and that are large enough to detect
safety signals. Such studies, especially when PK data are collected in all patients,
provide a robust repository of data to use for evaluating PK-PD relationships for
efficacy and/or safety endpoints. However, in the setting of highly resistant patho-
gens, large clinical studies are difficult to conduct in a reasonable time frame. An
important challenge for conducting such studies is the lack of frequency of patients
with such infections. Furthermore, when identified, study enrollment can be diffi-
cult to accomplish as these patients are often critically ill [1]. In order to develop a
given antimicrobial agent for MDR or XDR pathogens in a reasonable time frame,
clinical data for indications involving such pathogens will be less robust. Given that
data from in vitro or in vivo infection models have demonstrated similar PK-PD
relationships for efficacy among isolates with and without resistant determinants
[18, 49], the most efficient development program for antimicrobial agents for MDR
and XDR pathogens would be one that combines robust preclinical PK-PD data, the
data package for which includes MDR and/or XDR pathogens, with data from clini-
cal studies conducted in the UDR setting that are powered to demonstrate non-­
inferiority. While such programs, especially with even a limited number of clinical
cases with MDR and/or XDR pathogens should be adequate to allow for labeling
that includes indications for such pathogens, regulatory agencies to date have been
less willing to formally establish drug development paths based on this premise.
Instead, discussion has centered around a plan to encourage sponsors to assemble
robust preclinical PK-PD and Phase 1 PK data packages together with a limited
clinical data package to strengthen NDA submissions for indications due to MDR
and XDR [1, 50]. Regardless of the path, there is a common requirement for both
nonclinical and clinical PK-PD data to increase regulatory certainty.
22 The Role of Pharmacometrics in the Development of Antimicrobial Agents 693

22.5.1 Data Prerequisites

Irrespective of whether the clinical data package is robust or limited, the data derived
from PK-PD analyses are valuable. However, as described below, objectives of such
analyses will vary depending on the data package available. Important prerequisites
for both types of clinical data packages include the collection of PK data from all
patients and the evaluation of informative endpoints. As described in Sect. 22.3, the
benefit of developing a population PK model based on Phase 1 data is that the model
can be used to determine sparse PK sampling strategies for implementation in clini-
cal trials. Such strategies are designed to ensure that optimal information to estimate
drug exposure in each patient is obtained using a minimal number of blood samples
for drug assay as possible. Using these sparse PK data, the goal of additional popu-
lation PK analyses is to refine the existing model developed using Phase 1 data in
order to enable precise and unbiased estimation of drug exposure in individual
patients, including the applicable PK-PD index for efficacy (e.g., AUC:MIC ratio,
Cmax:MIC ratio, or %T > MIC).
In addition to reliable estimates of drug exposure, well-defined and reproducible
efficacy and safety endpoints are needed to evaluate PK-PD relationships for such
endpoints. Objective criteria, determined by observations collected at informative
time points, are required to assess drug effect. Clinical trial endpoints for efficacy
are typically categorical variables, such as clinical response to therapy (success or
failure) assessed at the test-of-cure visit (i.e., a window of time after the end of study
drug; TOC) and/or at the end of therapy. However, recent US FDA guidance for a
number of indications has described the assessment of efficacy endpoints evaluated
earlier in therapy [5, 6]. For patients with ABSSSI and CABP, clinical response is
assessed on Days 2 to 3 and 3 to 5, respectively. PK-PD relationships for efficacy
have been largely described using dichotomous efficacy endpoints assessed at TOC
[3, 31, 41, 51–54]. In contrast, there is comparatively less experience evaluating
efficacy endpoints assessed earlier in therapy [55]. Despite the lack of experience
with the latter, given the natural course of infection, which involves eradication of
the pathogen followed by macrophage and inflammatory modulator activity, which
is then followed by resolution of signs and symptoms, it may be difficult to identify
PK-PD relationships for efficacy early after therapy has been initiated [56–58].
Consequently, the time at which efficacy is assessed can influence the likelihood of
identifying PK-PD relationships for efficacy.
While dichotomous efficacy endpoints are typically evaluated in clinical trials
for antimicrobial agents and serve as primary endpoints upon which sample size is
determined, the evaluation of continuous or time-to-event efficacy endpoints can
also be informative. Examples of continuous endpoints include change in bacterial
density or lesion size, while examples of time-to-event endpoints include time to
resolution of signs and symptoms, lesion size reduction, or bacteriologic eradica-
tion. Continuous or time-to-event endpoints have the benefit of being more sensitive
than categorical endpoints for capturing drug effect. When measures of efficacy are
assessed serially, this provides the opportunity to identify the time period during
694 J. C. Bader et al.

which treatment effect is greatest [55, 59]. Evaluation of such endpoints for PK-PD
analyses for efficacy has the potential to inform decisions about dose and duration
using a relative smaller sample size than that for a dichotomous efficacy endpoint
[59]. For example, while evaluations of clinical or microbiological response for 38
tigecycline-treated patients with CABP failed to reveal PK-PD relationships for
efficacy, a relationship between free-drug AUC:MIC ratio and time to fever resolu-
tion was identified [60]. The median time to fever resolution was 12 and 24 h for
patients with a free-drug AUC:MIC ratio >12.8 and ≤12.8, respectively. Thus,
despite not representing the primary clinical trial endpoint for efficacy, relationships
for such endpoints can be used to support dose selection and even provide potential
insights about the duration of therapy.
For the assessment of PK-PD relationships for safety endpoints, the same prin-
ciples as described above are applicable. While safety endpoints, such as the pres-
ence or absence of a given safety event or a dichotomous threshold for a continuous
laboratory measure, are dichotomous in nature, the assessment of continuous
­endpoints including laboratory measures or physiologic measurements such as
blood pressure collected serially provides the opportunity to develop informative
multivariable models [61, 62]. Such models, which can be constructed to describe
the effect of varying drug exposures on laboratory measures over the course of
therapy in the context of other patient factors, can then be applied to simulated data
to discriminate among potential dosing regimens to be studied in Phase 3 trials. For
example, the percentage of simulated patients with laboratory measures above clini-
cally relevant folds of the upper limit of normal (ULN) (e.g., 3, 5 or 10 × ULN) or
in the case of systolic blood pressure, the percentage of patients with readings
≥160 mmHg can be determined for individual dosing regimens. This information,
together with assessments of the percent probabilities of achieving each efficacy
endpoints (based on clinical PK-PD relationships for efficacy) and/or nonclinical
PK-PD targets, can be used to balance considerations for safety and efficacy. Or
using multivariable models developed using Phase 2 and/or 3 data, percent proba-
bilities of elevation of these safety endpoints can be evaluated among all simulated
patients and subgroups at increased risk who receive intended dosing regimens. The
identification of patient populations at increased risk and the characterization of the
elevations for such safety endpoints can be used to inform use for labeling and/or
clinical practice guidelines.

22.5.2 Analysis Objectives

As described above, the robustness of the clinical data package guides the objectives
of the PK-PD analyses for efficacy. For antimicrobial agents for patients with infec-
tions arising from pathogens in the setting of UDR, the sample size of evaluable
populations is expected to be sufficient to support the assessment of PK-PD rela-
tionships for efficacy. Thus, the objective is to determine if PK-PD relationships for
efficacy endpoints can be identified. However, despite the robustness of the sample
22 The Role of Pharmacometrics in the Development of Antimicrobial Agents 695

size of analysis populations, other factors may influence the ability to characterize
PK-PD relationships, including the duration of therapy. It is important to evaluate
patients from the microbiologically evaluable population in order to consider
patients who received a sufficient number of doses and who had pathogen(s) iso-
lated at baseline. The former ensures that the lack of clinical response is not attrib-
uted to insufficient duration of drug exposure, and the latter allows for drug
exposures to be indexed to pathogen MIC values in order to enable PK-PD indices
to be determined. For infections for which there are baseline pathogens with antici-
pated or known (as determined by preclinical data) PK-PD characteristics that dif-
fer, evaluation of subpopulations may be necessary to characterize PK-PD
relationships for individual pathogens. Or in the setting of infections with polymi-
crobial pathogens, careful consideration needs to be given to how the primary
pathogen used for calculating the PK-PD index is identified. Additionally, consider-
ation needs to be given to the definitions for clinical failure. If the reasons for declar-
ing a clinical failure include those not related to study drug (e.g., an adverse events),
data for patients failing for these reasons should be excluded given the potential for
these data to impede the ability to identify PK-PD relationships.
Finally, and perhaps most importantly, while it is important to consider all of the
above-described factors to ensure that every opportunity has been provided to allow
for PK-PD relationships for efficacy based on a robust clinical data package to be
identified, the lack of identification of PK-PD relationships for efficacy is a predict-
able outcome when patients have received PK-PD optimized dosing regimens. In such
cases, it is still valuable to demonstrate that drug exposures from patients indexed to
MIC values from pathogens identified at baseline exceed nonclinical PK-PD targets
for efficacy to confirm the basis for dose selection. When PK-PD relationships for
efficacy are identified based on clinical data from patients who received PK-PD-
optimized dosing regimens, these relationships are usually based on a dichotomized
variable for the PK-PD index of interest, the threshold for which is optimally deter-
mined using a number of statistical approaches. These approaches can include using
the threshold of the PK-PD index representing the first split of a classification or
regression tree, a receiver operating characteristic curve, or using a model fit to esti-
mate a threshold for achieving a target efficacy outcome or probability. PK-PD rela-
tionships identified in this manner resemble step functions and allow for patients with
both lower PK-PD indices and percentages of successful response to be contrasted
from those with higher PK-PD indices and percentages of successful response [51,
59]. Table 22.1 summarizes the results of two separate PK-PD analyses of Phase 3
data for patients who received PK-PD optimized dalbavancin or oritavancin regimens.
In both cases, PK-PD relationships identified were based on two-group variables for
AUC:MIC ratio [63, 64]. The differences in the percentage of successful clinical
responses between patients in the lower and higher exposure groups were 10.9 and
13.6%, respectively, with percentage of patients with successful clinical response in
the lower AUC:MIC ratio groups of 89.1 and 82.6% for dalbavancin and oritavancin,
respectively. Thus, when PK-PD-optimized dosing regimens are studied and PK-PD
relationships based on two-group variables are identified, the differences between the
lower and higher exposures groups are unlikely to be impressive.
696 J. C. Bader et al.

Table 22.1 Summary of PK-PD relationships for efficacy for dalbavancin and oritavancin based
on dichotomous two-group AUC:MIC ratio variables
Percentage of patients <
or ≥ threshold achieving
Threshold the efficacy endpoint
value of (n/N)
Antimicrobial Efficacy PK-PD ≥
agent endpoint PK-PD index index < threshold threshold P-value
Dalbavancin Clinical AUCavg:MIC 21,267 89.1 (98/110) 100 0.01
[63] success at the ratiob (52/52)
test-of-cure
visita
Oritavancin Clinical AUC0– 11,982 82.6 (19/23) 96.2 0.029
[64] success at the 72:MIC ratiod (126/131)
post-therapy
evaluationc
a
The test-of-cure visit occurred 14 days [± 2 days] after the end of therapy
b
AUCavg:MIC ratio was calculated by dividing the average 24-h AUC from 0 to 120 h by the base-
line MIC of the infecting pathogen
c
The post-therapy evaluation occurred 7–14 days after the end of therapy
d
AUC0–72:MIC ratio was calculated by dividing the AUC from 0 to 72 h by the baseline MIC of the
infecting pathogen

However, for limited clinical data packages in support of indications involving

MDR or XDR pathogens, the sample size of evaluable patients will likely be insuf-
ficient to allow for formal analyses to be conducted. Thus, in such cases, the objec-
tive of the PK-PD analyses for efficacy will be to confirm that drug exposures
indexed to MIC values from pathogens isolated at baseline exceeded nonclinical
PK-PD targets for efficacy based on robust preclinical PK-PD data for all patients
studied. Such information will thereby serve to support dosing regimens selected.

22.5.3  istorical Data and Bayesian Approaches for Clinical

H
Trial Design

Given the current paradigm for obtaining robust preclinical PK-PD data and using
these data with Phase 1 PK data and Monte Carlo simulation to predict doses for
Phase 2 and 3 clinical trials, the likelihood for failed clinical trials has been reduced.
Evaluation of data based on contemporary clinical trials that did not make full use
of these approaches to select dose, together with innovative statistical approaches,
provides the opportunity to answer questions about the no-treatment effect. Such
data represent valuable inputs for power and sample size calculations for future
clinical trials in the setting of UDR. As described below, data for tigecycline from
61 patients with HABP/VABP who were microbiologically evaluable and who had
sufficient PK data, the clinical trial that failed to demonstrate non-inferiority
22 The Role of Pharmacometrics in the Development of Antimicrobial Agents 697

compared to imipenem/cilastatin in the clinically evaluable population [65], yielded

a number of useful PK-PD findings [53].
Panel A of Fig. 22.14 shows the fitted function and associated 95% pointwise
confidence bounds for the relationship between clinical response and free-drug
AUC:MIC ratio which was identified using univariable logistic regression [53, 66].
This function is overlaid on a histogram for the distribution of free-drug AUC:MIC
ratio. Three important observations based on these data were the following: (1) As
the free-drug AUC:MIC ratio increased, so too did the probability of clinical suc-
cess; (2) the 95% pointwise confidence bounds around the logistic function were
tight in the free-drug AUC:MIC ratio range in which the data density was high; and
(3) a large proportion of patients (31%) had observed free-drug AUC:MIC ratios
associated with a low probability of clinical success, an indicator that the chosen
tigecycline dosing regimen, 100 mg IV followed by 50 mg IV every 12 h, was sub-
optimal for patients with HABP/VABP [53].
The above-described analyses were based on frequentist inference. In a follow-
­up analysis, Bayesian inference, which provides the benefit of considering prior
information, was applied to reassess the PK-PD relationships for efficacy [66].
Specific objectives of the analyses were to determine and compare the magnitude of
treatment effect and the ability of clinical trial endpoints to capture drug benefit
using frequentist and Bayesian statistical approaches. Prior information that
informed the Bayesian analyses were based on data from in vivo studies conducted
using a neutropenic murine-thigh infection model. These data, which demonstrated
that increasing AUC:MIC ratio was associated with improved response, served to

Fig. 22.14 Frequentist (A) and Bayesian (B) logistic regression-estimated relationships between
clinical response and the tigecycline free-drug AUC:MIC ratio based on data from 61 patients with
HABP/VABP. The solid lines represent the fitted functions based on frequentist and Bayesian
logistic regression, while the dashed lines represent the upper and lower 95% pointwise confidence
and credible bounds, respectively. The green histogram represents the distribution of observed
values for free-drug AUC:MIC ratio. (Reproduced from Ref. [66] with permission from Antimicrob
Agents Chemother. Copyright © American Society for Microbiology)