Chapter 75

Biofabricated three-dimensional tissue

models
David B. Berry1,*, Claire Yu1 and Shaochen Chen1,2,3,4
1
Department of NanoEngineering, University of California, San Diego, La Jolla, CA, United States, 2Department of Bioengineering, University of
California, San Diego, La Jolla, CA, United States, 3Materials Science and Engineering Program, University of California, San Diego, La Jolla, CA,
United States, 4Chemical Engineering Program, University of California, San Diego, La Jolla, CA, United States

Introduction which affects function, signaling, and gene expression

[3 6]. Early evidence of this phenomenon was reported
Over the past decade, three-dimensional (3D) biofabrica- in 1989, when Dunn et al. demonstrated that hepatocytes
tion has combined principles of 3D bioprinting and 3D cultured “sandwiched” between two layers of collagen gel
bioassembly to replicate complex tissue structure and maintained normal morphology and albumin secretion for
function [1]. Using new advancements in 3D printing at least 42 days, while hepatocytes cultured on a single
technology, biomedical engineers can control the spatial collagen layer ceased albumin secretion within a week
positioning of biomaterials (natural, synthetic), biochem- [7]. This early work and similar studies have motivated
icals (drugs, growth factors), and/or living cells [induced developing biomimetic 3D environments in order to pre-
pluripotent stem cells (iPSCs), mature-differentiated cells, serve normal cellular function.
embryonic stem cells, etc.]. As such, this approach 3D tissue models can be separated into two categories:
enables the creation of precise tissue-engineered (TE) (1) scaffold-based and (2) scaffold-free designs. Scaffolds
models for drug screening, studying human development, serve as a biomimetic extracellular environment to pro-
fabrication of functional implants to replace damaged tis- vide chemical stimulation and/or mechanical support, in
sue, and delivery of biomolecules with temporal and spa- order to promote natural cellular interactions and organi-
tial cues to guide autologous tissue regeneration in vivo. zation. Scaffolds often consist of specially designed mate-
These transformative tools have allowed researchers to rials and are designed to degrade over time, coinciding
develop 3D models that accurately depict the structure with de novo extracellular matrix (ECM) production.
and function of normal and diseased tissues, resulting in These scaffolds can consist of one of more materials, in
more physiologically relevant behavior from the cell to order to simultaneously provide a mechanically
the whole tissue level than traditional two-dimensional stable structure for handling and an environment which
(2D) model systems. promotes cell adherence and migration. Cells can either
2D cell culturing has been used for over a century to be directly incorporated into the scaffold, or scaffolds can
study cellular responses to biophysical and biochemical rely on the autologous cellular environment to migrate
stimulation [2]. However, many of these cell responses and differentiate within when implanted in vivo.
differ in 2D culture than in vivo, making them poor-to- Challenges with scaffold-based models include inhomoge-
moderate platforms for various tissue models. To over- neous distributions of cells within the scaffold and lack of
come this, 3D model platforms have been utilized that gross complex tissue self-assembly. Scaffold-free models
more accurately mimic the natural complex biophysical rely on self-assembly of cells into larger constructs with
and biochemical stimuli cells experience in vivo. Studies autologous ECM deposition for support. The building
have routinely demonstrated that cells in 3D culture differ blocks of these constructs are often cell sheets, spheroids,
morphologically and physically from cells cultured in 2D, and tissue strands and do not require complex fabrication

*. These authors contributed equally to this publication.

Principles of Tissue Engineering. DOI: https://doi.org/10.1016/B978-0-12-818422-6.00077-0
Copyright © 2020 Elsevier Inc. All rights reserved. 1417
1418 PART | TWENTY ONE Emerging technologies

techniques, which are often utilized for scaffold-based interest. The effective use of these tools has resulted in
models. For example, tissue strands can be formed by advanced 3D models of many tissue systems. In the fol-
injecting a cell pellet into a microtubular capsule, allow- lowing sections, we will discuss key factors involved in
ing those cells to spontaneously adhere to one another model design, and specific examples of how these techni-
and remodel via cadherin-mediated cell cell interactions ques are applied to real-world situations.
[8,9]. These strands can then be 3D printed to form TE
constructs such as tissue patches. Scaffold-free models
Current methods of three-dimensional
are limited by size (they often form a necrotic core due to
inadequate nutrient diffusion) and have inadequate biofabrication
mechanical properties, leading to cell damage when han- The most commonly used methods for 3D bioprinting
dled [10]. Recently, synergistic methods combining include inkjet-based, extrusion-based, and light-assisted
scaffold-based and scaffold-free tissue models have been printing. All of these techniques have been used to either
proposed, which may allow for more complex, mechani- 3D print scaffolds onto which cells can be seeded or to
cally stable TE constructs. Silva et al. reported the use of directly encapsulate cells to form complex tissue models.
lockyballs—mechanically interlockable microscaffolds— The preprinted material is termed “bioink” and can con-
in order to provide a sound mechanical structure, within sist of cells, biochemicals, and/or biomaterials, depending
which spontaneous tissue organization was allowed to on the intended structure and printing method. Each of
occur [11]. Prefabricated human adipose stem cell spher- these methods has strengths and weaknesses that must be
oids were seeded into lockyballs and allowed to self- considered. Below, we will provide a detailed description,
assemble and differentiate toward adipogenic, osteogenic, evaluation, and comparison of each of these techniques.
and chondrogenic lineages, effectively combining In addition, these techniques are directly compared in
scaffold-based and scaffold-free approaches. Table 75.1.
Native tissues are made up of a heterogeneous cell
population consisting of fibroblasts, endothelial cells, and Inkjet printing
stem cell populations in addition to tissue-specific cells
The earliest forms of bioprinting systems were modified
that cofunction to regulate ECM production, biochemical
conventional desktop inkjet printers. Inkjet printers are
signaling, and overall tissue functionality. Traditional
able to deposit a very small volume (1 300 pL) of liquid
seeding of single or multiple types of cells on a scaffold
onto a substrate through small nozzles. Inkjet bioprinters
is a simple approach; however, it is difficult to replicate
normally utilize a bioink consisting of cells in either cul-
the complex multicellular tissue composition that exists
ture medium or a cross-linkable hydrogel or can be acel-
in vivo. Coculture platforms are of increasing interest as
lular. Current inkjet-based bioprinters can print at speeds
they provide both cell cell (direct) and cytokine (indi-
in the range of hundreds of millimeters per second and
rect) interactions to occur, which result in more advanced
deposit hundreds of thousands of droplets per second,
stages of tissue formation [12]. Recent advances of 3D
with a resolution as high as 50 μm [15]. The limit of reso-
bioprinting now allow for precise spatial control over cel-
lution of inkjet printers relies both on the diameter of the
lular deposition for both scaffold-based and scaffold-free
nozzle and the viscosity of the bioink, thus reducing the
tissue models. For example, Liu et al. demonstrated that
diameter of the nozzle alone increases the likelihood of
hepatocytes alone will cease albumin and urea secretion
clogging. This limits the types of materials that are able
after 2 weeks; however, secretion is protected when pat-
to be printed using this method to low-viscosity
terned with fibroblasts and endothelial cells in the same
(B30 mPa s) or water-based materials [54]. There cur-
model even if not all cell types are in direct contact with
rently exist three different approaches to inkjet printing
one another [13]. Of particular interest is the patterning of
(Fig. 75.1):
endothelial cells within a construct to enhance vasculari-
zation [14] because as the engineered tissues grow, a 1. Thermal—Thermal inject printing heats a small vol-
necrotic core is likely to develop due to the diffusion lim- ume of the bioink (up to 300 C) for a matter of micro-
its of oxygen and nutrients to cells. Thus vascularized seconds to vaporize the liquid and inflate an air
scaffolds would allow for larger and more clinically rele- bubble to force the ink out of the nozzle head. This
vant engineered tissues to be fabricated. method is most commonly used as it has the highest
3D biofabrication incorporates several powerful cell viability after printing, is user-friendly, and gener-
approaches to constructing complex tissue models. ally inexpensive [19].
However, the appropriate choice of 3D bioprinting tech- 2. Piezoelectric—Piezoelectric inkjet printing uses a
nology, biomaterials, and cells must be considered a priori mechanical pulse generated by a piezoelectric actua-
in order to adequately represent the key mechanical, bio- tor. While no difference in viability of piezoelectric
chemical, and structural properties of the tissue of printed versus unprinted fibroblasts has been found,
 Biofabricated three-dimensional tissue models Chapter | 75 1419

TABLE 75.1 Comparison of different three-dimensional bioprinting techniques.

Inkjet printing Extrusion printing DLP printing TPP printing References

Printing Serial (drop-by- Serial (line-by-line) Parallel and Serial (dot-by-dot) [19 25]
process drop) continuous
(projection-based)
Printing speed Medium (mm/s) Slow (10 50 μm/s) Fast (mm3/s) Medium (mm/s) [26 29]
Resolution ,1 pL droplets; .5 μm 1 μm B200 nm [15,29 32]
50 μm wide
Biomaterials Low-viscosity Polymers (natural/ Polymers (natural/ Polymers (natural/ [15,31 35]
suspensions of synthetic), plastics, synthetic), cells, synthetic), cells,
cells, cells, proteins nanoparticles nanoparticles
biomolecules,
growth factors
Biomaterial Low viscosity, Shear thinning, low Photopolymerizable, Photopolymerizable, [36]
requirements rheopectic surface tension, contain low-toxicity contain low-toxicity
behavior, low adhesion, rapid photo-initiators, photo-initiators,
nonfibrous, cross-linking, shape stability, high stability, high
rapid cross- retention mechanical strength mechanical strength
linking
Biomaterial 3.5 12 mPa s 30 to 6 3 107 mPa s 1 300 mPa s 1 300 mPa s [15,34,37 39]
viscosity
Mechanical Poor due to Poor due to Excellent due to Poor due to [17]
integrity interfaces interfaces scanningless and interfaces
continuous printing
Cell viability .85% 40% 80% 85% 95% .85% [40 43]
Compatible Thermo/pH/ Thermo/ Photosensitive Photosensitive [44 47]
biomaterials photosensitive photosensitive
Printer Simple Moderate Complex Complex [21,23]
construction
Advantages Wide Many Fast, nozzle free, no Nozzle free, highest [1,15,21,33,48 51]
availability, printable materials, contact, precise fabrication
low cost, scaffold-free control of geometry, resolution, single cell
ability to fabrication, easy to precise control of manipulation
introduce print multiple mechanical
concentration materials properties, high cell
gradients simultaneously viability
Disadvantages Limited vertical Low viability, Moderate capital High capital cost, [1,22,27,33,52,53]
structure, critical timing of cost, material waste, limited to small
susceptible to polymerization, limited number of structures, material
clogging, requires matching biomaterials waste, limited
thermal/ material densities number of
mechanical to preserve shape biomaterials
stress on cells

DLP, Digital light processing; TPP, two-photon polymerization.

Source: Adapted from Murphy SV, Atala A. 3D bioprinting of tissues and organs. Nat Biotechnol 2014;32:773 85 [15]; Li J, Chen M, Fan X, Zhou H.
Recent advances in bioprinting techniques: approaches, applications and future prospects. J Transl Med 2016;14:271 [16]; Zhu W, Ma X, Gou M, Mei D,
Zhang K, Chen S. 3D printing of functional biomaterials for tissue engineering. Curr Opin Biotechnol 2016;40:103 12 [17]; Jana S, Lerman A. Bioprinting a
cardiac valve. Biotechnol Adv 2015;33:1503 21 [18].

there is a concern that the range of frequencies 3. Electromagnetic—Electromagnetic inkjet printing uti-
employed by these printers (15 25 kHz) can cause lizes miniature solenoid valves to dispense fluid [56].
damage to the cell membrane and induce cell lysis This method produces much larger drop volumes than
[19,55]. other inkjet printing methods [57].
1420 PART | TWENTY ONE Emerging technologies

FIGURE 75.1 Schematic of dif-

ferent 3D bioprinting platforms.

While initially there was a lot of interest in inkjet bio- sacrificial bioinks that can be washed away to leave
printing, there has been little development toward the fab- voids. The viscosity of the bioinks available for extrusion
rication of large tissue constructs. It is very difficult to printing is greater than that of inkjet printing, and allows
layer low-viscosity liquid droplets on top of a solid sur- for the fabrication of larger, mechanically stable structures.
face with high fidelity. In addition, the deposition of cells Certain bioinks can be thermally, chemically or photo-
results in high thermal or shear stress which can affect crosslinked to form solid structures and increase
viability. While this method is cheap and flexible, chal- handleability.
lenges pertaining to bioink composition, resolution, layer- Several approaches have been taken in order to make
ing, and printing speed must be resolved in order to build models robust and complex, in order to more accurately
complex 3D models that other 3D bioprinting modalities replicate native tissue structure. These printers can be
are capable of. constructed with several nozzles, each containing a differ-
ent bioink, which allows for multimaterial, multicellular
Extrusion printing constructs. This allows for both scaffold-based and
Extrusion-based printers dispense bioink in a continuous scaffold-free tissue models to be fabricated using the
manner at precise points in space. These bioprinters typi- same machine. In addition, coaxial printer heads allow for
cally use piston-, screw-, or pneumatic-based controls to simultaneous bioprinting of a core bioink within a sheath
regulate the rate of bioink deposition, with linear actua- bioink, which can be used to create microfibrous or hol-
tors controlling the precise x y z location of the nozzle low constructs.
or substrate (Fig. 75.1). The shape of the fabricated con- There are a few limitations to extrusion-based 3D
struct can be designed using traditional computer-aided printing. Encapsulated cells undergo shear stress during
design modeling software, which allows for designs extrusion, which limits cell viability. While extrusion-
inspired directly from computed tomography (CT) or based 3D-printing allows for fabrication using more
magnetic resonance imaging (MRI) images. viscous mediums, these mediums often require more pres-
A vast array of bioinks are available for use with this sure, which affects cell viability more than the nozzle
method including encapsulated cells, naturally derived diameter, sometimes leading to worse cell viability than
polymers, synthetic hydrogels of varying viscosity, and inkjet-based approaches [26]. While decreasing pressure
 Biofabricated three-dimensional tissue models Chapter | 75 1421

would increase cell viability, extrusion-based 3D bioprint- more accurately depict models associated with pathology
ing already has a very slow printing speed (10 50 μm/s). or disease. When printing cells, no shear forces are
Typically, larger nozzle sizes are used which decreases applied, resulting in higher cell viability than inkjet- or
resolution, but allows for overall larger structures to be extrusion-based bioprinters. Taken together, DLP-based
fabricated. Therefore bioink viscosity, nozzle diameter, bioprinters can be used to fabricate complex 3D structures
extrusion pressure, and printing speed must all be with fine features in a matter of seconds.
accounted for when fabricating cell-laden structures using
this method. However, structures that do not require
Two-photon polymerization based bioprinting
microscale features such as bone, cartilage, and skin can
be easily fabricated using this method. TPP-based bioprinting utilizes a focused, near-infrared,
femtosecond laser to polymerize a monomer solution
(Fig. 75.1). Polymerization only occurs at the peak inten-
Light-assisted bioprinting sity area of the laser focal spot, where the energy is inten-
Light-assisted bioprinting is based on spatially controlling sive enough to trigger nonlinear absorption of the
solidification of a liquid photopolymerizable material femtosecond laser in the monomer solution and lead to
using light. This technique is of increasing interest in TE, photopolymerization for printing. This results in resolu-
as it allows for the encapsulation of cells with high viabil- tion below the diffraction limit and allows for high fidel-
ity, in addition to its rapid fabrication speeds, and high ity nanostructures with features less than 100 nm. 3D
resolution, which altogether affords tight control over structures, including overhanging structures, can be fabri-
mechanical, physical, and chemical properties of the fab- cated by rastering through the monomer solution. The
ricated models. There are two main forms of light- tradeoff of such a high-resolution printer is a decrease in
assisted bioprinting: digital light processing (DLP) and the fabrication size and speed, although good cell viability
two-photon polymerization (TPP) based bioprinting. is preserved. Similar to inkjet- and extrusion-based prin-
ters, since this is a noncontinuous printing process,
Digital light processing based bioprinting fabricated structures have interfaces between photopoly-
The basic components of a DLP-based bioprinting plat- merized parts, resulting in compromised mechanical
form are a light source, a digital micromirror device integrity.
(DMD) chip, a motorized computer-controlled stage, and Both DLP- and TPP-based bioprinters suffer from few
a probe (Fig. 75.1). The light source (UV or visible) is limitations. However, since no nozzles are used, the
reflected at the DMD chip, which contains 1 4 million photopolymers typically reside in a reservoir from which
micromirrors that can be rotated to be “on” or “off” using the objects are printed and can result in wasted materials
a simple binary mask. The light that is reflected off of the and increased cost. In addition, light-assisted printers uti-
DMD chip passes through a series of optics into a solution lize photopolymers, which must be chemically modified
of photopolymerizable biomaterials resulting in simulta- in order to facilitate 3D printing, thus limiting the number
neous printing of an entire plane of an optical pattern. By and type of materials which can be used.
moving the stage or the light focal plane along the z-
direction, complex patterns can be fabricated in 3D.
Biomaterials for three-dimensional fabrication
DLP-based bioprinters have several advantages over
inkjet- and extrusion-based bioprinters, thus highlighting Biomaterials are engineered substances that can be fabri-
their versatility. Structures fabricated with DLP-based cated to interact with living biological systems either as
bioprinting are smooth and have higher mechanical integ- therapeutics or diagnostics [58]. In order to fabricate
rity due to the lack of artificial interfaces between depos- physiologically relevant tissue models, it is necessary to
ited materials that inherently exist in inkjet- and closely mimic the natural physical and chemical proper-
extrusion-based bioprinted constructs. Moreover, this ties of the tissue of interest. Therefore biomaterial selec-
method results in a significant time advantage and allows tion is a crucial part of the design process and can govern
structures to be fabricated in a matter of seconds to min- which 3D fabrication methods may be used.
utes. The resolution of DLP-based printers is dependent The biomaterials used in 3D bioprinting are often
upon the focal size of the reflected light beam and is usu- referred to as bioinks. These bioinks are most often
ally on the order of a few microns. The stiffness of these hydrogels—a hydrophilic network consisting of cross-
structures can be modified by increasing power from the linked polymeric chains. Hydrogels can contain an aque-
light source or by modifying the duration of exposure ous solution of up to 1000 times its original weight [59].
which affects the degree of polymerization of the fabri- The ability of hydrogels to absorb water arises from func-
cated structure. This can be used to create a single struc- tional hydrophilic groups attached to their polymeric
ture with multiple different stiffness profiles, which could backbone and their ability to retain a structure is due to
1422 PART | TWENTY ONE Emerging technologies

the cross-links between their polymer chains [59]. biomaterials are poly(ethylene glycol) (PEG), Pluronic
Increasing cross-linking density increases the strength of F-127, and poly(ε-caprolactone) (PCL). These materials
a hydrogel, as well as decreasing the amount of water that are most commonly used as scaffolding materials due to
can be absorbed. Due to its structure, the hydrogel typi- their ease of printability and robust mechanical strength
cally retains high permeability to oxygen and nutrients, compared to naturally derived biomaterials.
which makes it an attractive medium for 3D tissue scaf- PEG-based bioinks are among the most common
folds. 3D-printable biomaterials can be divided into two hydrogels used in biomedical applications as they are
categories: (1) naturally derived and (2) synthetically nontoxic, nonimmunogenic, and have easily tunable
derived. Many differences exist between the printability, mechanical properties via molecular weight modification.
biocompatibility, mechanical strength, and biochemical PEG alone has fairly poor mechanical properties, how-
components of naturally and synthetically derived bioma- ever, chemical addition of PEG diacrylate (PEGDA) or
terials. For a more detailed review of biomaterials used in PEG methacrylate (PEGMA) groups facilitate photo-
3D bioprinting, the reader is directed to the comprehen- crosslinking when in the presence of a photoinitiator,
sive reviews by Hospodiuk et al., Skardal and Atala, which results in increased mechanical strength. Therefore
Parak et al., and Choudhury et al. [36,60 62]. PEGDA and PEGMA are the most common forms of
PEG used in 3D biofabrication. The addition of photopo-
Naturally derived biomaterials lymerizable groups allows the mechanical properties of
PEG to be further tuned based on the duration and inten-
Naturally derived bioinks that have been used for 3D tis-
sity of UV light exposure [103] wherein longer duration
sue fabrication include collagen, gelatin, agarose, hyaluro-
and higher light intensity increases stiffness. Due to its
nic acid, silk proteins, chitosan, alginate, and
easy printability, fast photopolymerization (seconds), and
decellularized ECM (dECM). Naturally derived biomater-
relatively low cost, PEGDA and PEGMA are widely used
ials are an attractive bioink for 3D printing, as they con-
in inkjet-, extrusion-, DLP-, and TPP-based printing.
tain the intrinsic biophysical and biochemical components
Common applications of PEGDA and PEGMA include
of the native ECM. This results in increased cell adhesion,
cell encapsulation [114,115], biomimetic scaffold fabrica-
proliferation, differentiation, and migration. Structures
tion [102,103,107,108,116 118], and microfluidic
made with naturally derived bioinks generally have high
devices [119 121].
biocompatibility when integrated with native host tissues.
Pluronic F-127 is a nontoxic poloxamer compound
The main detraction to naturally derived biomaterials is
that undergoes reverse polymerization—increasing cross-
that they are often mechanically weak on their own and
linking with increasing temperature—and is compatible
difficult to print with precision. Therefore naturally
with extrusion-based bioprinting. Pluronic F-127 is also
derived biomaterials are sometimes combined with syn-
compatible with DLP- and TPP-based bioprinting after
thetic biomaterials to form composites. In addition, as nat-
chemical modification [122,123] or combination with
urally derived biomaterials are derived from actual
other photopolymerizable materials [124,125]. The tem-
tissues, they are susceptible to batch-to-batch variability,
perature of cross-linking varies between 10 C and 40 C,
which effects model consistency. Nearly all naturally
depending on the molar mass, percentage of composites,
derived bioinks are compatible with inkjet- and extrusion-
and functionality [124,126]. This requires heating systems
based bioprinters but require chemical modification in
to be in place during the fabrication process, which, in
order to be compatible with DLP- and TPP-based bioprin-
combination with its relatively high viscosity, have pre-
ters. A list of commonly used naturally derived biomater-
cluded its use with inkjet-based bioprinters [36]. Pluronic
ials and their uses can be found in Table 75.2.
F-127 has been used to encapsulate cells, although it typi-
cally degrades fairly quickly (within hours), making it
Synthetically derived biomaterials more useful for cell delivery than as a long-term scaffold
Synthetically derived biomaterials are often used for 3D [127]. The rapid degradation time and reversible polymer-
tissue fabrication as they can be consistently produced ization characteristics of Pluronic F-127 make it an attrac-
and allow for easy control over mechanical properties, tive biomaterial for drug delivery and controlled release
degradation rate, and printability. These materials can applications [128], as well as serving as a sacrificial
also be easily combined with nanoparticles in order to bioink. Pluronic F-127 can be printed during the initial
create functionalized scaffolds such as detoxification 3D fabrication process, only to be washed out after the
devices [113]. In addition, the backbones of these syn- construct acquires enough rigidity to retain shape
thetic materials can be chemically modified to include [123,129 131]. This allows for the relatively simple con-
cell-binding moieties such as RGD and YISGR in order struction of conduits or voids within a construct.
to improve biocompatibility and cellular integration. A PCL is a thermoplastic commonly used with inkjet-
few of the most commonly used, synthetically derived and extrusion-based bioprinting platforms. PCL has a
TABLE 75.2 List of natural polymers commonly used for biofabrication.

Collagen Alginate Hyaluronic acid Matrigel

Printability Moderate Easy Moderate Moderate
Printer compatibility Inkjet, extrusion, DLP Inkjet, extrusion, DLP Inkjet, extrusion, DLP, Extrusion
TPP
Cross-linking method Thermo, pH, photo Ionic (Ca21), photo Chemical, photo Thermo
Cross-linking speed Seconds 1 h Seconds 5 30 min 20 min 1 h
Chemical Methacrylation [63 65] Methacrylation [66] Methacrylation [67,68] N/A
modification for Thiol(-ene) [60,69]
photopolymerization
(refs)
Advantages Promotes cell adhesion High biocompatibility, Enhanced Contains growth
and expansion, strong cheap, nonimmunogenic, chondrogenesis/ factors, facilitates
in vitro/in vivo fast cross-linking time, osteogenesis, cell growth/
biocompatibility, generally low viscosity biocompatible, adhesion
nonimmunogenic biodegradable, high
solubility
Disadvantages Slow cross-linking time Poor cell adhesion (no Must be chemically Poor mechanical
(0.5 1 h), natural cell-adhesive modified, poor properties, must be
nonhomogeneous cells moieties), slow mechanical properties, combined with
distribution degradation slow cross-linking other materials
Common Cell encapsulation, Composite bioinks, cell Cell encapsulation, cell Basement
applications functionalize material encapsulation, tubes, delivery, drug delivery, membrane,
surfaces strands wound healing increase
biocompatibility,
promote cell
growth
References [27,47,70 74] [75 85] [60,68,69,86 90] [91 93]

Agarose Fibrin Decellularized ECM Gelatin methacrylate

Printability Easy Easy Moderate Easy
Printer compatibility Inkjet, extrusion Inkjet, extrusion Inkjet, extrusion, DLP Inkjet, extrusion, DLP, TPP
Cross-linking method Thermo Fibrinogen 1 thrombin Thermo Thermo, photo
Cross-linking speed Seconds minutes Seconds Minutes Seconds
Chemical N/A N/A Mix w/other Methacrylation
modification for photopolymerizable
photopolymerization materials
(refs)
Advantages Fast polymerization, Fast polymerization, Contains native growth Retains cell-binding
thermally reversible highly adhesive, cell factors of target tissue, motifs, fast
cross-linking adherent, binds to high cell adherence, polymerization, good cell
growth factors high cell proliferation viability, tunable
mechanical properties
Disadvantages No cell adherence Poor mechanical Poor mechanical Not as durable as synthetic
proteins, poor properties, difficult to properties, batch-to- polymers
mechanical properties control geometry batch variation, slow
tunable by cross-linking time
concentration
Common Cell encapsulation, Adhesive, cell Incorporation with other Widely used, cell
applications structural support/ encapsulation, cell materials for tissue- encapsulation, increasing
scaffolding delivery specific cellular biocompatibility
response
References [91,94 98] [60,99 101] [102 106] [107 112]

DLP, Digital light processing; ECM, extracellular matrix; TPP, two-photon polymerization.
1424 PART | TWENTY ONE Emerging technologies

relatively low melting temperature of 60 C [132] and is for primary cells as this will affect the consistency in the
commonly used as a primary structural component in functional response of the models.
scaffolds [129]. PCL does not contain any natural peptide To address the issues regarding primary cells, cell
sequence motifs, so it is largely used in conjunction with lines are a convenient substitute as they have been altered
other naturally derived polymers or functionalized materi- via viral transfection to be able to proliferate in vitro
als to create composite structures [133,134]. As PCL is indefinitely by preventing normal cellular senescence,
relatively stiff compared to other synthetic polymers, it is thus enabling large populations to be cultivated. To date,
typically used in cartilage and bone tissue engineering a plethora of cell lines have been established for several
[135,136]. tissues including the immortalized normal adult kidney
cell line human kidney 2 cells, human liver carcinoma
cell line HepG2, human cervical cancer HeLa, mouse car-
Cell selection diac muscle cell line HL-1, and mouse fibroblast cell line
A critical step in the design of 3D tissue models is the 3T3 cells. While cell lines serve as important biological
appropriate selection of cell sources as this directly influ- tools by offering several advantages such as ease of use,
ences the performance as well as accuracy and relevancy reproducibility, cost-effectiveness, unlimited supply, and
of the resulting model. Moreover, the cells chosen for the circumventing ethical concerns regarding human and ani-
model must be able to replicate the key physiological mal tissue use, it is important to recognize that although
characteristics of the normal or pathological states for the cell lines display functional features of primary cells they
tissue of interest. Given these criteria, there are several are not identical in behavior [143]. This is because cell
cell sources available that can be classified as either pri- lines have been genetically manipulated and serial passag-
mary cells, cell lines, or stem cell derived cells. In the ing can result in variations of their native genotypic and
context of tissue biofabrication, it is also important that phenotypic expression profiles, in addition to changes in
the cell source selected is capable of ex vivo expansion to response to different stimuli [143]. For example, it was
meet the requirement of high cell numbers for bioprinting found in a comparative study that using the human hepa-
often ranging between 1 3 106 and 1 3 108 cells/mL toma cell lines, HepG2 and HepaRG cells, exhibited
depending on the application [137]. Reproducibility and reduced sensitivity to drug toxicity compounds compared
consistency of the cells produced in terms of maintaining to primary human hepatocytes [144]. As such, these ideal-
the desired phenotype and function are equally important ized cell types provide limited predictive value as they
postexpansion as well as during in vitro culture in a 3D have poor resemblance to native primary cells, and it is
format post fabrication. Other factors to consider also imperative that the data collected from in vitro models
include the ability for the cells to survive the physical produced from cell lines to be interpreted with caution.
stressors brought upon during the bioprinting process as More recently, the use of stem cell derived cells origi-
well as continued self-renewal thereafter to remodel the nating from embryonic stem cells, pluripotent stem cells,
tissue construct and maintain appropriate cellular density or adult stem cells has gained popularity as a viable source
in long-term culture. due to their self-renewing properties and differentiation
Primary cells are terminally differentiated and have capacity into various functional cell types. Embryonic stem
the quality of being a direct representation of the pheno- cells are totipotent thus enabling them to differentiate into
type, maturation state, and functional features of the specialized cell lineages derived from any of the three
native tissue functional unit [138]. These cells are har- embryonic germ layers [145]. For instance, this includes
vested from patient biopsies and have been reported to be plasticity to differentiate into cells of cardiac, neuronal, as
used successfully for generating bioprinted liver, carti- well as hematopoietic origins. Embryonic stem cells can be
lage, and skin tissues [139 141]. However, because these cultured continuously for long periods of time in the undif-
cells need to be isolated, a major drawback is their limited ferentiated state and are practical for generating large
availability as small samples can only be harvested at a populations, thus enabling their potentiality to provide an
time to avoid donor site morbidity and the procedures per- unlimited number of any specialized cell type [145].
formed can be complex and invasive [142]. Furthermore, However, with ethical concerns regarding the use of human
primary cell culture usage is further complicated by their embryos in addition to risks of teratoma formation and
limited capability for ex vivo culture as these cell types allogenic origin, these factors may hinder their application
often lack in vitro proliferative capacity, are sensitive to for cell-based regenerative therapies [146]. In the last
culture conditions, prone to dedifferentiation, and suscep- decade, techniques to generate iPSCs pioneered by
tible to changes in phenotype expression over prolonged Takahashi and Yamanaka resulted in a paradigm shift in
culture [142]. For the generation of reproducible tissue the cell source landscape [147]. By introducing the four
models, inherent donor-to-donor variability and risk of factors Klf4, Sox2, Oct3/4, and Myc this method enabled
obtaining cells in a diseased state also poses a challenge researchers to reprogram adult cells into iPSCs that
 Biofabricated three-dimensional tissue models Chapter | 75 1425

were capable of pluripotent differentiation and possess intricate layers and compartments. Understanding the
characteristics similar to embryonic stem cells with exten- development, structure, and degenerative disorders which
sive self-renewal capacity [147]. Currently, a large range of affect the brain is the main goal of neurobiology. Due to
protocols have been developed to produce specialized cell practical and ethical reasons, studying neural structure,
types from easily accessible human fibroblast, adipose- development, and disease has been largely reliant on ani-
derived stem cell (ADSC), or peripheral blood cell sources mal studies consisting of mainly rodent models. Brain
into hepatocytes, neural stem cells (NSCs), pancreatic cells, structure and development vary greatly between animal
and gastric epithelial cells [148]. The use of iPSCs also gar- and human models and thus has been a large motivation
ners a significant advantage by not only overcoming the for the development of in vitro models of the human
ethical concerns associated with embryonic stem cells but brain.
also enabling the possibility of patient-specific cell therapy 3D bioprinting neural structures present a number of
for clinical applications. More specifically, cells can be challenges that are unique to accurately replicate brain
obtained from patients to produce patient-specific tissue structure and organization. This may partially contribute
models for developing personalized medicine for the to the lack of successful 3D fabricated models in brain tis-
treatment of various rare diseases and elucidation of their sue compared to other more simplistic tissues. The human
associated pathological mechanisms. Finally, adult stem brain consists of roughly 86 billion neurons and 85 billion
cells which can be autologously or allogeneically har- nonneuronal cells [156]. The nonneuronal cells, referred
vested from adult tissues such as adipose [149], bone to as glial cells, include astrocytes, microglia, oligoden-
[150], umbilical cord blood [151], and skin [152] have drocytes, endothelial cells, and pericytes [157]. Glial cells
shown promise as a viable regenerative cell source in perform a wide spectrum of functions throughout the
recent years. Unlike embryonic stem cells and iPSCs, brain in order to maintain normal neuronal function.
adult stem cell populations are multipotent rendering Damage to or imbalance of these cells leads to neurode-
them capable of differentiating into two or more cell generative disorders in vivo. Therefore in order to
types but are more limited in lineage plasticity. For develop accurate in vitro models, the precise number and
instance, cells such as ADSCs and bone marrow derived distribution of multiple cell types must be simultaneously
stem cells have been demonstrated to successfully differ- accounted for, which is challenging but not impossible
entiate toward the adipogenic, chondrogenic, and osteo- with 3D bioprinting.
genic lineages in vitro [153,154]. Although promising as The capacity to differentiate into multiple cell types
an accessible cell source, adult stem cells are still met and the self-renewal capacity of NSCs have made them
with challenges in terms of variability in self-renewal attractive candidates for generating models of brain tissue.
capacity, differentiation potency, donor-to-donor differences, Using an extrusion-based printer, Gu et al. printed human
as well as the steps taken in the isolation, identification, and NSCs differentiated in situ into a complex structure con-
purification of the multipotent stem cell population from the sisting of neurons, as well as astrocytes and oligodendro-
stroma [155]. As a result, many of these factors pose the cytes [158]. Mature-fabricated models contained rounded
same issues as primary cells by affecting their reproducibil- soma and extensive neurite outgrowth, similar to native
ity and predictive capability which may compromise the neuronal morphology. These models also exhibited
quality of the final tissue model. bicuculline-induced increased calcium response, thus
demonstrating that the differentiated neurons had formed
into functional GABA receptors [159]. Hsieh et al. also
Three-dimensional tissue models for drug utilized extrusion-based bioprinting to develop a unique
screening, disease modeling, polyurethane bioink containing NSC. When implanted
therapeutics, and toxicology into a zebrafish model of traumatic brain injury, the zeb-
rafish recovered normal swimming ability and experi-
In the following sections the application of 3D biofabri-
enced reduced mortality [160].
cated constructs for in vitro disease modeling, high
Overall, 3D bioprinting brain tissue is still in its
throughput drug testing platforms, and in vivo therapeu-
infancy compared to other methods of in vitro brain
tics will be discussed. Specifically, the implementation of
modeling. The most complex models of the brain are
various biofabrication strategies to recapitulate both nor-
cerebral organoids. Cerebral organoids are a scaffold-free,
mal and diseased brain, nerve, cancer, heart, liver, and
human iPSC derived “mini-brain” model, which contains
vascularized tissue will be reviewed.
various discrete, interdependent regions, fabricated using
a hanging drop culture environment [161]. The culturing
Brain and nerve tissue models conditions promote self-organization and self-patterning
The brain is the most complex tissue in the body, consist- of iPSCs into various brain regions without the use of
ing of numerous different cell types organized into exogenous patterning factors. Recently, these organoids
1426 PART | TWENTY ONE Emerging technologies

have been combined with poly (lactic-co-glycolic acid) function are involved in pathology and progression of
(PLGA) microfilaments to enhance neuroectoderm forma- disease [163]. To develop an in vitro model of the BBB,
tion and form a distinct cortical plate, which is considered Nzou et al. utilized a hanging drop culture environment
the final step in corticogenesis [162]. These microfilament- to develop an organoid model comprised of the six con-
engineered cerebral organoids formed distinct layers and stituent cell types found within the cortex of the brain
had excellent neuronal organization compared to cere- [164]. The resulting spheroids are capable of junction
bral organoids alone (Fig. 75.2). A limitation of these formation and selective permeation of ions, making this
models is that they do not contain microvasculature, model attractive for testing a drug candidate’s ability to
which precludes the study of the blood brain barrier cross the BBB.
(BBB). The BBB is a highly selective semipermeable Nerve injuries can be grossly classified into two gen-
membrane that regulates ion, molecule, and cell transfer eral categories: (1) peripheral nerve injuries and (2) spinal
from vasculature to the brain, and alterations in its cord injuries. In particular, peripheral nerve injuries

FIGURE 75.2 Brain and nerve tissue models. (A) Schematic of the method for generating the engineered cerebral organoids. (B)
Immunohistochemical staining for laminin and the neuronal markers MAP2 and Ctip2 in engineered cerebral organoids. The organization of laminin,
MAP2, and Ctip2 in the cerebral organoid containing microfilaments (bottom) is typical of the CP in vivo. (C) Nestin staining reveals long radial glia
(arrowhead) organization that terminates outside the CP and MZ at the outside of the organoid. (D) Left: Axial slice of a spinal cord depicting the
regions of white and gray matter. Center: 3D bioprinted scaffold with physiologically informed geometry. Scaffold was printed using a DLP-based
bioprinter in 1.6 seconds. Right: A cross section through an implanted scaffold, 4 weeks after implantation. Green indicates axons. Scale bar 5 200 μm.
3D, Three-dimensional; CP, cortical plate; DLP, digital light processing; MZ, marginal zone. Reproduced with permission from (A C) Springer
Nature (2017) [162] and (D) Springer Nature (2019) [118].
 Biofabricated three-dimensional tissue models Chapter | 75 1427

require approximately 200,000 surgeries annually in the models for studying the complex tumor microenvironment
United States [165,166]. Current clinical approaches to (TME) in humans. While conventional 2D cancer models
healing nerve injuries involve trying to suture the proxi- have yielded some promising results [172], they inade-
mal and distal ends of the injury (,5 mm gap), or quately mimic the 3D TME [173,174]. This has given rise
through autologous or cadaveric nerve transposition to the use of 3D bioprinting as a technique to fabricate
[167,168]. While autografts are the current gold standard, in vitro cancer models to study tumor proliferation,
they require sacrificing nerve tissue from another part of metastasis, and drug response.
the body and the resulting nerve may not actually fit the TME is highly complex, heterogeneous, and disease
injury site [165]. This has motivated the development of dependent, affecting cell proliferation and tumor charac-
peripheral nerve guides, to try to bridge the gap in a teristics. Zhao et al. utilized 3D extrusion-based printing
peripheral nerve injury [167,168]. Zhu et al. utilized a to investigate differences in HeLa cell proliferation,
DLP-based 3D bioprinter to fabricate peripheral nerve matrix metalloproteinase (MMP) production, and che-
guides with anatomically informed geometry [117]. In a moresistance in 2D planar culture versus 3D tumor
rat injury model of peripheral nerve injury (sciatic tran- modeling [175]. HeLa cells were found to have higher
section) they were able to demonstrate improved function- proliferation and formed spheroids in 3D compared to
ality of motor and sensory function, as well as forming a monolayer in 2D culture. In addition, HeLa
histological verification of nerve growth into the scaffold. cells in 3D had higher MMP production and were more
Spinal cord injuries affect an estimated 285,000 people in resistant to paclitaxel, a commonly used chemotherapy
the United States and have poorer outcomes and fewer drug. Swaminathan et al. expanded upon this work to
surgical procedures available than peripheral nerve inju- demonstrate that preformed spheroids prior to printing,
ries [169]. Koffler et al. expanded upon the prior study demonstrated greater resistance to paclitaxel than individ-
and fabricated nerve guides for spinal cord injury repair ually printed cells in a breast cancer model [176]. The
using the same DLP-based 3D bioprinter [118]. The nerve findings from these studies suggest that more biomimetic
guide geometry was informed by the anatomical structure cell cell and cell matrix interactions may contribute to
of the spinal cord, and additionally included “neural the differences in cancer cell behavior and function in 2D
relays” consisting of neural progenitor cells integrated versus 3D microenvironments.
into the scaffold (Fig. 75.2). In a rat spinal cord injury Tumor metastasis leads to 90% of cancer death and
model (T3 spinal cord transection), implanted scaffolds decreases 5-year survival rates. Therefore there has been
were found to integrate with native injured host axons much interest in evaluating how TME affects tumor
and significantly improve functional outcomes. metastasis in vitro. Using a DLP-based 3D printing plat-
Importantly, both of these studies demonstrated the capa- form, Soman et al. fabricated a scaffold with a log-pile
bility to print human-sized scaffolds, as well as scaffolds architecture with tunable material properties [177]. The
with geometry extracted from MRI images. These patient- differences in cell migration in normal and TWIST
specific scaffolds hold promise for a future clinical trans- oncogene transformed breast endothelial cells were eval-
lation of 3D bioprinting. uated in 2D and 3D in scaffolds with soft or stiff sub-
While 3D fabrication strategies for nerve are promis- strates. They found that while no differences in cell
ing, complex models of 3D fabricated brain tissue still migration were found in 2D between soft and stiff sub-
need to be developed. Brain tissue contains such a vast strates, substantial differences were found between soft
array of cells and has a unique microenvironment com- and stiff substrates for the cancer cells in 3D. This is
pared to the rest of the body. New biofabrication strate- important as it demonstrates that observations conducted
gies must therefore focus on recapitulating the layered, in a 2D system cannot be directly extrapolated to 3D.
complex organization of neural tissue in models that are Huang et al. further evaluated cancer metastasis utilizing
capable of being handled. a DLP-based bioprinter by fabricating a biomimetic
microstructure with vasculature informed channel sizes
(Fig. 75.3) [121]. The migration behavior of normal fibro-
Cancer models blast (10T1/2) and HeLa cells was evaluated in channels
Over 18 million new cases of cancer and over 9 million with 25, 45, and 120 μm width. Channel width had no
cancer-related deaths were estimated to have occurred in effect on average migratory speed of fibroblasts, but
2018 [170]. Despite its prevalence, the lack of under- HeLa cell speed was found to significantly decrease with
standing of tumorigenesis and metastases affects the abil- increasing channel width (Fig. 75.3). This suggests that
ity to develop successful drugs to fight cancer. Currently, the different responses of normal and cancerous cells to
the percentage of cancer drugs that transition into success- different geometric environments could be used as a tool
ful clinical therapeutics is about 8% [171]. This is par- to screen cancerous cells. These studies demonstrate how
tially due to the lack of suitable in vitro and animal the 3D environment can drastically change the behavior
1428 PART | TWENTY ONE Emerging technologies

FIGURE 75.3 Cancer tissue models. (A) Optical microscope images of HeLa cells seeded on fabricated PEGDA microstructures. Scale
bar 5 100 μm. (B) Average instantaneous speed of HeLa cells (left) and 10T1/2 cells (right) cultured on microstructures. Channel width had a signifi-
cant effect on instantaneous cell speed for cancer cells but not fibroblasts. (C) Bioprinted constructs consisting of primary breast cancer cells (21PT;
green) and ADMSCs (red). Constructs were fabricated with thin, moderate, and thick layers of ADMSCs around the 21PT printed disk. Scale
bar 5 250 μm. (D) Density of Caspase-3 positive (apoptotic) 21PT cells in constructs with thin, moderate, and thick layers of ADMSCs after treatment
with doxorubicin (chemotherapy drug). Increasing the amount of ADMSCs around 21PT cells decreases the sensitivity of 21PT cells to doxorubicin.
ADMSCs, Adipose-derive mesenchymal stem/stromal cells; PEGDA, poly(ethylene glycol) diacrylate. (A and B) Reproduced with permission from
Springer Science 1 Business Media (2014) [121]; (C and D) reprinted (adapted) with permission from Heinrich MA, Bansal R, Lammers T, Zhang
YS, Michel Schiffelers R, Prakash J. 3D-bioprinted mini-brain: a glioblastoma model to study cellular interactions and therapeutics. Adv Mater
2019;31:e1806590. r2018 American Chemical Society [179].

of cancer cell mobility and can be used to inform future with macrophages were found to have a much higher
studies investigating tumor metastasis. growth rate than tumors cultured alone. These studies
Stroma tumor interactions have been increasingly demonstrate that the inclusion of stromal cells into the
identified as a key factor in treatment response to various TME enhances clinical relevance of the models in vitro.
drugs. In fact, stroma-induced drug resistance and stroma- 3D bioprinting of cancer models has the potential to
induced synthetic lethality are linked to the TME with provide insight into how the TME affects tumor develop-
which tumor cells interact [178]. Wang et al. demon- ment, behavior, metastasis, and invasion. Future works
strated that the addition of a layer of stromal cells around focusing on patient-specific cells have the potential to
a disk of bioprinted breast cancer models decreased the provide more insight on patient-specific as well as stage-
breast cancer cell’s sensitivity to a chemotherapy drug specific behavior of various types of cancer. Further, 3D
(doxorubicin) (Fig. 75.3) [179]. Further, the thicker the bioprinting patient-specific in vitro tumor models has the
layer of stromal cells, the less breast cancer cell apoptosis tremendous potential to screen drugs which may or may
was observed. This printing system consisted of multiple not be effective, potentially resolving the disease faster
syringe extrusion-based printer, separate bioinks for each and improving patient outcomes. However, additional
cell type, each encapsulated in a combination of metha- work is needed to develop the materials and cell sources
crylated gelatin, hyaluronic acid, and gelatin. During the in order to create repeatable, physiologically relevant
extrusion process, the construct had to be photo- tumor models.
crosslinked three separate times in order to fabricate a
stable structure. Heinrich et al. recently utilized a similar
setup in order to investigate the relationship between Heart tissue models
glioblastoma-associated macrophages and glioblastoma Worldwide, cardiovascular diseases (CVD) are a major
multiforme [180]. They printed a “mini-brain” consisting cause of morbidity and mortality [181]. Myocardial
of a glioblastoma tumor surrounded by a large number of infarction (MI) and valvular heart disease (VHD) are
macrophages. They demonstrated that glioblastoma cells some of the main forms of CVD. In the United States, an
actively recruit macrophages and polarize them to have a estimated 720,000 people experience a new MI with
more migratory phenotype. In addition, tumors cocultured 335,000 having a recurrent MI annually [181] and an
 Biofabricated three-dimensional tissue models Chapter | 75 1429

estimated 2.5% of the population are thought to have along the direction of the patterning (Fig. 75.4).
VHD [182]. As cardiac muscle has limited capacity for Furthermore, the deflection of pillar structures in which
self-repair, current treatment strategies include grafting the cardiomyocytes were printed onto enabled analysis of
diseased tissues or inserting artificial prostheses to cardiac force generation and drug response in a high
attempt to restore function [183]. However, each of these throughput manner [116]. These studies demonstrate the
approaches is complicated by their own respective disad- ability to rapidly fabricate in vitro cardiac models, capa-
vantages which include lack of donor tissue, immune ble of supporting rapid drug testing. Some drugs may dis-
rejection, anticoagulation therapy, and limited durability proportionately benefit or harm certain genotypes,
[184]. These factors have driven a large amount of ethnicities, sexes, and ages [192]. Since the cardiomyo-
resources to develop predictive preclinical platforms for cytes used in this platform are derived from human
cardiac drug testing, as well as durable patient-specific iPSCs, this allows for in vitro drug testing for all potential
therapies to treat CVD and physical models for presurgi- patient populations.
cal planning [185,186]. During MI, reduced blood flow to the heart results in
Current models of 3D cardiac tissue focus around ischemia and subsequently leads to large necrotic patches
seeding cells atop, or encapsulating cells within a of heart tissue. TE cardiac patches are designed to mimic
hydrogel-based scaffold [52,187], then utilizing 3D the native ECM and provide mechanical support as well
aligned cultures [188], mechanical stretching [189], or as delivery of cells to the area of MI. For example, Jang
electrical pacing [189,190] to promote mature cardiomyo- et al. utilized a dual nozzle, extrusion-based printer to
cyte phenotypes. However, by using 3D printing, it is pos- fabricate a prevascularized stem cell patch [193]. Both
sible to achieve mature aligned cardiomyocytes without cardiac progenitor cells (CPCs) and mesenchymal stem
any external stimulus. For instance, Liu et al. employed a cells (MSCs) were printed in a dECM bioink containing
DLP-based printer to directly print human iPSC derived vascular endothelial growth factor (VEGF), and cultured
cardiomyocytes into aligned patterned slabs suspended for 5 days. Using a standardized model of MI in a rat,
between two pillars [191]. While both slabs and aligned implants were surgically implanted before sacrifice at up
cardiomyocytes exhibited spontaneous synchronous beat- to 56 days’ postinjury. Patterned patches were found to
ing, cardiomyocytes printed in slabs beat without direc- increase ejection fraction and decrease fibrosis compared
tional preference, whereas aligned cardiomyocytes beat to patches consisting of simply mixed CPCs and MSCs.

FIGURE 75.4 Heart tissue models. (A) Top: Two sets of digital patterns including a slab and aligned patterns for encapsulated cell printing with a
DLP-based bioprinter. Bottom: The masks for the multilayer pattern to produce 3D scaffolds. First, a hyaluronic acid base layer is printed; second, a
gelatin methacrylate cantilever system; and third, parallel lines of hESC-CMs. (B) Optical image of the resulting patterned hESC-CMs printed in the
force gauge scaffold after culture for 21 days. Scale bar 5 500 μm. (C) Designed fiber bundle structure for muscle organization. PCL pillars (green)
were used to maintain the structure and to induce the compaction phenomenon for cell alignment. (D) 3D patterning outcome of designed muscle
organization (left) before and (after) removing the sacrificial material (Pluronic F-127). The printed construct was cross-linked with thrombin solution
to induce gelation of fibrinogen and the uncross-linked sacrificial material was removed by dissolving with cold medium. (E) The Live/Dead staining
of the encapsulated cells in the fiber structure indicates high cell viability after the printing process (green: live cells; red: dead cells). (F)
Immunofluorescent staining for myosin heavy chain of the 3D printed muscle organization after 7 days of differentiation. The encapsulated myoblasts
aligned along the longitudinal direction of the fiber structure. 3D, Three-dimensional; DLP, digital light processing; hESC-CM, human embryonic
stem cell derived cardiomyocytes; PCL, poly(ε-caprolactone). Reproduced with permission from (A and B) Elsevier (2019) [191] and (C F)
Springer Nature (2016) [129].
1430 PART | TWENTY ONE Emerging technologies

Furthermore, revascularization of the damaged myocar- Liver tissue models

dium was observed. These results demonstrate that an Liver is a densely populated organ composed of approxi-
organized 3D bioprinted scaffold can not only improve mately 120 million cells/gram in humans and is made up
cardiac function via mechanical support but also promote of mainly hepatocytes as well as several tissue-specific
neovascularization of the damaged tissue. cell types including Kupffer cells, epithelial cells, stromal
Cardiac valves play a major role in the proper func- cells, and sinusoidal endothelial cells [198]. These cells
tioning of the cardiovascular system by ensuring proper are arranged into highly organized hexagonal hepatic lob-
blood flow through the four chambers of the heart. In ule functional units. Within each hepatic lobule, hepato-
severe cases of VHD, valves must be surgically cytes are situated into plates that radiate in an outward
replaced. There is increased interest in developing a TE direction from the central vein with portal triads located
heart valve rather than relying on mechanical prosthet- at each vertex composed of the hepatic arteriole, portal
ics, donor supplied allografts, or porcine/bovine xeno- venule, and bile duct. Venous blood and arterial blood
grafts. Since cardiac valves have highly specific enter through the portal vein and hepatic artery, respec-
geometry integral to proper function, 3D bioprinting has tively, and flow through the sinusoidal regions and drain
been sought as an attractive fabrication method. into the central vein of each hepatic lobule.
Hockaday et al. used an extrusion-based 3D bioprinter Since liver is a major site for drug adsorption, distri-
to fabricate a biocompatible heterogeneous valve with bution, metabolism, and excretion there is a critical need
flexible leaflets and a rigid root [194]. Duan et al. for novel liver tissue models as tools to more accurately
implemented a similar model using extrusion-based 3D predict drug metabolism, pharmacokinetics, and hepato-
bioprinting and incorporated valvular interstitial cells toxicity for the development of potential new therapies.
and smooth muscle cells encapsulated in a combination Namely, this demand has largely driven strong motivation
of methacrylated hyaluronic acid and methacrylated gel- in the pharmaceutical industry for the development of bet-
atin rather than a synthetic polymer [195]. Further work ter liver models for drug-induced liver injury (DILI) as
by van der Valk et al. sought to use a similar approach this is a major cause for drug failure resulting in liver
by encapsulating valvular interstitial cells in order to transplantation or morbidity in patients [199]. To address
develop an in vitro model of calcific aortic valve disease this issue, it was recently shown that a 3D bioprinted liver
(CAVD), the most prevalent form of VHD [196]. tissue using patient-derived primary hepatocytes and non-
Microdissection and careful mechanical testing of a parenchymal endothelial and hepatic stellate cells were
cadaveric human ventricle with CAVD was performed suitable as a multicellular model for DILI [139]. Using a
to accurately model the disease progression. Mechanical scaffold-free assembly process of cellular spheroids,
properties of the fabricated construct were varied by Nguyen et al. patterned the different cells types using an
modulating the ratios of the main constituents of their extrusion-based bioprinter into a microarchitecture com-
bioink [gelatin methacrylate (GelMA) and methacry- posed of several compartments filled with hepatocytes
lated hyaluronic acid]. The resulting multilayered 3D and surrounded by nonparenchymal cells (Fig. 75.5).
construct recapitulated the layer-specific mechanical Using this bioprinted liver, they demonstrated that the
properties of the dissected valve and provided a novel printed tissue was capable of maintaining long-term func-
3D model for studying CAVD. tion over 4 weeks in culture including albumin secretion
Overall, 3D bioprinting is a promising technique for and expression of drug-induced enzyme activities of cyto-
the fabrication of in vitro cardiac models to in vivo car- chrome P450s [139]. In particular, it was shown that
diac therapies. A common thread between successful plat- model was sensitive to trovafloxacin dose-dependent tox-
forms is the importance of cellular alignment and icity at clinically relevant doses, which is normally unde-
mechanical properties of the scaffolds, which are integral tected in standard in vitro systems [200]. Similarly, Ma
features of the heart architecture. These features are not et al. developed a complex heterogeneous liver tissue
only integral factors to consider in cardiac muscle but in model that uniquely replicates the physiological micro-
skeletal muscle as well. 3D fabrication of skeletal muscle architecture and microscale features of the hexagonal
models for in vitro disease modeling and in vivo muscle lobule unit for drug screening [108]. In this work, a DLP-
repair consistently demonstrates that fiber alignment dur- based 3D bioprinter was employed to pattern human
ing the initial formation of a scaffold leads to organized iPSC derived hepatic progenitor cells (iPSC-HPCs)
muscle tissue with superior functional outcomes within a GelMA matrix, as well as human umbilical vein
(Fig. 75.4) [129,197]. In tissues such as cardiac and skele- endothelial cells (HUVECs) combined with ADSCs in a
tal muscle, where aligned cellular organization is essential GelMA and glycidyl methacrylate hyaluronic acid (GM-
for normal tissue function, 3D printing is a key tool for HA) matrix to serve as the supporting cell population to
fabricating physiologically accurate tissues. form a triculture 3D liver model as shown in Fig. 75.5.
 Biofabricated three-dimensional tissue models Chapter | 75 1431

FIGURE 75.5 Liver tissue models. (A) Scaffold-free spheroid-based bioprinted liver tissue and (B) representative H&E image showing compartmen-
talization of the hepatocytes and nonparenchymal cell populations indicated by the dashed line. Scale bar 5 25 μm. (C) Images of H&E stained bio-
printed liver tissues with and without trovafloxacin treatment. (D) Schematic of a DLP-based 3D bioprinter, (E) hexagonally shaped digital patterns
used to fabricate the liver tissue model, and (F) resulting fluorescent images of the printed liver tissue showing iPSC-HPCs (green) surrounded by sup-
portive cells (red). Scale bar 5 500 μm. (G) Representative Live/Dead images of HepG2 cells encapsulated in varying stiffnesses of liver
dECM GelMA hydrogels. Scale bar 5 500 μm. (H) Images of CellTracker stained HepG2 cells to visualize their invasion into surrounding stromal
regions over time from soft (red), medium (green), and stiff (yellow) conditions. Scale bar 5 500 μm. 3D, Three-dimensional; dECM, decellularized
extracellular matrix; DLP, digital light processing; GelMA, gelatin methacrylate; iPSC-HPCs, induced pluripotent stem cells derived hepatic progeni-
tor cells. Reproduced with permission from (A C) PLoS One (2016) [139]; (D F) National Academy of Sciences, r2016 [108]; and (G and H)
Elsevier (2018) [103].

Compared to 2D monolayer controls as well as iPSC- in the triculture model compared to untreated controls,
HPC only bioprinted liver constructs, Ma et al. demon- while no observed significant increase was detected in the
strated improved metabolic activity with regards to 2D monolayer and bioprinted iPSC-HPC only groups.
increased albumin and urea secretion over time in the tri- These results suggest that the triculture model provided a
culture system over 19 days in culture. Maturation state suitable environment for drug induction potential of
also increased in the triculture system as characterized by iPSC-HPCs in vitro.
a significantly higher expression of albumin (ALB), trans- The pathological mechanisms of liver fibrogenesis
thyretin (TTR), hepatocyte nuclear factor 4 alpha have also been largely studied as this condition develops
(HNF4α) along with a downregulation of the early marker over time in many types of chronic liver diseases and is a
alpha fetal protein (AFP) [108]. When treated with the consequence of an imbalance in the reparative process
hepatotoxic drug, rifampicin, CYP expression levels (i.e., following injury. As such, this results in changes in the
CYP2C9, CYP2C19, and CYP3A4) significantly increased liver tissue matrix due to an abnormal overproduction of
1432 PART | TWENTY ONE Emerging technologies

fibrillar collagen that disrupts normal liver microarchitec- were measured to evaluate their growth and invasiveness
ture, as well as the production of inflammatory cytokines over 7 days. The results demonstrated characteristic
and growth factors, which altogether deteriorates normal restricted growth and upregulation of invasive markers of
liver function [201]. Modeling liver fibrosis in vitro is these cancer cells under cirrhotic stiffness compared to
challenging to recapitulate since the process in vivo is healthy stiffness controls, which confirmed the validity of
complex and involves multiple interactions between both the tissue model. This work affirms the feasibility of
resident cells as well as recruited cells. Since macro- using 3D bioprinting to study HCC growth and invasion
phages play an important role in liver fibrogenesis, in a pathological mechanical environment.
Norona et al. aimed to further elucidate the role of Overall, these advanced engineered liver tissues high-
Kupffer cells, which are specialized macrophages located light their potential use as more reliable, sensitive, and
in the liver sinusoids and involved in homeostatic func- predictive models for improving the drug screening pro-
tionality [202], following drug-induced fibrogenesis in a cess. However, many of these 3D liver models are still in
3D bioprinted liver model [203]. Using a NovoGen early development and limited in terms of their applica-
Bioprinter platform (Organovo, San Diego, CA) a tion in high throughput arrangements accustomed in the
scaffold-free, two-compartment tissue geometry was pro- pharmaceutical industry. Moreover, to ensure the efficacy
duced comprising of hepatic stellate cells and endothelial of future 3D liver models, they will need to satisfy the
cells within the border regions while primary human following criteria: (1) demonstrate significant advantage
hepatocytes were positioned within the compartments in over conventional models such as 2D cultured hepato-
the presence or absence of Kupffer cells [203]. Over 28 cytes, (2) exhibit long-term maintenance of normal func-
days in culture, the liver tissues were treated with TGF-β1 tion and phenotype as well as drug-metabolizing enzyme
and methotrexate to induce fibrotic injury. Interestingly, activities to enable prolonged in vitro hepatotoxicity test-
levels of lactate dehydrogenase activity, which was used ing, and (3) produce measurable outcomes in response to
as a general biomarker of cytotoxic response over the the development of DILI such as changes in mitochon-
treatment period, was reduced in the liver tissue contain- drial function and the formation of reactive oxygen spe-
ing Kupffer cells for both TGF-β1 and methotrexate cies [204].
[203]. Furthermore, the release of miR-122, which was
used as a specific hepatocyte injury biomarker, verified
that the early onset of injury in response to TGF-β1 was Vascular tissue models
delayed in the Kupffer cell containing model compared to Adequate vascularization is critical for long-term survival
the standard model without Kupffer cells [203]. On the by facilitating the delivery of oxygen and nutrients to
other hand, miR-122 levels were also delayed in the cells within thick constructs to overcome diffusion limita-
methotrexate treated samples during initial exposure; tions and is an essential element for realizing the potential
however, the overall levels increased relative to the stan- of constructing full-scale tissues and organs [205]. With
dard model across the treatment period. While both drugs the development of complex tissue model systems, the
resulted in similar biochemical responses, gene expression incorporation of functional vascular networks is important
profiling further elucidated that Kupffer cells played an as they play a key role in several biological processes
important role on baseline tissue function and the varied including wound repair, progression of cancer and tumor
responses these cells have on different drug compounds pathologies, developmental angiogenesis, as well as aid-
[203]. Specifically, it was hypothesized that TGF-β1 acts ing in graft implantation and integration into the host tis-
after hepatocyte injury occurs during the intermediate sues [206 209]. To date, methods to produce vasculature
stages of the response while methotrexate acts upstream in vitro can be divided into several strategies: (1) via the
by instigating hepatocyte injury prior to downstream incorporation of proangiogenic growth factors (e.g.,
fibrotic events [203]. In a different examination, a plat- VEGF, basic fibroblast growth factor, and epidermal
form for studying the effects of pathologically relevant growth factor) to stimulate de novo vessel formation, (2)
3D matrix stiffness on hepatocellular carcinoma (HCC) use of sacrificial materials for the construction of hollow
progression and invasion was developed by Ma et al. microchannels, (3) use of microfluidic systems, or (4) pat-
(Fig. 75.5) [103]. Here, a 3D liver construct composed of terning of encapsulated endothelial cell populations [210].
decellularized porcine liver ECM and GelMA was devel- However, engineering functional blood vessels at multiple
oped such that the mechanical properties of the engi- length scales within a construct ranging from small
neered liver tissue were precisely tuned to recapitulate capillaries to large arteries has been a challenge in the
varying stages of fibrotic liver conditions ranging from field due to limitations in biofabrication technologies.
physiologically relevant softer than normal (0.5 kPa), nor- Recently, the flexibility and resolution offered by
mal (5 kPa), and cirrhotic (15 kPa) moduli [103]. Next, advanced 3D bioprinting platforms have enabled the pro-
encapsulated HepG2 cells within the bioprinted scaffold duction of vascular networks possessing complex
 Biofabricated three-dimensional tissue models Chapter | 75 1433

FIGURE 75.6 Vascular tissue models. (A) Image of an extrusion printed vascularized tissue contained in a perfusion chamber. Scale bar 5 5 mm.
(B) Fluorescent images of direct DLP-based 3D bioprinted prevascularized tissues showing stained HUVECs (green, CD31) and supportive mesenchy-
mal stem cells (purple, alpha-smooth muscle actin) of varying widths (i iii) ranging 50 250 μm. Scale bar 5 100 μm. (C and D) Schematic of 3D
printed health and stenotic vascular models, as well as (E) an image of the PDMS chip containing the vascular microchannels. (F) Confocal images of
HUVECs stained with F-actin (green) and nuclei (blue) lining the interior microchannel walls. Scale bar 5 200 μm (top and side view) and 50 μm
(confluent interior). (G) Microfluidic device with healthy geometry before (top left) and after 15 minutes of blood perfusion (top right) demonstrates
no evidence of platelet aggregation. A microfluidic device with stenotic geometry after 1 minute (bottom left) and 2.5 minutes (bottom right) of blood
perfusion demonstrates thrombosis at the apex of the stenosis. Scale bar 5 200 μm. (H) Schematic of an extrusion 3D bioprinting process to form a
vascularized hydrogel thrombosis model as shown in (I). (J) Images of endothelialization within bifurcated microchannels showing (K) a confluent
monolayer of CD31 (green) and nuclei (blue) stained HUVECs along the walls. 3D, Three-dimensional; DLP, digital light processing; HUVECs,
human umbilical vein endothelial cells; PDMS, polydimethylsiloxane. Reproduced with permission from (A) National Academy of Sciences, r2016
[130]; (B) Elsevier (2017) [107]; (C F) The Royal Society of Chemistry (RSC) (2017) [218]; and (H K) PubMed Central (PMC) (2016) [219].

geometries over traditional approaches by way of sacrifi- density found in native vessels [211]. This was demon-
cial, spheroid-based, core shell extrusion, and light- strated by depositing spheroids of Chinese hamster ovary
based printing techniques. For instance, using sacrificial cells and human skin fibroblasts (HSFs) onto supporting
biomaterials that can be evacuated to produce intricate agarose rod molds. After 5 7 days in culture to allow for
microchannel networks and subsequently populated with cell fusion, the spheroids assembled into hollow vessels
endothelial cells is a simple and effective technique to measuring 900 μm in diameter with a wall thickness of
create functional vessels. Using this approach, Kolesky 300 μm [211]. However, issues regarding spheroid scal-
et al. employed an extrusion-based bioprinter and formed ability, long fusion durations, and homogeneity of the
intricate networks with a fugitive ink composed of final vessel upon spheroid fusion limit the use of this
Pluronic F-127, which was subsequently cast with a soft approach for building larger structures. As such, this
gelatin fibrinogen matrix containing human neonatal der- group demonstrated a more efficient technique via the
mal fibroblasts (Fig. 75.6) [130]. Following removal of extrusion of multicellular cylinders composed of human
the Pluronic F-127, the resulting open microchannels, umbilical vein smooth muscle cells and HSFs onto aga-
approximately 400 500 μm in diameter, were seeded rose rod molds, which enabled fusion to complete within
with HUVECs into a confluent monolayer and it was 2 4 days resulting in branched tube structures with
demonstrated that the perfused tissue chip was capable of double-layered walls [211]. While promising, spatial reso-
supporting a tissue thickness of 1 cm for at least 6 weeks lution remains a challenge using this approach since it is
in culture [130]. In another approach, Norotte et al. uti- restricted by the spheroid size or micropipette tip dia-
lized a scaffold-free bioprinting method that relies on the meters (i.e., 300 or 500 μm). Furthermore, a direct method
self-assembly of patterned cell spheroids to maximize to construct vasculature that circumvents the multiple
direct cell cell interactions and achieve high cellular steps and time involved in sacrificial and spheroid-based
1434 PART | TWENTY ONE Emerging technologies

printing strategies would enable a more rapid, stream- vascular pathogenesis and disorders. In particular, throm-
lined, and scalable fabrication approach. To address this bosis has been linked as a major contributor and cause for
challenge, Jia et al. developed a technique to directly print global disability and mortality due to its association as an
perfusable vascular constructs using an extrusion bioprin- underlying factor for cardiovascular disorders [213,214].
ter fitted with a trilayered coaxial nozzle to deposit a For instance, venous thromboembolism, in the form of
blend bioink composed of sodium alginate, GelMA, and deep vein thrombosis or pulmonary embolism, is a preva-
4-arm PEG-tetra-acrylate (PEGTA) [212]. Ionic cross- lent medical condition that results in a blood clot occur-
linking of the alginate component was first induced via ring in the vein and affects 1 out of 1000 adults annually
the simultaneous delivery of calcium ions in the core in the United States [215]. This condition derives from
channel and ambient spray in the outer ring of the print patient risk factors such as age, obesity, surgery, genetic
nozzle to ensure temporary mechanical stability during factors, trauma, cancer, and immobility that can lead to
the extrusion process. Afterward, the GelMA and PEGTA chronic pain, tissue swelling, formation of venous ulcers,
components were UV photopolymerized to permanently and in severe cases lung collapse resulting in heart failure
stabilize the final construct. By varying the nozzle inter- [216]. Similarly, arterial thrombosis is also a leading
nal and external needle sizes, different combinations of cause of death in the United States due to occlusions
diameters (i.e., outer 5 500 1500 μm, caused by clots or arterial plaque that when left untreated,
inner 5 400 1000 μm) and wall thicknesses (i.e., can manifest into a host of secondary health complica-
60 280 μm) were achieved using this technique in addi- tions including stroke, heart attack, and critical limb
tion to being able to readily form various designs and ischemia [217]. Due to its prevalence, there is a need for
geometries of perfusable networks. Furthermore, it was models of thrombosis that accurately recapitulates the
demonstrated that coencapsulated HUVECs and human vessel geometries, local flow patterns, as well as underly-
MSCs within the blend bioink provided an optimal micro- ing cellular interactions to develop better future treatment
environment for the cell viability, proliferation, and options. Recently, Costa et al. used a stereolithography-
spreading to form perfusable constructs resembling native based 3D printer to create a replica of a miniaturized cor-
vasculature after 21 days in culture [212]. However, the onary artery model based on CT angiography (CTA)
ability to form multiscale vasculature structures in a con- (Fig. 75.6) [218]. Using this approach, they were able to
tinuous single step using this approach remains difficult produce polydimethylsiloxane microfluidic chips of
as it will involve physically interchanging different nozzle healthy and stenotic coronary arteries that were seeded
sizes throughout the printing process. One strategy is to with a monolayer of HUVECs and perfused with human
utilize light-based printing that does not involve physical whole blood containing fluorescently labeled platelets
contact during fabrication and instead works by projecting [218]. Under physiologically relevant arterial shear rates,
a series of digital patterns onto a photopolymerizable res- it was demonstrated that thrombosis occurred within the
ervoir to quickly form vasculature structures in a direct stenotic geometries after 2.5 minutes of perfusion but not
manner. Zhu et al. demonstrated this technique by rapidly in the healthy controls as expected. Computational fluid
constructing complex hierarchal branched networks by dynamics further confirmed the validity of their scaled
using a microscale continuous optical bioprinting platform microfluidic model geometries to closely correlated to the
to encapsulate HUVECs and supportive C3H/10T1/2 cells original CTA scale geometries under the selected flow
within a GelMA and GM-HA blend prepolymer velocity profiles and shear rate distributions used in this
(Fig. 75.6) [107]. Meanwhile, the areas surrounding the study for both the healthy and stenotic conditions [218].
printed vasculature networks were printed with a GelMA This work highlights several advancements regarding the
bioink embedded with HepG2 to form a heterogeneous ability to recapitulate real vessel architectures possessing
tissue construct. Intricate vessel structures with micro- smooth walls to elucidate factors affecting hemodynam-
channel widths ranging between 50 and 250 μm were ics. In particular, changes in fluidic velocity gradients,
readily fabricated and over time the endothelial cell popu- increased shear rates, platelet formation in the recircula-
lation remodeled the hydrogel to form lumen-like struc- tion zone, and activation of the intrinsic coagulation path-
tures after 1 week in culture [107]. In vivo implantation way within stenotic vessel geometries were probable root
into a subcutaneous mouse model of the prevascularized causes of thrombosis in their model [218]. In another
tissues also revealed integration and anastomosis to the study of thrombosis, Zhang et al. examined the pathology
host circulation after 2 weeks compared to nonprevascu- of fibrosis within vessels by using an extrusion-based 3D
larized controls [107]. printer to create hollow bifurcated microchannels within a
With the promising application of 3D bioprinting tech- GelMA hydrogel via the removal of sacrificial Pluronic
nologies to engineer vascularized tissues, several studies F-127 (Fig. 75.6) [219]. To evaluate the migration of
have also utilized the flexibility of 3D bioprinters to cre- fibroblasts in exacerbating the formation of a fibrotic clot
ate models to better understand the mechanisms of due to a damaged endothelium three groups were tested:
 Biofabricated three-dimensional tissue models Chapter | 75 1435

(1) microchannels lined with HUVECs and no fibroblasts the protocols employed to control differentiation and
in the GelMA matrix as a negative control, (2) HUVECs maturation toward the desired lineage in vitro. Specialized
lined microchannels with fibroblasts in the GelMA matrix cell types differentiated from iPSCs are not equivalent in
to serve as a normal vessel control, and (3) fibroblasts in their maturation state as primary cells and are considered
the GelMA matrix with no HUVECs lining the micro- immature with resemblance more similar to fetal cells of
channels to simulate the damaged endothelium [219]. the native tissue. As such, the successful use of stem
Thrombosis was induced by administering CaCl2 into cell derived cells hinges on new methods to facilitate
human whole blood which was then perfused into the their differentiation pathway to achieve mature adult phe-
microchannels for each group and allowed to culture for 2 notypes and functionalities in order to more accurately
weeks. Interestingly, as expected there was no visible extrapolate data for disease processes observed in adult
staining of collagen I present within the clot in both the tissues and organs. In another aspect the multicellular
negative control and normal vessel control due to the population inherent in native tissues comprised of both
absence of fibroblasts and barrier formed by the endothe- parenchymal and nonparenchymal cell types is critical in
lium, respectively. In contrast, the group simulating the the design of physiologically relevant tissue models.
damaged endothelium had significant invasion of fibro- Simplistic 3D tissue models contain a homogenous popu-
blasts found in the regions of the thrombus along with lation comprised of a single cell type representing the
deposition of collagen I secreted by the cells, which is a functional unit of the tissue or organ. However, growing
representative of the in vivo setting [219]. Overall, these evidence in recent literature has shown that supportive
realistic models hold great potential to serve as physiolog- nonparenchymal cells play a critical role in overall tissue
ically relevant platforms for studying the mechanisms of function [108,203]. As such, future designs will need to
thrombus formation as well as other vascular disorders by continue to recapitulate tissue or organ cellular heteroge-
being able to closely mimic the native vascular neity to build toward improved engineered tissue models
microenvironment. that possess greater relevancy to native behavior that will
increase their efficacy. Finally, reliable and reproducible
manufacturing methods will be an essential factor for the
Conclusion and future directions scalability of 3D bioprinted tissue models in high
Over the years, 3D bioprinting has surfaced as a flexible throughput applications. This is necessary as future tissue
biofabrication tool that enables the creation of complex tis- models will need to be produced in a manner that is able
sue model systems to better recapitulate the native micro- to interface directly with existing commercial platforms
environment. In particular, the successful outcome of accustomed in industry, such as for drug screening and
engineered tissue models relies on several key factors diagnostics, without compromising performance, consis-
including mechanical properties (i.e., stiff/soft and elastic/ tency, and predictive value. Considering these factors is
brittle), biochemical composition (ECM components, an important step toward achieving the full potential of
growth factors, etc.), and complexity of the final tissue 3D tissue models for use in the development of new ther-
(cell source, multiple cell types, shape, and organization). apies and elucidating pathological mechanisms.
Taken together, careful consideration of these factors can Nonetheless, technological advancements in 3D bioprint-
influence which method of 3D bioprinting is best suited for ing continue to play a critical role in defining new direc-
a specific tissue application. To date, a growing variety of tions and possibilities for realizing the production of
engineered tissues have been successfully produced using a tissue or organ substitutes in both research and clinical
range of different 3D bioprinting modalities to serve as applications.
models for improving diagnostic and therapeutic outcomes
including nerve, cancer, heart, liver, and vascularized tis-
sues as highlighted above. While these advanced tissue Acknowledgments
models represent a step closer to physiologically relevant This work was supported in part by the National Institutes of Health
tissue platforms, many challenges remain regarding their (R21AR074763, R01EB021857) and the National Science
maturation state, multicellular complexity, and scalability Foundation (CMMI1644967). Scholarship funding for Dr. Claire Yu
for high throughput manufacturing. was provided by the Natural Sciences and Engineering Research
With the emergence of complex 3D tissue models, the Council (NSERC) Postdoctoral Fellowship Scholarship of Canada.
trend toward the selection of stem cell derived cells, par-
ticularly iPSCs, has gained popularity in terms of their
practicality in research and translational advantages. References
While promising, it is important to recognize that the rele- [1] Woodfield T, Lim K, Morouço P, Levato R, Malda J, Melchels F.
vancy and performance of tissue models composed of Biofabrication in tissue engineering. Compr Biomater II
these cell types are highly dependent on the efficacy of 2017;236 66.
1436 PART | TWENTY ONE Emerging technologies

[2] Duval K, Grover H, Han L-H, Mou Y, Pegoraro AF, Fredberg J, [21] You S, Li J, Zhu W, Yu C, Mei D, Chen S. Nanoscale 3D printing
et al. Modeling physiological events in 2D vs. 3D cell culture. of hydrogels for cellular tissue engineering. J Mater Chem, B
Physiology 2017;32:266 77. 2018;6:2187.
[3] Edmondson R, Broglie JJ, Adcock AF, Yang L. Three- [22] Yu C, Zhu W, Sun B, Mei D, Gou M, Chen S. Modulating physi-
dimensional cell culture systems and their applications in drug dis- cal, chemical, and biological properties in 3D printing for tissue
covery and cell-based biosensors. Assay Drug Dev Technol engineering applications. Crit Appl Phys Rev 2018;5:41107.
2014;12:207 18. [23] Chen S, Lawrence N, Zhu W, Victorine G, Hwang HH. 3D-
[4] Baharvand H, Hashemi SM, Ashtiani SK, Farrokhi A. printing of functional biomedical microdevices via light- and
Differentiation of human embryonic stem cells into hepatocytes in extrusion-based approaches. Small Methods 2017;2:1700277.
2D and 3D culture systems in vitro. Int J Dev Biol 2006;50:645 52. [24] Koch L, Gruene M, Unger C, Chichkov B. Laser assisted cell
[5] Benya PD, Shaffer JD. Dedifferentiated chondrocytes reexpress printing. Curr Pharm Biotechnol 2013;14:91 7.
the differentiated collagen phenotype when cultured in agarose [25] Jakab K, Damon B, Neagu A, Kachurin A, Forgacs G. Three-
gels. Cell 1982;30:215 24. dimensional tissue constructs built by bioprinting. Biorheology
[6] Nelson CM, Bissell MJ. Modeling dynamic reciprocity: engineer- 2006;43:509 13.
ing three-dimensional culture models of breast architecture, func- [26] Nair K, Gandhi M, Khalil S, Yan KC, Marcolongo M, Barbee K,
tion, and neoplastic transformation. Semin Cancer Biol et al. Characterization of cell viability during bioprinting pro-
2005;15:342 52. cesses. Biotechnol J 2009;4:1168 77.
[7] Dunn JC, Yarmush ML, Koebe HG, Tompkins RG. Hepatocyte [27] Smith CM, Stone AL, Parkhill RL, Stewart RL, Simpkins MW,
function and extracellular matrix geometry: long-term culture in a Kachurin AM, et al. Three-dimensional bioassembly tool for gen-
sandwich configuration. FASEB J 1989;3:174 7. erating viable tissue-engineered constructs. Tissue Eng
[8] Yu Y, Moncal KK, Li J, Peng W, Rivero I, Martin JA, et al. Three- 2004;10:1566 76.
dimensional bioprinting using self-assembling scalable scaffold-free [28] Demirci U, Montesano G. Single cell epitaxy by acoustic picolitre
“tissue strands” as a new bioink. Sci Rep 2016;6:28714. droplets. Lab Chip 2007;7:1139 45.
[9] Yu Y, Ozbolat IT. Tissue strands as “bioink” for scale-up organ [29] Guillotin B, Souquet A, Catros S, Duocastella M, Pippenger B,
printing. In: 2014 36th annual international conference of the Bellance S, et al. Laser assisted bioprinting of engineered tissue
IEEE Engineering in Medicine and Biology Society (EMBC). with high cell density and microscale organization. Biomaterials
IEEE; 2014. p. 1428 31. 2010;31:7250 6.
[10] Ovsianikov A, Khademhosseini A, Mironov V. The synergy of [30] Wang Z, Abdulla R, Parker B, Samanipour R, Ghosh S, Kim K. A
scaffold-based and scaffold-free tissue engineering strategies. simple and high-resolution stereolithography-based 3D bioprinting
Trends Biotechnol 2018;36:348 57. system using visible light crosslinkable bioinks. Biofabrication
[11] Silva KR, Rezende RA, Pereira FDAS, Gruber P, Stuart MP, 2015;7:45009.
Ovsianikov A, et al. Delivery of human adipose stem cells spher- [31] Bajaj P, Schweller RM, Khademhosseini A, West JL, Bashir R.
oids into lockyballs. PLoS One 2016;11:e0166073. 3D biofabrication strategies for tissue engineering and regenera-
[12] Kook YM, Jeong Y, Lee K, Koh WG. Design of biomimetic cel- tive medicine. Annu Rev Biomed Eng 2014;16:247 76.
lular scaffolds for co-culture system and their application. J Tissue [32] Khalil S, Sun W. Biopolymer deposition for freeform fabrication
Eng 2016;8 204173141772464. of hydrogel tissue constructs. Mater Sci Eng C 2007;27:469 78.
[13] Liu Y, Wei J, Lu J, Lei D, Yan S, Li X. Micropatterned coculture [33] Chia HN, Wu BM. Recent advances in 3D printing of biomater-
of hepatocytes on electrospun fibers as a potential in vitro model ials. J Biol Eng 2015;9:4.
for predictive drug metabolism. Mater Sci Eng C [34] Guillemot F, Souquet A, Catros S, Guillotin B, Lopez J, Faucon
2016;63:475 84. M, et al. High-throughput laser printing of cells and biomaterials
[14] Paulsen SJ, Miller JS. Tissue vascularization through 3D printing: for tissue engineering. Acta Biomater 2010;6:2494 500.
will technology bring us flow? Dev Dyn 2015;244:629 40. [35] Peltola SM, Melchels FPW, Grijpma DW, Kellomäki M. A review
[15] Murphy SV, Atala A. 3D bioprinting of tissues and organs. Nat of rapid prototyping techniques for tissue engineering purposes.
Biotechnol 2014;32:773 85. Ann Med 2008;40:268 80.
[16] Li J, Chen M, Fan X, Zhou H. Recent advances in bioprinting [36] Hospodiuk M, Dey M, Sosnoski D, Ozbolat IT. The bioink: a
techniques: approaches, applications and future prospects. J Transl comprehensive review on bioprintable materials. Biotechnol Adv
Med 2016;14:271. 2017;35:217 39.
[17] Zhu W, Ma X, Gou M, Mei D, Zhang K, Chen S. 3D printing of [37] Kim JD, Choi JS, Kim BS, Chan Choi Y, Cho YW. Piezoelectric
functional biomaterials for tissue engineering. Curr Opin inkjet printing of polymers: stem cell patterning on polymer sub-
Biotechnol 2016;40:103 12. strates. Polymer (Guildf) 2010;51:2147 54.
[18] Jana S, Lerman A. Bioprinting a cardiac valve. Biotechnol Adv [38] Chang CC, Boland ED, Williams SK, Hoying JB. Direct-write
2015;33:1503 21. bioprinting three-dimensional biohybrid systems for future regen-
[19] Cui X, Boland T, D’Lima DD, Lotz MK. Thermal inkjet printing erative therapies. J Biomed Mater Res, B: Appl Biomater
in tissue engineering and regenerative medicine. Recent Pat Drug 2011;98:160 70.
Deliv Formul 2012;6:149 55. [39] Guillotin B, Guillemot F. Cell patterning technologies for organo-
[20] Benam KH, Dauth S, Hassell B, Herland A, Jain A, Jang K-J, typic tissue fabrication. Trends Biotechnol 2010;29:183 90.
et al. Engineered in vitro disease models. Annu Rev Pathol Mech [40] Xu T, Jin J, Gregory C, Hickman JJ, Boland T. Inkjet printing of
Dis 2015;10:195 262. viable mammalian cells. Biomaterials 2005;26:93 9.
 Biofabricated three-dimensional tissue models Chapter | 75 1437

[41] Xu T, Gregory CA, Molnar P, Cui X, Jalota S, Bhaduri SB, et al. [58] Williams DF. On the nature of biomaterials. Biomaterials
Viability and electrophysiology of neural cell structures generated 2009;30:5897 909.
by the inkjet printing method. Biomaterials 2006;27:3580 8. [59] Ahmed EM. Hydrogel: preparation, characterization, and applica-
[42] Chang R, Nam J, Sun W. Effects of dispensing pressure and noz- tions: a review. J Adv Res 2015;6:105 21.
zle diameter on cell survival from solid freeform fabrica- [60] Skardal A, Atala A. Biomaterials for integration with 3-D bio-
tion based direct cell writing. Tissue Eng, A 2008;14:41 8. printing. Ann Biomed Eng 2015;43:730 46.
[43] Hopp B, Smausz T, Kresz N, Barna N, Bor Z, Kolozsvári L, et al. [61] Kumar P, Parak A, Choonara YE, Pillay V, du Toit LC, Pradeep
Survival and proliferative ability of various living cell types after P. Functionalizing bioinks for 3D bioprinting applications. Drug
laser-induced forward transfer. Tissue Eng 2005;11:1817 23. Discov Today 2018;198 205.
[44] Koch L, Kuhn S, Sorg H, Gruene M, Schlie S, Gaebel R, et al. [62] Choudhury D, Tun HW, Wang T, Naing MW. Organ-derived
Laser printing of skin cells and human stem cells. Tissue Eng, C: decellularized extracellular matrix: a game changer for bioink
Methods 2010;16:847 54. manufacturing? Trends Biotechnol 2018;36:787 805.
[45] Murphy SV, Skardal A, Atala A. Evaluation of hydrogels for bio- [63] Brinkman WT, Nagapudi K, Thomas BS, Chaikof EL. Photo-
printing applications. J Biomed Mater Res, A 2013;101 cross-linking of type I collagen gels in the presence of smooth
A:272 84. muscle cells: mechanical properties, cell viability, and function.
[46] Smith CM, Christian JJ, Warren WL, Williams SK. Biomacromolecules 2003;4:890 5.
Characterizing environmental factors that impact the viability of [64] Ahmed I, Drzewiecki KE, Malavade JN, Lowe CJ, Shreiber DI. A
tissue-engineered constructs fabricated by a direct-write bioassem- thermoreversible, photocrosslinkable collagen bio-ink for free-
bly tool. Tissue Eng 2007;13:373 83. form fabrication of scaffolds for regenerative medicine.
[47] Michael S, Sorg H, Peck CT, Koch L, Deiwick A, Chichkov B, Technology 2017;05:185 95.
et al. Tissue engineered skin substitutes created by laser-assisted [65] Gaudet ID, Shreiber DI. Characterization of methacrylated Type-I
bioprinting form skin-like structures in the dorsal skin fold cham- collagen as a dynamic, photoactive hydrogel. Biointerphases
ber in mice. PLoS One 2013;8:e57741. 2012;7:1 9.
[48] Koch L, Deiwick A, Schlie S, Michael S, Gruene M, Coger V, [66] Rouillard AD, Berglund CM, Lee JY, Polacheck WJ, Tsui Y,
et al. Skin tissue generation by laser cell printing. Biotechnol Bonassar LJ, et al. Methods for photocrosslinking alginate hydro-
Bioeng 2012;109:1855 63. gel scaffolds with high cell viability. Tissue Eng, C: Methods
[49] Chan V, Collens MB, Jeong JH, Park K, Kong H, Bashir R. 2011;17:173 9.
Directed cell growth and alignment on protein-patterned 3D [67] Khunmanee S, Jeong Y, Park H. Crosslinking method of
hydrogels with stereolithography. Virtual Phys Prototyp hyaluronic-based hydrogel for biomedical applications. J Tissue
2012;7:219 28. Eng 2017;8 2041731417726464.
[50] Kwon IK, Matsuda T. Photo-polymerized microarchitectural con- [68] Poldervaart MT, Goversen B, De Ruijter M, Abbadessa A,
structs prepared by microstereolithography (μSL) using liquid Melchels FPW, Öner FC, et al. 3D bioprinting of methacrylated
acrylate-end-capped trimethylene carbonate-based prepolymers. hyaluronic acid (MeHA) hydrogel with intrinsic osteogenicity.
Biomaterials 2005;26:1675 84. PLoS One 2017;12:e0177628.
[51] Nakamura M, Kobayashi A, Takagi F, Watanabe A, Hiruma Y, [69] Stichler S, Böck T, Paxton N, Bertlein S, Levato R, Schill V,
Ohuchi K, et al. Biocompatible inkjet printing technique for et al. Double printing of hyaluronic acid/poly(glycidol) hybrid
designed seeding of individual living cells. Tissue Eng hydrogels with poly(E-caprolactone) for MSC chondrogenesis.
2005;11:1658 66. Biofabrication 2017;9:044108.
[52] Ma X, Liu J, Zhu W, Tang M, Lawrence N, Yu C, et al. 3D bio- [70] Yanez M, Rincon J, Dones A, De Maria C, Gonzales R, Boland T.
printing of functional tissue models for personalized drug screen- In vivo assessment of printed microvasculature in a bilayer skin graft
ing and in vitro disease modeling. Adv Drug Deliv Rev to treat full-thickness wounds. Tissue Eng, A 2015;21:224 33.
2018;132:235 51. [71] Ferreira AM, Gentile P, Chiono V, Ciardelli G. Collagen for bone
[53] Duocastella M, Fernández-Pradas JM, Morenza JL, Zafra D, Serra tissue regeneration. Acta Biomater 2012;8:3191 200.
P. Novel laser printing technique for miniaturized biosensors prep- [72] Lee VK, Lanzi AM, Ngo H, Yoo SS, Vincent PA, Dai G.
aration. Sens Actuators, B: Chem 2010;145:596 600. Generation of multi-scale vascular network system within 3D
[54] Seerden KAM, Reis N, Evans JRG, Grant PS, Halloran JW, hydrogel using 3D bio-printing technology. Cell Mol Bioeng
Derby B. Ink-jet printing of wax-based alumina suspensions. J 2014;7:460 72.
Am Ceram Soc 2001;84:2514 20. [73] Duarte Campos DF, Blaeser A, Korsten A, Neuss S, Jäkel J, Vogt
[55] Saunders RE, Gough JE, Derby B. Delivery of human fibroblast M, et al. The stiffness and structure of three-dimensional printed
cells by piezoelectric drop-on-demand inkjet printing. hydrogels direct the differentiation of mesenchymal stromal cells
Biomaterials 2008;29:193 203. toward adipogenic and osteogenic lineages. Tissue Eng, A
[56] Andò B, Marletta V. An all-inkjet printed bending actuator with 2015;21:740 56.
embedded sensing feature and an electromagnetic driving mecha- [74] Chang CC, Boland ED, Williams SK, Hoying JB. Direct-write
nism. Actuators 2016;5:21. bioprinting three-dimensional biohybrid systems for future regen-
[57] Kollamaram G, Hopkins SC, Glowacki BA, Croker DM, Walker erative therapies. J Biomed Mater Res, B: Appl Biomater 2011;98
GM. Inkjet printing of paracetamol and indomethacin using elec- B:160 70.
tromagnetic technology: rheological compatibility and polymor- [75] Axpe E, Oyen ML. Applications of alginate-based bioinks in 3D
phic selectivity. Eur J Pharm Sci 2018;115:248 57. bioprinting. Int J Mol Sci 2016;17:1976.
1438 PART | TWENTY ONE Emerging technologies

[76] Lee KY, Mooney DJ. Alginate: properties and biomedical appli- [96] Forget A, Blaeser A, Miessmer F, Köpf M, Campos DFD,
cations. Prog Polym Sci 2012;37:106 26. Voelcker NH, et al. Mechanically tunable bioink for 3D bioprint-
[77] Zhang Y, Yu Y, Akkouch A, Dababneh A, Dolati F, Ozbolat IT. ing of human cells. Adv Healthc Mater 2017;6:1700255.
In vitro study of directly bioprinted perfusable vasculature con- [97] Duarte Campos DF, Blaeser A, Weber M, Jäkel J, Neuss S,
duits. Biomater Sci 2015;3:134 43. Jahnen-Dechent W, et al. Three-dimensional printing of stem
[78] de Vos P, Faas MM, Strand B, Calafiore R. Alginate-based cell-laden hydrogels submerged in a hydrophobic high-density
microcapsules for immunoisolation of pancreatic islets. fluid. Biofabrication 2013;5:2202 11.
Biomaterials 2006;27:5603 17. [98] Duarte Campos DF, Blaeser A, Weber M, Jäkel J, Neuss S,
[79] Ozbolat IT, Hospodiuk M. Current advances and future perspec- Jahnen-Dechent W, et al. Three-dimensional printing of stem
tives in extrusion-based bioprinting. Biomaterials 2016;76:321 43. cell-laden hydrogels submerged in a hydrophobic high-density
[80] Purcell EK, Singh A, Kipke DR. Alginate composition effects on fluid. Biofabrication 2013;5:15003 14.
a neural stem cell seeded scaffold. Tissue Eng, C: Methods [99] Cui X, Boland T. Human microvasculature fabrication using
2009;15:541 50. thermal inkjet printing technology. Biomaterials
[81] Hill E, Boontheekul T, Mooney DJ. Designing scaffolds to 2009;30:6221 7.
enhance transplanted myoblast survival and migration. Tissue [100] Sahni A, Francis CW. Vascular endothelial growth factor binds
Eng 2006;12:1295 304. to fibrinogen and fibrin and stimulates endothelial cell prolifera-
[82] Alsberg E, Anderson KW, Albeiruti A, Rowley JA, Mooney DJ. tion. Blood 2000;96:3772 8.
Engineering growing tissues. Proc Natl Acad Sci USA [101] England S, Rajaram A, Schreyer DJ, Chen X. Bioprinted fibrin-
2002;99:12025 30. factor XIII-hyaluronate hydrogel scaffolds with encapsulated
[83] Rowley JA, Madlambayan G, Mooney DJ. Alginate hydrogels as Schwann cells and their in vitro characterization for use in nerve
synthetic extracellular matrix materials. Biomaterials regeneration. Bioprinting 2017;5:1 9.
1999;20:45 53. [102] Yu C, Ma X, Zhu W, Wang P, Miller KL, Stupin J, et al.
[84] Gudapati H, Dey M, Ozbolat I. A comprehensive review on Scanningless and continuous 3D bioprinting of human tissues with
droplet-based bioprinting: past, present and future. Biomaterials decellularized extracellular matrix. Biomaterials 2018;194:1 13.
2016;102:20 42. [103] Ma X, Yu C, Wang P, Xu W, Wan X, Lai CSE, et al. Rapid 3D
[85] Yan J, Huang Y, Chrisey DB. Laser-assisted printing of alginate bioprinting of decellularized extracellular matrix with regionally
long tubes and annular constructs. Biofabrication 2013;5:015002. varied mechanical properties and biomimetic microarchitecture.
[86] Hemshekhar M, Thushara RM, Chandranayaka S, Sherman LS, Biomaterials 2018;185:310 21.
Kemparaju K, Girish KS. Emerging roles of hyaluronic acid [104] Kim J, Shim IK, Hwang DG, Lee YN, Kim M, Kim S-W, et al.
bioscaffolds in tissue engineering and regenerative medicine. Int 3D cell printing of islet-laden pancreatic tissue-derived extracel-
J Biol Macromol 2016;86:917 28. lular matrix bioink constructs for enhancing pancreatic functions.
[87] Highley CB, Prestwich GD, Burdick JA. Recent advances in hya- J Mater Chem, B 7, 2019, 1773 1781;.
luronic acid hydrogels for biomedical applications. Curr Opin [105] Toprakhisar B, Nadernezhad A, Bakirci E, Khani N, Skvortsov
Biotechnol 2016;40:35 40. GA, Koc B. Development of bioink from decellularized tendon
[88] Yoo HS, Lee EA, Yoon JJ, Park TG. Hyaluronic acid modified extracellular matrix for 3D bioprinting. Macromol Biosci
biodegradable scaffolds for cartilage tissue engineering. 2018;18:1800024.
Biomaterials 2005;26:1925 33. [106] Pati F, Jang J, Ha DH, Won Kim S, Rhie JW, Shim JH, et al.
[89] Hölzl K, Lin S, Tytgat L, Van Vlierberghe S, Gu L, Ovsianikov Printing three-dimensional tissue analogues with decellularized
A. Bioink properties before, during and after 3D bioprinting. extracellular matrix bioink. Nat Commun 2014;5:3935.
Biofabrication 2016;8:032002. [107] Zhu W, Qu X, Zhu J, Ma X, Patel S, Liu J, et al. Direct 3D bio-
[90] Burdick JA, Prestwich GD. Hyaluronic acid hydrogels for bio- printing of prevascularized tissue constructs with complex micro-
medical applications. Adv Mater 2011;23:H41 56. architecture. Biomaterials 2017;124:106 15.
[91] Fan R, Piou M, Darling E, Cormier D, Sun J, Wan J. Bio- [108] Ma X, Qu X, Zhu W, Li Y-S, Yuan S, Zhang H, et al.
printing cell-laden Matrigel-agarose constructs. J Biomater Appl Deterministically patterned biomimetic human iPSC-derived
2016;31:684 92. hepatic model via rapid 3D bioprinting. Proc Natl Acad Sci USA
[92] Poldervaart MT, Gremmels H, Van Deventer K, Fledderus JO, 2016;113:2206 11.
Öner FC, Verhaar MC, et al. Prolonged presence of VEGF pro- [109] Mandt D, Gruber P, Markovic M, Tromayer M, Rothbauer M,
motes vascularization in 3D bioprinted scaffolds with defined Krayz SRA, et al. Fabrication of placental barrier structures
architecture. J Control Release 2014;184:58 66. within a microfluidic device utilizing two-photon polymerization.
[93] Snyder J, Hamid Q, Wang C, Chang R, Emami K, Wu H, et al. Int J Bioprinting 2018;4.
Bioprinting cell-laden Matrigel for radioprotection study of liver [110] Nichol JW, Koshy ST, Bae H, Hwang CM, Yamanlar S,
by pro-drug conversion in a dual-tissue microfluidic chip. Khademhosseini A. Cell-laden microengineered gelatin methac-
Biofabrication 2011;3:34112 21. rylate hydrogels. Biomaterials 2010;31:5536 44.
[94] Lee H, Hong SH, Ahn DJ. Stable patterning of sensory agarose [111] Rutz AL, Hyland KE, Jakus AE, Burghardt WR, Shah RN. A
gels using inkjet printing. Macromol Res 2014;23:124 7. multimaterial bioink method for 3D printing tunable, cell-
[95] Forget A, Blaeser A, Miessmer F, Köpf M, Campos DFD, compatible hydrogels. Adv Mater 2015;27:1607 14.
Voelcker NH, et al. Mechanically tunable bioink for 3D bioprint- [112] Gao G, Schilling AF, Hubbell K, Yonezawa T, Truong D, Hong
ing of human cells. Adv Healthc Mater 2017;6:25011. Y, et al. Improved properties of bone and cartilage tissue from
 Biofabricated three-dimensional tissue models Chapter | 75 1439

3D inkjet-bioprinted human mesenchymal stem cells by simulta- [129] Kang HW, Lee SJ, Ko IK, Kengla C, Yoo JJ, Atala A. A 3D
neous deposition and photocrosslinking in PEG-GelMA. bioprinting system to produce human-scale tissue constructs with
Biotechnol Lett 2015;37:2349 55. structural integrity. Nat Biotechnol 2016;34:312 19.
[113] Gou M, Qu X, Zhu W, Xiang M, Yang J, Zhang K, et al. Bio- [130] Kolesky DB, Homan KA, Skylar-Scott MA, Lewis JA. Three-
inspired detoxification using 3D-printed hydrogel nanocompo- dimensional bioprinting of thick vascularized tissues. Proc Natl
sites. Nat Commun 2014;5:3774. Acad Sci USA 2016;113:3179 84.
[114] Cui X, Gao G, Yonezawa T, Dai G. Human cartilage tissue fabri- [131] Kolesky DB, Truby RL, Gladman AS, Busbee TA, Homan KA,
cation using three-dimensional inkjet printing technology. J Vis Lewis JA. 3D bioprinting of vascularized, heterogeneous cell-
Exp 2014;88:51294. laden tissue constructs. Adv Mater 2014;26:3124 30.
[115] Bryant SJ, Anseth KS. Hydrogel properties influence ECM pro- [132] Liu F, Vyas C, Poologasundarampillai G, Pape I, Hinduja S,
duction by chondrocytes photoencapsulated in poly(ethylene gly- Mirihanage W, et al. Structural evolution of PCL during melt
col) hydrogels. J Biomed Mater Res 2002;59:63 72. extrusion 3D printing. Macromol Mater Eng 2018;303:1700494.
[116] Ma X, Dewan S, Liu J, Tang M, Miller KL, Yu C, et al. 3D [133] Zhang K, Fu Q, Yoo J, Chen X, Chandra P, Mo X, et al. 3D bio-
printed micro-scale force gauge arrays to improve human cardiac printing of urethra with PCL/PLCL blend and dual autologous
tissue maturation and enable high throughput drug testing. Acta cells in fibrin hydrogel: an in vitro evaluation of biomimetic
Biomater 2019;95:319 27. mechanical property and cell growth environment. Acta
[117] Zhu W, Tringale KR, Woller SA, You S, Johnson S, Shen H, Biomater 2017;50:154 64.
et al. Rapid continuous 3D printing of customizable peripheral [134] Kundu J, Shim JH, Jang J, Kim SW, Cho DW. An additive
nerve guidance conduits. Mater Today 2018;21:951 9. manufacturing-based PCL-alginate-chondrocyte bioprinted scaf-
[118] Koffler J, Zhu W, Qu X, Platoshyn O, Dulin JN, Brock J, et al. fold for cartilage tissue engineering. J Tissue Eng Regen Med
Biomimetic 3D-printed scaffolds for spinal cord injury repair. 2015;9:1286 97.
Nat Med 2019;25:263 9. [135] Park JY, Choi JC, Shim JH, Lee JS, Park H, Kim SW, et al.
[119] Xue D, Wang Y, Zhang J, Mei D, Wang Y, Chen S. A comparative study on collagen type i and hyaluronic acid
Projection-based 3D printing of cell patterning scaffolds with dependent cell behavior for osteochondral tissue bioprinting.
multiscale channels. ACS Appl Mater Interfaces Biofabrication 2014;6:035004.
2018;10:19428 35. [136] Xu T, Binder KW, Albanna MZ, Dice D, Zhao W, Yoo JJ, et al.
[120] Liu J, Hwang HH, Wang P, Whang G, Chen S. Direct 3D- Hybrid printing of mechanically and biologically improved con-
printing of cell-laden constructs in microfluidic architectures. structs for cartilage tissue engineering applications.
Lab Chip 2016;16:1430 8. Biofabrication 2013;5:015001.
[121] Huang TQ, Qu X, Liu J, Chen S. 3D printing of biomimetic [137] Al-Sabah A, Jessop ZM, Whitaker IS, Thornton C. Cell prepara-
microstructures for cancer cell migration. Biomed Microdevices tion for 3D bioprinting. 3D bioprinting for reconstructive sur-
2014;16:127 32. gery, Elsevier. 2018. p. 75 88.
[122] López-Barrón CR, Chen R, Wagner NJ, Beltramo PJ. Self- [138] Caddeo S, Boffito M, Sartori S. Tissue engineering approaches
assembly of Pluronic F127 diacrylate in ethylammonium nitrate: in the design of healthy and pathological in vitro tissue models.
structure, rheology, and ionic conductivity before and after Front Bioeng Biotechnol 2017;5:40.
photo-cross-linking. Macromolecules 2016;49:5179 89. [139] Nguyen DG, Funk J, Robbins JB, Crogan-Grundy C, Presnell
[123] Müller M, Becher J, Schnabelrauch M, Zenobi-Wong M. SC, Singer T, et al. Bioprinted 3D primary liver tissues allow
Nanostructured Pluronic hydrogels as bioinks for 3D bioprinting. assessment of organ-level response to clinical drug induced tox-
Biofabrication 2015;7:035006. icity in vitro. PLoS One 2016;11:e0158674.
[124] Gioffredi E, Boffito M, Calzone S, Giannitelli SM, Rainer A, [140] Taniguchi D, Matsumoto K, Tsuchiya T, MacHino R, Takeoka
Trombetta M, et al. Pluronic F127 hydrogel characterization and Y, Elgalad A, et al. Scaffold-free trachea regeneration by tissue
biofabrication in cellularized constructs for tissue engineering engineering with bio-3D printing. Interact Cardiovasc Thorac
applications. Procedia CIRP 2016;49:125 32. Surg 2018;26:745 52.
[125] Jhaveri SJ, McMullen JD, Sijbesma R, Tan L-S, Zipfel W, Ober [141] Rimann M, Bono E, Annaheim H, Bleisch M, Graf-Hausner U.
CK. Direct three-dimensional microfabrication of hydrogels via Standardized 3D bioprinting of soft tissue models with human
two-photon lithography in aqueous solution. Chem Mater primary cells. J Lab Autom 2016;21:496 509.
2009;21:2003 6. [142] Buttery LDK, Bishop AE. Introduction to tissue engineering.
[126] Wang S, Lee JM, Yeong WY. Smart hydrogels for 3D bioprint- Biomaterials, artificial organs and tissue engineering. Woodhead
ing. Int J Bioprinting 2015;1. Publishing; 2005. p. 193 200.
[127] Khattak SF, Bhatia SR, Roberts SC. Pluronic F127 as a cell [143] Kaur G, Dufour JM. Cell lines: valuable tools or useless artifacts.
encapsulation material: utilization of membrane-stabilizing Spermatogenesis 2012;2:1 5.
agents. Tissue Eng 2005;11:974 83. [144] Sison-Young RL, Lauschke VM, Johann E, Alexandre E,
[128] Gong C, Shi S, Dong P, Zheng X, Fu S, Guo G, et al. In vitro Antherieu S, Aerts H, et al. A multicenter assessment of single-cell
drug release behavior from a novel thermosensitive composite models aligned to standard measures of cell health for prediction
hydrogel based on Pluronic F127 and poly(ethylene glycol)-poly of acute hepatotoxicity. Arch Toxicol 2017;91:1385 400.
(epsilon-caprolactone)-poly(ethylene glycol) copolymer. BMC [145] Itskovitz-Eldor J, Schuldiner M, Karsenti D, Eden A, Yanuka O,
Biotechnol 2009;9:8. Amit M, et al. Differentiation of human embryonic stem cells
1440 PART | TWENTY ONE Emerging technologies

into embryoid bodies compromising the three embryonic germ [164] Nzou G, Wicks RT, Wicks EE, Seale SA, Sane CH, Chen A,
layers. Mol Med 2000;6:88 95. et al. Human cortex spheroid with a functional blood brain bar-
[146] Blum B, Benvenisty N. The tumorigenicity of human embryonic rier for high-throughput neurotoxicity screening and disease
stem cells. Adv Cancer Res 2008;100:133 58. modeling. Sci Rep 2018;8.
[147] Takahashi K, Yamanaka S. Induction of pluripotent stem cells [165] Kehoe S, Zhang XF, Boyd D. FDA approved guidance conduits
from mouse embryonic and adult fibroblast cultures by defined and wraps for peripheral nerve injury: a review of materials and
factors. Cell 2006;126:663 76. efficacy. Injury 2012;43:553 72.
[148] Zhao J, Jiang W, Sun C, Hou C, Yang X-M, Gao J. Induced plu- [166] Ichihara S, Inada Y, Nakamura T. Artificial nerve tubes and their
ripotent stem cells: origins, applications, and future perspectives. application for repair of peripheral nerve injury: an update of
J Zhejiang Univ Sci, B 2013;14:1059 69. current concepts. Injury 2008;39(Suppl.):29 39.
[149] Yu C, Kornmuller A, Brown C, Hoare T, Flynn LE. [167] Lizama MDLAP, Takemoto RM, Ranzani-Paiva MJT, Ayroza
Decellularized adipose tissue microcarriers as a dynamic culture LMDS, Pavanelli GC. Relação parasito-hospedeiro em peixes de
platform for human adipose-derived stem/stromal cell expansion. pisciculturas da região de Assis, Estado de São Paulo, Brasil. 2.
Biomaterials 2017;120:66 80. Piaractus mesopotamicus (Holmberg, 1887). Acta Sci—Biol Sci
[150] Astori G, Soncin S, Lo Cicero V, Siclari F, Sürder D, Turchetto 2007;29:437 45.
L, et al. Bone marrow derived stem cells in regenerative medi- [168] Kim YT, Haftel VK, Kumar S, Bellamkonda RV. The role of
cine as advanced therapy medicinal products. Am J Transl Res aligned polymer fiber-based constructs in the bridging of long
2010;2:285 95. peripheral nerve gaps. Biomaterials 2008;29:3117 27.
[151] Tang X-P, Zhang M, Yang X, Chen L-M, Zeng Y. Differentiation [169] M. Qiu, S. Rao, J. Zhu, P.P. Chen, S. Fu, W. Yuan, et al.,
of human umbilical cord blood stem cells into hepatocytes in vivo Mechanical properties of MJ-class toroidal magnet wound by
and in vitro. World J Gastroenterol 2006;12:4014 19. composite HTS conductor. IEEE Transactions on Applied
[152] Chunmeng S, Tianmin C. Skin: a promising reservoir for adult Superconductivity 27, 2017.
stem cell populations. Med Hypotheses 2004;62:683 8. [170] Bray F, Ferlay J, Soerjomataram I, Siegel RL, Torre LA, Jemal
[153] Bunnell BA, Estes BT, Guilak F, Gimble JM. Differentiation of A. Global cancer statistics 2018: GLOBOCAN estimates of inci-
adipose stem cells. Methods in molecular biology, Humana dence and mortality worldwide for 36 cancers in 185 countries.
Press. 2008. p. 155 71. CA Cancer J Clin 2018;68:394 424.
[154] Wang C, Meng H, Wang X, Zhao C, Peng J, Wang Y. [171] Mak IW, Evaniew N, Ghert M. Lost in translation: animal mod-
Differentiation of bone marrow mesenchymal stem cells in osteo- els and clinical trials in cancer treatment. Am J Transl Res
blasts and adipocytes and its role in treatment of osteoporosis. 2014;6:114 18.
Med Sci Monit 2016;22:226 33. [172] Mcmillin DW, Negri JM, Mitsiades CS. The role of tumour-
[155] Russell AL, Lefavor R, Durand N, Glover L, Zubair AC. stromal interactions in modifying drug response: challenges and
Modifiers of mesenchymal stem cell quantity and quality. opportunities. Nat Rev Drug Discov 2013;12:217 28.
Transfusion 2018;58:1434 40. [173] Leonard F, Godin B. 3D in vitro model for breast cancer research
[156] Gurǎu C, Gurǎu G, Alexandru P, Ghiban N. Ultrafine microstruc- using magnetic levitation and bioprinting method. Methods Mol
tures of steel X60 induced by severe plastic deformation. Metal Biol 2016;1406:239 51.
Int 2013;18:37 41. [174] Padrón JM, Van Der Wilt CL, Smid K, Smitskamp-Wilms E,
[157] Freire-Regatillo A, Argente-Arizón P, Argente J, Garcı́a-Segura Backus HHJ, Pizao PE, et al. The multilayered postconfluent cell
LM, Chowen JA. Non-neuronal cells in the hypothalamic adapta- culture as a model for drug screening. Crit Rev Oncol Hematol
tion to metabolic signals. Front Endocrinol (Lausanne) 2017;8:51. 2000;36:141 57.
[158] Gu Q, Tomaskovic-Crook E, Lozano R, Chen Y, Kapsa RM, [175] Zhao Y, Yao R, Ouyang L, Ding H, Zhang T, Zhang K, et al.
Zhou Q, et al. Functional 3D neural mini-tissues from printed Three-dimensional printing of Hela cells for cervical tumor
gel-based bioink and human neural stem cells. Adv Healthc model in vitro. Biofabrication 2014;6:035001.
Mater 2016;5:1429 38. [176] Swaminathan S, Hamid Q, Sun W, Clyne AM. Bioprinting of 3D
[159] Johnston GAR. Advantages of an antagonist: bicuculline and breast epithelial spheroids for human cancer models.
other GABA antagonists. Br J Pharmacol 2013;169:328 36. Biofabrication 2019;11:025003.
[160] Hsieh F-Y, Lin H-H, Hsu S. 3D bioprinting of neural stem cell- [177] Soman P, Kelber JA, Lee JW, Wright TN, Vecchio KS, Klemke
laden thermoresponsive biodegradable polyurethane hydrogel RL, et al. Cancer cell migration within 3D layer-by-layer micro-
and potential in central nervous system repair. Biomaterials fabricated photocrosslinked PEG scaffolds with tunable stiffness.
2015;71:48 57. Biomaterials 2012;33:7064 70.
[161] Lancaster MA, Renner M, Martin CA, Wenzel D, Bicknell LS, [178] McMillin DW, Delmore J, Weisberg E, Negri JM, Geer DC,
Hurles ME, et al. Cerebral organoids model human brain devel- Klippel S, et al. Tumor cell-specific bioluminescence platform to
opment and microcephaly. Nature 2013;501:373 9. identify stroma-induced changes to anticancer drug activity. Nat
[162] Lancaster MA, Corsini NS, Wolfinger S, Gustafson EH, Phillips Med 2010;16:483 9.
AW, Burkard TR, et al. Guided self-organization and cortical [179] Wang Y, Shi W, Kuss M, Mirza S, Qi D, Krasnoslobodtsev A,
plate formation in human brain organoids. Nat Biotechnol et al. 3D bioprinting of breast cancer models for drug resistance
2017;35:659 66. study. ACS Biomater Sci Eng 2018;4:4401 11.
[163] Daneman R, Prat A. The blood-brain barrier. Cold Spring Harb [180] Heinrich MA, Bansal R, Lammers T, Zhang YS, Michel
Perspect Biol 2015;7:a020412. Schiffelers R, Prakash J. 3D-bioprinted mini-brain: a
 Biofabricated three-dimensional tissue models Chapter | 75 1441

glioblastoma model to study cellular interactions and therapeu- [198] Sohlenius-Sternbeck AK. Determination of the hepatocellularity
tics. Adv Mater 2019;31:e1806590. number for human, dog, rabbit, rat and mouse livers from protein
[181] Benjamin EJ, Virani SS, Callaway CW, Chamberlain AM, concentration measurements. Toxicol In Vitro 2006;20:1582 6.
Chang AR, Cheng S, et al. Heart disease and stroke statistics— [199] Fontana RJ, Hayashi PH, Gu J, Reddy KR, Barnhart H, Watkins
2018 update: a report from the American Heart Association. PB, et al. DILIN Network, Idiosyncratic drug-induced liver injury
Circulation 2018;137:E67 492. is associated with substantial morbidity and mortality within 6
[182] Nkomo VT, Gardin JM, Skelton TN, Gottdiener JS, Scott CG, months from onset. Gastroenterology 2014;147:96 108.e4.
Enriquez-Sarano M. Burden of valvular heart diseases: a [200] Shaw PJ, Ganey PE, Roth RA. Idiosyncratic drug-induced liver
population-based study. Lancet 2006;368:1005 11. injury and the role of inflammatory stress with an emphasis on
[183] Duan B. State-of-the-art review of 3D bioprinting for cardiovas- an animal model of trovafloxacin hepatotoxicity. Toxicol Sci
cular tissue engineering. Ann Biomed Eng 2017;45:195 209. 2010;118:7 18.
[184] Bouten CVC, Dankers PYW, Driessen-Mol A, Pedron S, Brizard [201] Moreira RK. Hepatic stellate cells and liver fibrosis. Arch Pathol
AMA, Baaijens FPT. Substrates for cardiovascular tissue engi- Lab Med 2007;131:1728 34.
neering. Adv Drug Deliv Rev 2011;63:221 41. [202] Krenkel O, Tacke F. Liver macrophages in tissue homeostasis
[185] El Sabbagh A, Eleid MF, Al-Hijji M, Anavekar NS, Holmes DR, and disease. Nat Rev Immunol 2017;17:306 21.
Nkomo VT, et al. The various applications of 3D printing in car- [203] Norona LM, Nguyen DG, Gerber DA, Presnell SC, Mosedale M,
diovascular diseases. Curr Cardiol Rep 2018;20:47. Watkins PB. Bioprinted liver provides early insight into the role
[186] Giannopoulos AA, Mitsouras D, Yoo SJ, Liu PP, Chatzizisis YS, of Kupffer cells in TGF-β1 and methotrexate-induced fibrogen-
Rybicki FJ. Applications of 3D printing in cardiovascular dis- esis. PLoS One 2019;14:e0208958.
eases. Nat Rev Cardiol 2016;13:701 18. [204] Otieno MA, Gan J, Proctor W. Status and future of 3D cell cul-
[187] Hirt MN, Hansen A, Eschenhagen T. Cardiac tissue engineering : ture in toxicity testing. Methods in pharmacology and toxicology.
state of the art. Circ Res 2014;114:354 67. New York: Humana Press; 2018. p. 249 61.
[188] Wang PY, Yu J, Lin JH, Tsai WB. Modulation of alignment, [205] Carmeliet P, Jain RK. Angiogenesis in cancer and other diseases.
elongation and contraction of cardiomyocytes through a combi- Nature 2000;407:249 57.
nation of nanotopography and rigidity of substrates. Acta [206] Folkman J. Angiogenesis in cancer, vascular, rheumatoid and
Biomater 2011;7:3285 93. other disease. Nat Med 1995;1:27 30.
[189] Stoppel WL, Kaplan DL, Black LD. Electrical and mechanical [207] Roskoski R. Vascular endothelial growth factor (VEGF) signaling
stimulation of cardiac cells and tissue constructs. Adv Drug in tumor progression. Crit Rev Oncol Hematol 2007;62:179 213.
Deliv Rev 2016;96:135 55. [208] DiPietro LA. Angiogenesis and wound repair: when enough is
[190] Lasher RA, Pahnke AQ, Johnson JM, Sachse FB, Hitchcock enough. J Leukoc Biol 2016;100:979 84.
RW. Electrical stimulation directs engineered cardiac tissue to an [209] Pashneh-Tala S, MacNeil S, Claeyssens F. The tissue-engineered
age-matched native phenotype. J Tissue Eng 2012;3:1 15. vascular graft-past, present, and future. Tissue Eng, B: Rev
[191] Liu J, He J, Liu J, Ma X, Chen Q, Lawrence N, et al. Rapid 3D 2015;22:68.
bioprinting of in vitro cardiac tissue models using human embry- [210] Lovett M, Lee K, Edwards A, Kaplan DL. Vascularization strate-
onic stem cell-derived cardiomyocytes. Bioprinting 2019;13: gies for tissue engineering. Tissue Eng, B: Rev 2009;15:353 70.
e00040. [211] Norotte C, Marga FS, Niklason LE, Forgacs G. Scaffold-free
[192] Ribas J, Sadeghi H, Manbachi A, Leijten J, Brinegar K, Zhang vascular tissue engineering using bioprinting. Biomaterials
YS, et al. Cardiovascular organ-on-a-chip platforms for drug dis- 2009;30:5910 17.
covery and development. Appl In Vitro Toxicol 2016;2:82 96. [212] Jia W, Gungor-Ozkerim PS, Zhang YS, Yue K, Zhu K, Liu W,
[193] Jang J, Park HJ, Kim SW, Kim H, Park JY, Na SJ, et al. 3D et al. Direct 3D bioprinting of perfusable vascular constructs
printed complex tissue construct using stem cell-laden decellular- using a blend bioink. Biomaterials 2016;106:58 68.
ized extracellular matrix bioinks for cardiac repair. Biomaterials [213] Wendelboe AM, Raskob GE. Global burden of thrombosis. Circ
2017;112:264 74. Res 2016;118:1340 7.
[194] Hockaday LA, Kang KH, Colangelo NW, Cheung PYC, Duan B, [214] Raskob GE, Angchaisuksiri P, Blanco AN, et al. Contributor to
Malone E, et al. Rapid 3D printing of anatomically accurate and global disease burden. Semin Thromb Hemost 2014;40:724 35.
mechanically heterogeneous aortic valve hydrogel scaffolds. [215] White RH. The epidemiology of venous thromboembolism.
Biofabrication 2012;4:035005. Circulation 2003;107:I4 8.
[195] Duan B, Hockaday LA, Kang KH, Butcher JT. 3D Bioprinting of [216] Cushman M. Epidemiology and risk factors for venous thrombo-
heterogeneous aortic valve conduits with alginate/gelatin hydro- sis. Semin Hematol 2007;44:62 9.
gels. J Biomed Mater Res, A 2013;101 A:1255 64. [217] Lyaker MR, Tulman DB, Dimitrova GT, Pin RH, Papadimos TJ.
[196] van der Valk D, van der Ven C, Blaser M, Grolman J, Wu P-J, Arterial embolism. Int J Crit Illn Inj Sci 2013;3:77 87.
Fenton O, et al. Engineering a 3D-bioprinted model of human [218] Costa PF, Albers HJ, Linssen JEA, Middelkamp HHT, van der
heart valve disease using nanoindentation-based biomechanics. Hout L, Passier R, et al. Mimicking arterial thrombosis in a 3D-
Nanomaterials 2018;8:296. printed microfluidic in vitro vascular model based on computed
[197] Choi YJ, Kim TG, Jeong J, Yi HG, Park JW, Hwang W, et al. tomography angiography data. Lab Chip 2017;17:2785 92.
3D cell printing of functional skeletal muscle constructs using [219] Zhang YS, Davoudi F, Walch P, Manbachi A, Luo X, Dell’Erba
skeletal muscle-derived bioink. Adv Healthc Mater V, et al. Bioprinted thrombosis-on-a-chip. Lab Chip
2016;5:2636 45. 2016;16:4097 105.