Jensen et al.

Diabetol Metab Syndr (2018) 10:4

https://doi.org/10.1186/s13098-018-0307-8 Diabetology &
Metabolic Syndrome

RESEARCH Open Access

Dietary fat stimulates development

of NAFLD more potently than dietary fructose
in Sprague–Dawley rats
Victoria Svop Jensen1,2*, Henning Hvid2, Jesper Damgaard2, Helle Nygaard2, Camilla Ingvorsen3,
Erik Max Wulff4, Jens Lykkesfeldt1 and Christian Fledelius2

Abstract 
Background:  In humans and animal models, excessive intake of dietary fat, fructose and cholesterol has been linked
to the development of non-alcoholic fatty liver disease (NAFLD). However, the individual roles of the dietary compo-
nents remain unclear. To investigate this further, we compared the effects of a high-fat diet, a high-fructose diet and a
combination diet with added cholesterol on the development of NAFLD in rats.
Methods:  Forty male Sprague–Dawley rats were randomized into four groups receiving either a control-diet (Con-
trol: 10% fat); a high-fat diet (HFD: 60% fat, 20% carbohydrate), a high-fructose diet [HFr: 10% fat, 70% carbohydrate
(mainly fructose)] or a high-fat/high-fructose/high-cholesterol-diet (NASH: 40% fat, 40% carbohydrate (mainly fruc-
tose), 2% cholesterol) for 16 weeks.
Results:  After 16 weeks, liver histology revealed extensive steatosis and inflammation in both NASH- and HFD-fed
rats, while hepatic changes in HFr-rats were much more subtle. These findings were corroborated by significantly
elevated hepatic triglyceride content in both NASH- (p < 0.01) and HFD-fed rats (p < 0.0001), elevated hepatic cho-
lesterol levels in NASH-fed rats (p < 0.0001), but no changes in HFr-fed rats, compared to Control. On the contrary,
only HFr-fed rats developed dyslipidemia as characterized by higher levels of plasma triglycerides compared to all
other groups (p < 0.0001). Hepatic dysfunction and inflammation was confirmed in HFD-fed rats by elevated levels of
hepatic MCP-1 (p < 0.0001), TNF-alpha (p < 0.001) and plasma β-hydroxybutyrate (p < 0.0001), and in NASH-fed rats by
elevated levels of hepatic MCP-1 (p < 0.01), increased hepatic macrophage infiltration (p < 0.001), and higher plasma
levels of alanine aminotransferase (p < 0.0001) aspartate aminotransferase (p < 0.05), haptoglobin (p < 0.001) and
TIMP-1 (p < 0.01) compared to Control.
Conclusion:  These findings show that dietary fat and cholesterol are the primary drivers of NAFLD development and
progression in rats, while fructose mostly exerts its effect on the circulating lipid pool.
Keywords:  Non-alcoholic fatty liver disease, NAFLD, Animal models, Diet, Fat, Fructose, Cholesterol, Non-alcoholic
steatohepatitis, NASH, Rat

Background syndrome [1]. At present, NAFLD is assuming epidemic

Non-alcoholic fatty liver disease (NAFLD) constitutes an proportions affecting more than 25% of the world’s popu-
increasingly prevalent liver disorder and has been sug- lation, likely related to a concurrent rise in the prevalence
gested to be the hepatic manifestation of the metabolic of obesity and type 2 diabetes [2], although an increasing
proportion of normal weight individuals are also affected
pointing towards dyslipidemia as an important independ-
*Correspondence: [email protected] ent risk factor [3]. NAFLD denotes a broad range of liver
2
Insulin Pharmacology, Novo Nordisk A/S, Novo Nordisk Park 1, pathologies, spanning from simple hepatic steatosis, to
2760 Maaloev, Denmark
Full list of author information is available at the end of the article
advanced non-alcoholic steatohepatitis (NASH), which

© The Author(s) 2018. This article is distributed under the terms of the Creative Commons Attribution 4.0 International License
(http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium,
provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license,
and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/
publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
Jensen et al. Diabetol Metab Syndr (2018) 10:4 Page 2 of 13

without intervention may progress to transplant-requir- Methods

ing cirrhosis and hepatocellular carcinoma. Whereas Animals
simple hepatic steatosis is mainly characterized by lipid Forty male Sprague–Dawley rats were purchased from
accumulation in > 5% of hepatocytes, NASH is a condi- Charles River Laboratories (Sulzfeld, Germany). Ani-
tion further complicated by lobular inflammatory infil- mals were acclimatized for 2 weeks upon arrival and were
trations and the presence of ballooning hepatocytes with 12 weeks old at the beginning of the experiment, weigh-
or without concurrent fibrosis [4]. ing approximately 440–460 g. They were housed two per
It has been hypothesized that progression through cage with access to non-chlorinated, non-acidific tap-
NAFLD stages involves multiple adverse “hits” where water and with unrestricted access to standard rodent
both hepatic inflammation and oxidative stress are key chow (Altromin 1324, Brogaarden, Denmark) until initia-
facilitators [5]. Accordingly, serum and hepatic levels of tion of the experiment. Temperature in the animal rooms
proinflammatory cytokines such as tumor-necrosis factor was maintained at 20–25  °C with a light/dark cycle of
alpha (TNF-α) and monocyte-chemoattractant protein-1 12/12  h, a relative humidity of 30–70%, and air change
(MCP-1) have been found to be increased in patients 8–15 times/h.
with both simple hepatic steatosis and NASH [6–9].
Also, the glycoproteins haptoglobin and tissue-inhibitor Experimental design
of metalloproteinase 1 (TIMP-1) have been shown to be Immediately prior to study start, rats were block-ran-
elevated in plasma of patients with more advanced stages domized into four diet groups based on body weight
of NAFLD and have therefore recently been suggested as (n  =  10/group). Briefly, rats were sorted in order of
useful clinical plasma biomarkers, indicative of hepato- ascending weight into blocks, and within each block, rats
cyte ballooning and liver fibrosis, respectively [10, 11]. were then randomly assigned to receive either a control
Even though the specific etiology of NAFLD remains diet (Control), a high-fat diet (HFD), a high-fructose diet
unclear, dietary fat and cholesterol have been linked to (HFr) or a high-fat/high-fructose/high-cholesterol-diet
the development of hepatic steatosis and NASH in both (NASH), (Research Diets, NJ, USA) for 16  weeks. The
humans and animal models [12–14] and more recently, length of the study period was chosen to enhance the
the marked increase in consumption of the simple car- probability that NASH with concurrent fibrosis would
bohydrate fructose, has been pointed out as another pos- be established, as has previously been described in
sible contributory factor [15, 16]. The realization that Sprague–Dawley rats fed a high-fat diet [19]. The exact
diet is an important contributor to the pathogenesis of nutritional composition of each diet is shown in Table 1.
NAFLD has resulted in a considerable variety of diet- All diets were stored at −  20  °C throughout the study
induced animal models, where the majority (apart from until use, and feed-remains were exchanged with fresh
those based on nutritional deficiencies) is based on a feed twice weekly to prevent purification. To obtain base-
high-fat diet with varying levels of simple carbohydrate line values for all parameters assessed at the 16 week time
and cholesterol [17, 18]. However, the individual roles of point, animals were qCT-scanned, quantitative magnetic
fat, carbohydrates and cholesterol in the development of resonance (qMR)-scanned and blood sampled at study
NAFLD are still not entirely clear. Thus, improved insight start. After 16 weeks, animals were euthanized by exsan-
from animal studies on the individual contribution of fat, guination under isoflurane anesthesia by incision of the
simple carbohydrates and cholesterol to the metabolic abdominal aorta. Immediately following exsanguination,
and inflammatory characteristics of NAFLD is an impor-
tant prerequisite in understanding the disease complexity
in patients. Furthermore, establishment of animal models Table 1  Nutritional composition of diets
of NAFLD/NASH that more adequately mimic human Nutrient composition Control NASH HFD HFr
pathology will be valuable when testing novel therapeu-
tics for NAFLD/NASH. Fat (%) 10 40 60 10
The aim of the present study was therefore to compare Carbohydrates (total) (%) 70 40 20 70
the effects of diets high in either dietary fat, dietary fruc-  Fructose (%) 0 20 0 60
tose or a combination diet with added cholesterol on devel-  Sucrose (%) 0 10 7 1
opment of NAFLD, dyslipidemia and inflammation in  Corn starch (%) 55 0 0 9
Sprague–Dawley rats. Development of NAFLD was evalu-  Maltodextrin 10 (%) 15 10 13 0
ated over time both histologically and biochemically and Protein (%) 20 20 20 20
quantitative computed tomography (qCT) was used as a Cholesterol (%) – 2 – –
non-invasive marker of hepatic steatosis, enabling monitor- Metabolizable energy (kcal/g) 3.85 4.49 5.24 3.85
ing of NAFLD progression throughout the study period. NASH NASH-diet, HFD high-fat diet, HFr high-fructose diet
Jensen et al. Diabetol Metab Syndr (2018) 10:4 Page 3 of 13

the liver, a 5 cm section of the proximal jejunum and the qCT-scans were performed after 8 and 16  weeks in iso-
two epididymal fat depots were excised, weighed and flurane-anesthetized rats using a Latheta CT-scanner
freeze clamped for further analysis. Food intake and body (LCT-200 series, Aloka co. LTD, Tokyo, Japan). Changes
weight was measured twice weekly throughout the study. in liver density were calculated by subtracting baseline
Food intake was recorded by calculating the difference qCT values from week 8 and 16 time point values.
between the amount of administered food and remain-
ing food in the cage. These data were used to calculate Liver biochemistry
the accumulated energy intake after the 16 weeks. qCT- Levels of hepatic TG, TC and liver glycogen were ana-
and qMR-scans was performed on a bi-monthly basis, lyzed on homogenized liver tissue sampled from the left
and blood samples were collected at baseline and at study lateral lobe using a Cobas 6000 c501 instrument (Roche
termination. Diagnostics GmbH 68206 Mannheim, Germany) accord-
ing to manufacturer’s instructions and as previously
Plasma samples described [22].
Blood samples were taken from the sublingual vein in
conscious non-fasted animals. They were collected Histology
in ­K3-EDTA microvettes and after centrifugation The right medial, the left lateral, and the caudate lobe
plasma was isolated and kept at −  20  °C until further of the liver were collected for histological examination.
analysis. Triglycerides (TG), total cholesterol (TC), Samples were fixed in 10% Neutral Buffered Formalin,
high-density lipoprotein cholesterol (HDL-C), free processed to paraffin, imbedded and cut in 2–4 µm sec-
fatty acids (FFA), alanine aminotransferase (ALAT), tions. In addition, a sample from the left medial lobe was
aspartate aminotransferase (ASAT), Haptoglobin and snap frozen for cryo-sectioning. Steatosis was evaluated
β-hydroxybutyrate were measured using a Cobas 6000 in all three lobes with Mayer’s Haematoxylin and Eosin
c501 instrument (Roche Diagnostics GmbH, D-68296 (H&E) on paraffin section (10 animals/gr) and confirmed
Mannheim, Germany), according to manufacturer’s by Oil Red O-stain on frozen cryo-sections of the left
instructions. Plasma levels of MCP-1 and TIMP-1 were medial lobe (2–3 animals/group). Inflammation and
analyzed using a multiplex assay, (K15179-C1, Mes- fibrosis were evaluated based on the morphology on the
oscale Discovery, MD, USA). H&E stain, collagen deposition on a Picro Sirius stain
Additionally, plasma samples were collected from 4-h and macrophage infiltration based on a CD68 immu-
fasted animals at week 15 (collected 1 week prior to week nohistochemistry (IHC) stain. For the CD68 IHC stain,
16 blood samples, to avoid the fasting interfering with antigens were first retrieved by TEG buffer pH 9.0 and
plasma lipid parameters) and assayed for endogenous subsequently blocked with 0.5% hydrogen peroxide fol-
insulin and glucose. Samples for blood glucose meas- lowed by avidin and biotin (Invitrogen, CA, USA). The
urements (10  µL) were collected in capillary tubes and slides were then incubated with primary antibody (3 µg/
transferred to 500  µL system solution. Blood glucose mL, MCA341R, AbD serotec, CA, USA) for 60  min,
levels were analyzed using the glucose oxidase method biotinylated secondary antibody for 60  min (1  µg/mL,
at a Biosen apparatus (EKF Diagnostics, Barleben, Ger- 715-065-151, Jackson ImmunoSearch, PA, USA) and
many) according to manufacturer’s instructions. Samples ABC reagent (Vector Laboratories, CA, USA). Finally,
for endogenous insulin measurements were collected in macrophages were visualised with a DAB reagent (Dako,
­K3-EDTA microvettes and after centrifugation; plasma Glostrup, Denmark). All sections were scanned with
was isolated and analyzed as previously described [20]. a Hamamutsu slide scanner and later evaluated using
Leptin levels were quantified using a Milliplex assay NanoZoomer Digital Pathology Image software (Hama-
(RADPKMAG80-K, Merck, Hellerup, DK). matsu, Hamamatsu City, Japan). The collagen and mac-
rophage areas were quantified by Visiomorph software
qMR and qCT (Visiopharm, Hoersholm, Denmark).
To determine total fat mass, all animals underwent qMR-
scans after 8 and 16 weeks using an EchoMRI Body Com- Inflammatory markers in tissues
position Analyser (EchoMRI, Houston, TX, USA). Mass As described above, livers, epididymal fat depots, and
measurements of fat tissue were performed according to jejunal segments were excised from the animals immedi-
manufacturer’s instructions and as previously described ately after sacrifice and stored at −  80  °C until analysis.
[21]. Tissue protein concentrations of the three tissues were
In order to assess the development and progression first determined using a Pierce BCA Protein Assay Kit
of NAFLD, qCT scans were used to quantify liver den- (Thermo Fisher Scientific, MA, USA) method accord-
sity as an indirect measure of hepatic fat content. Liver ing to manufacturer’s instruction. TNF-α and MCP-1
Jensen et al. Diabetol Metab Syndr (2018) 10:4 Page 4 of 13

levels were subsequently determined in tissue homogen- Table 2 Baseline characteristics of rats immediately
ates with enzyme-linked immunosorbent assay (ELISA) after randomization to either Control-, NASH-, HFD- or
kits (AB100785 and AB100778; Abcam, Cambridge, HFr-diets
UK) according to manufacturer’s instructions. Absorb- Control NASH HFD HFr
ance was read using a Spectramax 340PC384 microplate
Body weight (g) 454.9 ± 6.6 457.7 ± 5.7 442.7 ± 7.0 447.7 ± 8.8
reader at 450 nm (Molecular Devices, CA, USA).
Fat mass (g) 43.6 ± 2.6 44.4 ± 2.4 46.7 ± 2.0 44.9 ± 2.6
Liver density (HU) 40.0 ± 1.0 39.6 ± 0.6 40.7 ± 1.0 38.9 ± 0.8
Statistical analysis
Plasma TG (mM) 0.9 ± 0.1 0.9 ± 0.1 1.0 ± 0.1 1.3 ± 0.2
Statistical analyses were done using GraphPad Prism ver-
sion 6.05 (GraphPad Software Inc., La Jolla, CA, USA). Plasma FFA (mM) 0.3 ± 0.02 0.3 ± 0.03 0.4 ± 0.03 0.3 ± 0.02
Data were assumed to be normally distributed and con- Plasma cholesterol 1.6 ± 0.1 1.5 ± 0.1 1.7 ± 0.1 1.6 ± 0.1
(mM)
firmed by visual inspection of qq-plots. In case of severe
Plasma HDL-C (mM) 1.0 ± 0.04 1.0 ± 0.1 1.1 ± 0.04 1.0 ± 0.1
deviations from normality, statistical analyses were per-
ALAT (U/L) 33.0 ± 2.1 31.9 ± 1.7 35.7 ± 2.8 33.1 ± 1.9
formed on log-transformed data (natural logarithm) or
ASAT (U/L) 62.4 ± 1.9 59.9 ± 3.5 61.5 ± 2.9 63.2 ± 2.3
with the use of non-parametric tests. Data are presented
Results are presented as mean ± SEM
as mean ± SEM, except for log-transformed data, which
HU Hounsfield units, TG triglycerides, FFA free fatty acids, HDL-C high-density-
are presented as geometric means with 95% confidence lipoprotein-cholesterol, ALAT alanine aminotransferase, ASAT aspartate
intervals. Differences in means between diet groups for aminotransferase. Sample sizes: n = 10/group
each parameter were analyzed using one-way ANOVA,
repeated measures two-way ANOVA or Kruskal–Wallis
tests where appropriate, and compared after 16 weeks on was significantly greater in NASH and HFD compared
the diets. Bonferroni or Dunn corrections, respectively, to Control (p < 0.0001 and p < 0.05) and in NASH com-
were used to adjust for multiple comparisons. Outliers in pared to HFD and HFr (p < 0.0001).
data sets were identified and removed using the ROUT- Livers from NASH-fed rats weighed significantly more
function in GraphPad Prism. p-values  <  0.05 were con- compared to Control (p < 0.0001), HFD (p < 0.0001) and
sidered significant. HFr (p < 0.001, Table 3), as did livers from HFr-fed rats
compared to HFD (p  <  0.05). Hepatic TG content was
Results higher in NASH- and HFD-fed rats compared to both
Baseline characteristics for Control, HFD, HFr and Control (p < 0.01 and p < 0.0001) and HFr (p < 0.01 and
NASH-groups are given in Table 2 and show that groups p  <  0.0001, Fig.  2b). Additionally, HFD-fed animals had
were comparable at study start for all parameters. higher hepatic TG levels compared to NASH (p < 0.01).
Hepatic TG levels in rats fed HFr did not differ from
Effects of diet on body weight, energy intake and fat those fed the Control diet. Hepatic TC levels were signifi-
distribution cantly elevated in NASH-fed animals compared to Con-
Figure  1 shows changes in body weight, fat mass and trol, HFD, and HFr (p  <  0.0001, Fig.  2c). Liver glycogen
energy intake. Body weight increased throughout the was significantly lower in both NASH and HFD com-
study in all groups, but did not differ from Control after pared to Control and HFr (p < 0.0001, Fig. 2d). At week
16  weeks (Fig.  1a). Fat mass was significantly increased 16, ALAT-levels were significantly higher in NASH-fed
in HFD-fed animals compared to both those fed Con- rats compared to Control- (p  <  0.0001, Fig.  3a), HFD-
trol and HFr-diets (p < 0.05, Fig. 1b) after 16 weeks. After and HFr-fed rats (p < 0.01). Plasma levels of ASAT were
8  weeks, accumulated energy intake was significantly higher in the NASH-group compared to both Control-
higher in NASH-fed rats, compared to both Control- and HFr-groups (p < 0.05, Fig. 3b). Furthermore, plasma
and HFr-fed rats (p < 0.001 and p < 0.05, Fig. 1c) and in levels of the β-oxidation marker β-hydroxybuturate were
HFD-fed rats compared to Control (p < 0.05). This differ- increased in HFD-fed rats compared to Control-, NASH-
ence remained significant only in NASH-fed compared to and HFr-groups (p  <  0.0001, p  <  0.01 and p  <  0.0001,
Control-fed rats after 16 weeks (p < 0.05). Table 3).
The histological evaluation is presented in Fig. 4. After
Effects of diet on the liver 16  weeks, hepatic steatosis was present in NASH, HFD
Throughout the study, liver density (used as an indirect and HFr-fed rats as confirmed by corresponding Oil Red
measure of liver fat content) continuously declined in O stains. No pathological changes were observed in liv-
all four diet groups as measured by qCT with the effect ers of Control-fed rats. Steatosis was phenotypically dis-
being most pronounced in the NASH-group (Fig.  2a). tinguishable between NASH-, HFD- and HFr-fed rats.
At both 8 and 16  weeks, the decrease in liver density While zonal distribution was comparable (originating in
Jensen et al. Diabetol Metab Syndr (2018) 10:4 Page 5 of 13

Fig. 1  Effects on body weight, fat mass and energy intake of Control-, NASH-, HFD- and HFr-diets. a Body weight increased in all 4 diet groups
throughout the study, but did not differ after 16 weeks. b After 8 weeks fat mass was significantly increased in HFD-fed rats compared to Con-
trol- and HFr-fed rats. This effect was still present at week 16. c After 8 weeks, accumulated energy intake was significantly higher in NASH-fed rats,
compared to both Control- and HFr-fed rats and in HFD-fed rats compared to Control. This remained significant only in NASH-fed compared to
Control-fed rats after 16 weeks. Comparisons between groups: *NASH vs. Control; #NASH vs. HFr; †HFD vs. Control. Statistical significance: *p < 0.05;
***p < 0.001; #p < 0.05; ##p < 0.01; †p < 0.05

zone 1, periportally), steatosis in HFD-rats was almost NASH-group were more pronounced, as exemplified by
exclusively found to be of the microvesicular type; in larger and more disseminated infiltrates. Accordingly,
HFr-fed rats almost exclusively macrovesicular; while the hepatic macrophage numbers were two- to threefold
NASH-fed rats represented an intermediate between the increased in the NASH-group compared to Control- and
two, with both macro- and microvesicular steatosis. Fur- HFD-fed rats (p < 0.001 and p < 0.01, Fig. 5a, b). Hepatic
thermore, steatosis in the NASH-group at week 16 was deposition levels of collagen was higher in NASH-
much more pronounced, involving not only zone 1, but animals than in other groups, albeit only significantly
occasionally also zones 2 and 3 and in some animals, only increased compared to HFD (p < 0.001, Fig. 5c, d).
few areas displayed normal liver architecture. Inflamma-
tory hepatic infiltration was identified in both NASH-, Effects of diet on plasma lipids and general metabolic state
HFD- and HFr-fed rats. In all three groups, inflammation Plasma FFA did not differ between groups (Table  3).
was characterized by being of primarily mononuclear HFr-feeding induced a significant increase in circu-
composition with occasional neutrophilic involvement. lating TG, compared to Control-, NASH- and HFD
However in HFr-fed rats, inflammatory cells appeared to (p  <  0.0001, Table  3). Plasma HDL-c levels were simi-
center around lipid-loaded hepatocytes and form struc- lar between groups, as were the plasma cholesterol lev-
tures resembling lipogranulomas, whereas inflamma- els (Table  3). Fasting hyperglycemia was present in all
tory infiltrates observed in HFD-fed rats were randomly groups compared to Control after 16 weeks (NASH and
scattered and not consistently associated with steato- HFr: p < 0.01, HFD: p < 0.0001, Table 3). However, fast-
sis. As with the steatosis, inflammatory changes in the ing plasma insulin levels did not differ between groups
Jensen et al. Diabetol Metab Syndr (2018) 10:4 Page 6 of 13

Fig. 2  Effects on liver in Control-, NASH-, HFD- and HFr-fed rats. a Throughout the study, liver density continuously declined in all four diet groups,
with the effect being more pronounced in the NASH-group. Both halfway through the study period and at the week 16 time point the decline in
liver density was significantly greater in NASH and HFD compared to Control and in NASH compared to HFD and HFr. b Liver triglyceride content
was significantly elevated in NASH- and HFD-fed rats compared to Control and HFr. Additionally, HFD-fed rats had significantly higher hepatic TG
levels compared to NASH-fed. Hepatic TG levels in rats fed HFr did not differ from those fed the Control-diet. c Hepatic cholesterol levels were sig-
nificantly elevated in NASH-fed animals compared to Control, HFD, and HFr. d The level of liver glycogen was significantly lower in both NASH and
HFD compared to Control and HFr. a Comparisons between groups: *Control vs. NASH; †Control vs. HFD; ¤NASH vs. HFD; #NASH vs. HFr. Statistical
significance: ****p < 0.0001; †p < 0.05; ¤¤¤¤p < 0.0001; ####p < 0.0001. b–d. Statistical significance: **p < 0.01; ****p < 0.0001. TG triglyceride, TC total
cholesterol, NASH NASH-diet, HFD high-fat diet, HFr high fructose diet. Results are shown as mean ± SEM

(Table  3). Circulating leptin levels were significantly (p < 0.01, Fig. 3d). Haptoglobin was significantly increased
increased in the HFD-fed rats compared to both Control- in plasma of NASH-fed rats compared to rats fed Con-
(p < 0.05), NASH- (p < 0.05) and HFr-fed rats (p < 0.01, trol (p  <  0.001), HFD (p  <  0.0001) and HFr (p  <  0.001,
Table 3). Fig.  3e). Concentrations of MCP-1 and TNF-alpha were
not elevated in adipose or intestinal tissue in any of the
Effects of diet on markers of systemic and tissue‑specific diet-groups compared to Control; however, HFr-fed rats
inflammation had significantly higher levels of adipose tissue MCP-1
Hepatic levels of MCP-1 were significantly increased in compared to both NASH- and HFD-fed rats (p  <  0.05,
rats on NASH (p < 0.01), HFD diet (p < 0.0001) and HFr Table 3). MCP-1 was below detection limit in plasma in all
diet (p < 0.01) compared to Control (Fig. 3c). Additionally, groups. Plasma TIMP-1 levels was significantly increased
hepatic TNF-α levels were higher in HFD fed rats com- in the NASH-fed rats, compared to all other groups (vs.
pared to Control- (p < 0.001), NASH (p < 0.0001) and HFr Control: p < 0.01; vs. HFD and HFr: p < 0.001, Fig. 3f ).
Jensen et al. Diabetol Metab Syndr (2018) 10:4 Page 7 of 13

Table 3  Metabolic and inflammatory effects in rats after 16 weeks on NASH-, HFD- and HFr-diet

Control NASH HFD HFr

Liver weight (g) 18.9 ± 3.4# 32.7 ± 7.9*†‡ 16.5 ± 2.5# ‡ 23.1 ± 4.4#†

b #†‡
Fasting plasma glucose (mM) 6.4 ± 0.2 7.2 ± 0.2* 7.5 ± 0.1* 7.16 ± 0.2*
Fasting plasma insulin (pM)b 560.6 ± 79.5 486.7 ± 93.7 603.0 ± 43.7 695.8 ± 83.7
Plasma β-hydroxybutyrate (µmol/L) 38.4 ± 4.2† 64.0 ± 5.3† 115.4 ± 16.8*#‡ 38.4 ± 2.8†
† † #‡
Plasma leptin (ng/mL) 3283.3 ± 1035.2 3803.0 ± 1341.1 9366.6 ± 2052.8* 2638.0 ± 808.5†
‡ ‡ ‡
Plasma TG (mmol/L) 1.6 ± 0.2 1.7 ± 0.1 1.5 ± 0.1 3.9 ± 0.5*#†
Plasma FFA (mmol/L) 0.3 ± 0.1 0.4 ± 0.03 0.4 ± 0.02 0.3 ± 0.02
Plasma Cholesterol (mmol/L) 2.1 ± 0.1 2.5 ± 0.1 2.1 ± 0.1 2.3 ± 0.2
Plasma HDL-C (mmol/L) 1.3 ± 0.1 1.0 ± 0.1 1.3 ± 0.1 1.1 ± 0.1
Jejunal MCP-1 (ng/mL)a 45.3 [36.3; 55.3] 58.1 [48.5; 69.6] 49.5 [40.1; 60.5] 56.4 [48.1; 65.5]
Jejunal TNF-α (ng/mL)a 179.9 [129.3; 247.7] 87.6 [61.7; 125.5] 171.1 [76.1; 384.6] 213.2 [112.4; 404.3]
Adipose tissue MCP-1 (ng/mL)a 14.8 [13.4; 16.5] 14.1 [11.9; 16.8]‡ 14.2 [11.3; 17.7]‡ 25.6 [15.2; 42.6]#†
a
Adipose tissue TNF-α (ng/mL) 15.4 [14.4; 16.4] 16.1 [14.3; 18.2] 15.4 [13.6; 18.2] 16.2 [14.6; 18.0]
TG triglycerides, FFA free fatty acids, HDL-C high-density-lipoprotein-cholesterol, MCP-1 monocyte-chemoattractant-protein-1, TNF-α tissue-necrosis-factor-alpha
Results are presented as mean ± SEM. Metabolic parameters denoted with superscript a, were analysed on log-transformed data (due to non-normal distribution), and
are consequently presented as geometric means with 95% confidence intervals. Superscripts (*,#,†,‡) indicate significant (p < 0.05) differences in readouts compared to
Control-, NASH-, HFD- and HFr-groups respectively. Sample sizes at week 16: n = 6–10/group. b Fasting plasma glucose and fasting plasma insulin levels were assessed
after 15 weeks

Discussion also significantly elevated only in HFD-fed rats, reflect-

The present study shows that feeding rats diets high in ing the increase in adipose tissue depot size within this
either dietary fat or dietary fructose results in distinctly group. Fasting hyperglycemia was present in all groups
different hepatic, metabolic and inflammatory profiles in at study termination, indicating disturbances in glucose
rats. When comparing the effect of dietary fat and fruc- metabolism, even though fasting insulin levels remained
tose, high-fat feeding more potently induce development similar between groups. These disturbances were further
of fatty liver and associated hepatic inflammation without corroborated by the increased levels of hepatic TG and
affecting the circulating lipid pool. In contrast, high-fruc- decreased/unaltered hepatic glycogen observed in both
tose feeding appear to have the most pronounced effects NASH-, HFD- and HFr-fed rats, indicative of selective
on the plasma lipid profile, while only subtle effects on insulin resistance [23].
the liver are observed. The combination of fat, fructose, Hepatic steatosis was rapidly induced in both NASH-
and cholesterol exacerbated and intensified overall effects and HFD-fed rats and was shown using qCT to progress
on the liver. The strength of this study is the direct com- in a time-dependent manner. In line with this, higher
parison of dietary fat (with very restricted amounts of levels of plasma β-hydroxybutyrate were observed in
carbohydrate) and dietary fructose (with very restricted HFD-fed rats, suggestive of increased β-oxidation due
amounts of fat) on parameters associated with NAFLD. to hepatic lipid overload. The role of hepatic β-oxidation
This enables a more detailed evaluation of the individual in the pathogenesis of NAFLD remains controversial,
roles of these macronutrients in disease progression. Fur- but several studies in human NASH-patients have found
thermore, the use of qCT allows the non-invasive assess- serum levels of β-hydroxybutyrate to be increased [24,
ment of NAFLD-progression throughout the study, a 25]. Although not significant, there was a trend towards
method which to our knowledge has not previously been an increase of this marker in NASH-fed rats as well. The
applied in rat studies comparing NAFLD progression development of hepatic steatosis in NASH- and HFD-fed
after administration of different diets. rats agrees well with recent studies in other rodent spe-
NAFLD in humans is often associated with obesity and cies suggesting fat and cholesterol to be important driv-
insulin resistance, and these metabolic derangements ers in the development of experimental NAFLD/NASH
were more closely reflected in the HFD-group. Accord- [14, 22]. The addition of dietary cholesterol (NASH-diet)
ingly, only HFD-fed rats became obese and even though caused the most detrimental liver changes, evident by
the cumulative energy intake in this group was transiently significant accumulation of both hepatic TG and choles-
higher than Control animals at week 8, differences in terol, pronounced decreases in liver density, and severe
energy intake did no longer account for the HFD-induced morphological alterations in liver histology. In contrast
obesity after 16  weeks. Circulating leptin levels were to what was observed in HFD- and HFr-fed rats, these
Jensen et al. Diabetol Metab Syndr (2018) 10:4 Page 8 of 13

Fig. 3  Markers of hepatic function, inflammation and fibrosis. a After 16 weeks ALAT-levels were significantly higher in NASH-fed rats compared to
Control-, HFD- and HFr-fed rats. b Plasma levels of ASAT were higher in the NASH-group compared to both Control and HFr-groups. c Rats on NASH,
HFD and HFr-diet had increased hepatic levels of MCP-1 compared to Control. d Hepatic TNF-α levels were significantly increased in the HFD fed
rats compared to Control. e Haptoglobin was significantly increased in plasma of NASH-fed rats compared to rats fed Control, HFD and HFr. f Plasma
TIMP-1 was significantly increased in the NASH-fed rats, compared to all other groups. Statistical significance: *p < 0.05; **p < 0.01; ***p < 0.001;
****p < 0.0001. Results are shown as mean ± SEM
Jensen et al. Diabetol Metab Syndr (2018) 10:4 Page 9 of 13

Fig. 4  Histological evaluation of liver sections from Control-, NASH-, HFD- and HFr-fed rats. Row 1: representative H&E-stains of normal liver from
a Control-fed rat, and hepatic steatosis in b NASH-fed rat, c HFD-fed rat and d HFr-fed rat. Hepatic steatosis in HFD-fed rats was almost exclusively
found to be microvesicular (c); in HFr-fed rats almost exclusively macrovesicular (d); while steatosis in NASH-fed rats represented an intermediate
between the two, with both macro- and microvesicular steatosis (b). Row 2: higher magnification of representative H&E stained liver sections. e
Liver morphology appeared normal in Control-fed rats, whereas inflammatory infiltrates were observed in livers from f NASH-fed rats, g HFD-fed
rats and h HFr-fed rats Row 3: Oil Red O stains of liver sections from j NASH-fed rats, k HFD-fed rats and l HFr-fed rats confirmed hepatic steatosis
observed in b–d. i Oil Red O staining of liver from Control-fed rat

changes were also accompanied by increased plasma lev- the more aggravated inflammatory histological response,
els of ALAT and ASAT, suggesting impaired liver func- supported by the significant increase in infiltrating mac-
tion. It has previously been observed in both mice [26, rophages observed in livers of NASH-fed rats compared
27] and humans [28] that the combined intake of dietary to both HFD and Control groups. Further studies are
fat and carbohydrates, or of fat and cholesterol [29] can needed to determine if these changes are caused exclu-
result in accelerated and more severe effects in the liver. sively by the cholesterol content, by a synergistic effect of
The majority of these changes could be driven by the fat and fructose, or by a combination of all.
amount of cholesterol in the NASH-diet (2%), as hepatic While fat and cholesterol both seem to contribute to
levels of cholesterol measured in NASH-fed rats were the development of NAFLD, previous studies in rats have
very high. Supporting this, two previous studies in mice shown that also high levels of dietary fructose are capable
have suggested that dietary cholesterol is important in of inducing hepatic steatosis, inflammation and increase
facilitating progression from simple steatosis to NASH oxidative stress markers within a relatively short time-
[30] and that free cholesterol loading can sensitize the span of only 2–5  weeks [32–34]. These are in contrast
liver to cytokine-mediated hepatocellular death, inflam- to our findings with the HFr-diet, which only induced
mation and oxidative stress [31]. This could account for subtle changes in the liver. While steatosis was observed
Jensen et al. Diabetol Metab Syndr (2018) 10:4 Page 10 of 13

Fig. 5  Macrophage infiltration and collagen deposition. A significant increase in macrophage infiltration was seen in NASH-fed rats, compared
to Control- and HFD.fed rats (a) and collagen deposition in NASH-fed rats trended towards being increased, although only significantly when
compared to HFD-rats (b). Representative CD68- (c) and Picro Weigert-stains (d) of liver from NASH-fed rat. Statistical significance: **p < 0.01;
***p < 0.001. Results in a and c are shown as mean ± SEM

histologically in some animals in the HFr-group the con- confirming the heterogeneously distributed steatosis
dition was not confirmed by increased levels of liver TG would be interesting in terms of evaluating and compar-
at the week 16 time point. The apparent discrepancy ing histology, biochemistry and imaging of the liver in
between histological and biochemical analyses within rodent models of NAFLD/NASH.
the HFr-group could result from sampling variation, a Dyslipidemia is one of the hallmarks of the metabolic
problem also commonly encountered in biopsy-guided syndrome and has been shown to be strongly associated
diagnosis of human NAFLD/NASH [35]. Biochemical with NAFLD [36]. Dyslipidemia associated with NAFLD
analyses of hepatic TG-content was performed on liver is typically characterized by elevated levels of circulat-
tissue sampled from the left lateral lobe and histologic ing TG and low-density-lipoprotein cholesterol (LDL-
evaluation of livers in the HFr-group (performed on both C), as well as decreased HDL-C levels [37]. Only rats
left lateral, right medial and the caudate lobe) did in fact, fed HFr-diet developed dyslipidemia, as defined by the
in some animals, show specific lobes to be more severely presence of hypertriglyceridemia. Previous studies have
affected than others, indicating that steatosis, at least also found that dietary fructose potently increase TG in
in rats fed fructose-enriched diets, may not be homog- plasma within a relatively short timeframe in both rats
enously distributed throughout the liver. Further studies, and mice [34, 38]. In rats, this has been suggested to be
Jensen et al. Diabetol Metab Syndr (2018) 10:4 Page 11 of 13

caused in part by the ability of fructose to both increase pathology. Despite increased hepatic MCP-1 levels in all
hepatic very-low-density-lipoprotein (VLDL)-TG secre- groups compared to Control, MCP-1 plasma levels were
tion and decrease VLDL-TG clearance from the circula- below detection limit.
tion, even in the absence of hyperinsulinemia [39, 40]. It has been hypothesized that fructose and fat may
Mechanistically, the increase in VLDL-TG secretion has induce their inflammatory action in the liver not only
been proposed to result from a combined effect of fruc- by dietary overload, but also by stimulation of bacterial
tose-induced hepatic stress responses [40], and activation overgrowth in the intestine, increasing intestinal perme-
of hepatic enzymes involved in the de novo synthesis of ability and thereby facilitating translocation of endotox-
fatty acids (de novo lipogenesis) [41]. ins across the intestinal barrier that are then transported
In contrast to the HFr-diet, neither the HFD- nor to the liver [46, 47]. In the present study, we were not
the NASH-diet resulted in increased plasma TG, even able to detect increased intestinal levels of TNF-alpha
though high-fat diets have previously been shown to or MCP-1 in any diet group. Moreover, we did not find
induce hypertriglyceridemia in rats [42]. However, our increased levels of MCP-1 and TNF-alpha in visceral
findings agree with a recent study in mice investigat- adipose tissue in any of the groups, even though inflam-
ing the contribution of dietary fat and cholesterol to mation particularly in the adipose tissue compartment
NAFLD-development, which showed no effect of dietary is strongly associated with NAFLD/NASH in humans
fat and cholesterol on circulating TG or FFA levels, inde- [48]. Notably, the cytokine analyses in this study were
pendently of whether these macronutrients were fed sep- performed on epididymal fat depots, which may not
arately or in combination [29]. One explanation could be adequately represent visceral adipose tissue depots in
that accelerated fat accumulation in the liver imposes a humans [49].
significant strain on metabolic pathways responsible for Although fibrosis is not a prerequisite for the NASH
the release of lipids from the liver into plasma. Accord- diagnosis, it often accompanies the three recognized
ingly, it has been shown both in vitro and in vivo that high diagnostic hallmarks (steatosis, lobular inflammation
levels of fat and cholesterol in the liver result in increased and hepatocyte ballooning), and can be used as a grading
stress in the endoplasmic reticulum, limiting secre- tool for evaluation of NASH severity [50]. In the present
tion of apolipoprotein B100 (an essential component in study, neither cellular ballooning nor fibrosis was present
the assembly of VLDL-particles) and thereby inhibits in any of the groups. However, increased levels of plasma
hepatic export of TG [29, 43]. The inflammatory profile haptoglobin and TIMP-1 were observed in NASH-fed
was clearly distinguishable between groups in the present rats and are considered predictors of these pathological
study. Systemic inflammation was present only in NASH- liver changes, indicating that the NASH diet might be
fed rats with higher circulating levels of haptoglobin. promising in terms of modelling more advanced stages
We also found increased levels of the pro-inflammatory of NASH within a relatively short time-frame. The ten-
cytokines MCP-1 and TNF-alpha in liver tissue homoge- dency towards increased collagen deposition in NASH-
nates in NASH-, HFD- and HFr- and in HFD-fed rats, fed rats further supports this. The inability of the diets
respectively, suggesting progression of hepatic steatosis used in this study to induce liver fibrosis in rats after only
towards NASH with the increased inability of the liver 16  weeks is not entirely surprising. Successful dietary
to cope with fat infiltration. The inflammatory infiltrates induction of fibrosis in rodent models of NASH has so
observed histologically in livers of HFr- and NASH-fed far mainly been associated with either genetic manipula-
rats at week 16, were reflected only in significantly higher tion, forced overfeeding or diets low in methionine and
levels of hepatic MCP-1, not TNF-alpha. Compared to choline, [51, 52], except for the guinea pig that has been
the other diet-groups, NASH-fed rats might be expected shown to develop NASH after 16 weeks of high fat feed-
to have the highest levels of liver inflammatory cytokines, ing [22]. However, mild hepatic fibrosis in addition to a
due to the higher macrophage count, increased liver NASH-like phenotype has also been induced in C57BL/6
enzymes, and the more severe liver pathology. However, mice on a high-fat diet, but only after 50  weeks [53]. In
hepatic macrophages are a remarkably heterogeneous a recent study, using levels of dietary cholesterol compa-
population of immune cells, whose effector function rable to the level in the NASH-diet in our study, devel-
depends on origin, underlying pathogenesis and disease opment of fibrosis in Sprague–Dawley rats was observed
stage, which can result in high diversity in inflammatory after only 9  weeks. [54]. However, the diets used by
cytokines released and cell surface markers [44, 45]. This Ichimura and colleagues included relatively high levels of
could explain why the NASH-fed group does not have dietary cholate (2%), which is known to potently induce
the highest hepatic MCP-1 and TNF-alpha levels despite hepatotoxicity and upregulate specific genes related to
the higher macrophage count and the more severe liver hepatic fibrosis [55].
Jensen et al. Diabetol Metab Syndr (2018) 10:4 Page 12 of 13

Conclusion Received: 19 October 2017 Accepted: 16 January 2018

Dietary fat seems to primarily drive NAFLD development
in Sprague–Dawley rats with potent effects on hepatic
fat accumulation and inflammation, whereas dietary
fructose primarily affects circulating lipids with much References
more subtle effects on the liver. Combining fat, fructose 1. Katsiki N, Mikhailidis DP, Mantzoros CS. Non-alcoholic fatty liver disease
and dyslipidemia: an update. Metabolism. 2016;65:1109–23.
and cholesterol accelerates NAFLD development and
2. Bellentani S, Scaglioni F, Marino M, Bedogni G. Epidemiology of non-
increase the overall severity of changes observed in the alcoholic fatty liver disease. Dig Dis. 2010;28:155–61.
liver. 3. Hojland Ipsen D, Tveden-Nyborg P, Lykkesfeldt J. Normal weight dyslipi-
demia: is it all about the liver? Obesity (Silver Spring). 2016;24:556–67.
4. Cheung O, Sanyal AJ. Recent advances in nonalcoholic fatty liver disease.
Curr Opin Gastroenterol. 2010;26:202–8.
Abbreviations
5. Hebbard L, George J. Animal models of nonalcoholic fatty liver disease.
ALAT: alanine aminotransferase; ASAT: aspartate aminotransferase; ELISA:
Nat Rev Gastroenterol Hepatol. 2011;8:35–44.
enzyme-linked immunosorbent assay; FFA: free fatty acid; H&E: Haematoxylin
6. Crespo J, Cayon A, Fernandez-Gil P, Hernandez-Guerra M, Mayorga
and Eosin; HDL-C: high-density lipoprotein cholesterol; HFD: high-fat diet; HFr:
M, Dominguez-Diez A, Fernandez-Escalante JC, Pons-Romero F.
high-fructose diet; HU: Hounsfield unit; IHC: immunohistochemistry; LDL-C:
Gene expression of tumor necrosis factor alpha and TNF-receptors,
low-density lipoprotein cholesterol; MCP-1: monocyte-chemoattractant-
p55 and p75, in nonalcoholic steatohepatitis patients. Hepatology.
protein 1; NAFLD: non-alcoholic fatty liver disease; NASH: non-alcoholic
2001;34:1158–63.
steatohepatitis; qCT: quantitative computed tomography; qMR: quantitative
7. Greco D, Kotronen A, Westerbacka J, Puig O, Arkkila P, Kiviluoto T, Laitinen
magnetic resonance; TC: total cholesterol; TG: triglyceride; TIMP-1: tissue-
S, Kolak M, Fisher RM, Hamsten A, et al. Gene expression in human
inhibitor of metalloproteinase-1; TNF-α: tumor-necrosis-factor alpha; VLDL:
NAFLD. Am J Physiol Gastrointest Liver Physiol. 2008;294:G1281–7.
very-low-density lipoprotein.
8. Hui JM, Hodge A, Farrell GC, Kench JG, Kriketos A, George J. Beyond
insulin resistance in NASH: TNF-alpha or adiponectin? Hepatology.
Authors’ contributions
2004;40:46–54.
The study was designed by VSJ, CF, HH, JL, EMW, CI, JD and HN. The experi-
9. Haukeland JW, Damas JK, Konopski Z, Loberg EM, Haaland T, Goverud I,
ments were carried out by VSJ, JD and HN. Initial data analysis was performed
Torjesen PA, Birkeland K, Bjoro K, Aukrust P. Systemic inflammation in non-
by VSJ followed by data interpretation by all authors. The draft manuscript was
alcoholic fatty liver disease is characterized by elevated levels of CCL2. J
drafted by VSJ and subsequently edited by all authors. All authors read and
Hepatol. 2006;44:1167–74.
approved the final manuscript.
10. Kamada Y, Akita M, Takeda Y, Yamada S, Fujii H, Sawai Y, Doi Y, Asazawa H,
Nakayama K, Mizutani K, et al. Serum fucosylated haptoglobin as a novel
Author details
1 diagnostic biomarker for predicting hepatocyte ballooning and nonalco-
 Department of Veterinary and Animal Science, Faculty of Health and Medi-
holic steatohepatitis. PLoS ONE. 2013;8:e66328.
cal Sciences, University of Copenhagen, Ridebanevej 9, 1870 Frederiksberg,
11. Abdelaziz R, Elbasel M, Esmat S, Essam K, Abdelaaty S. Tissue inhibitors of
Denmark. 2 Insulin Pharmacology, Novo Nordisk A/S, Novo Nordisk Park 1,
metalloproteinase-1 and 2 and obesity related non-alcoholic fatty liver
2760 Maaloev, Denmark. 3 Histology and Imaging, Novo Nordisk A/S, Novo
disease: is there a relationship. Digestion. 2015;92:130–7.
Nordisk Park 1, 2760 Maaloev, Denmark. 4 Obesity and Diabetes Pharmacology,
12. Alkhouri N, Dixon LJ, Feldstein AE. Lipotoxicity in nonalcoholic fatty liver
Novo Nordisk A/S, Novo Nordisk Park 1, 2760 Maaloev, Denmark.
disease: not all lipids are created equal. Expert Rev Gastroenterol Hepatol.
2009;3:445–51.
Acknowledgements
13. Puri P, Baillie RA, Wiest MM, Mirshahi F, Choudhury J, Cheung O, Sargeant
We wish to thank Susanne Juul Rasmussen, Sara Louise Riisberg, Marianne
C, Contos MJ, Sanyal AJ. A lipidomic analysis of nonalcoholic fatty liver
Schiødt, Anita Svendsen and Lotte Gottlieb Sørensen for providing excellent
disease. Hepatology. 2007;46:1081–90.
technical assistance with biomarker analyses and Hyeran Yang for guidance
14. Mells JE, Fu PP, Kumar P, Smith T, Karpen SJ, Anania FA. Saturated fat and
on diet selection.
cholesterol are critical to inducing murine metabolic syndrome with
robust nonalcoholic steatohepatitis. J Nutr Biochem. 2015;26:285–92.
Competing interests
15. Vos MB, Kimmons JE, Gillespie C, Welsh J, Blanck HM. Dietary fructose
The authors declare that they have no competing interests. The authors VSJ,
consumption among US children and adults: the Third National Health
CF, HH, EMW, CI, JD and HN are employees at Novo Nordisk.
and Nutrition Examination Survey. Medscape J Med. 2008;10:160.
16. Lim JS, Mietus-Snyder M, Valente A, Schwarz JM, Lustig RH. The role of
Availability of data and materials
fructose in the pathogenesis of NAFLD and the metabolic syndrome. Nat
The datasets used and/or analysed during the current study are available from
Rev Gastroenterol Hepatol. 2010;7:251–64.
the corresponding author on reasonable request.
17. Lau JK, Zhang X, Yu J. Animal models of non-alcoholic fatty liver disease:
current perspectives and recent advances. J Pathol. 2017;241:36–44.
Consent for publication
18. Koteish A, Diehl AM. Animal models of steatosis. Semin Liver Dis.
Not applicable.
2001;21:89–104.
19. Svegliati-Baroni G, Candelaresi C, Saccomanno S, Ferretti G, Bachetti T,
Ethics approval and consent to participate
Marzioni M, De Minicis S, Nobili L, Salzano R, Omenetti A, et al. A model of
The study was approved by the Danish Animal Experiments Inspectorate
insulin resistance and nonalcoholic steatohepatitis in rats: role of peroxi-
under the Ministry of Food, Agriculture and Fisheries in accordance with
some proliferator-activated receptor-alpha and n-3 polyunsaturated fatty
European Union directive 2010/63/EU.
acid treatment on liver injury. Am J Pathol. 2006;169:846–60.
20. Hvid H, Blouin MJ, Birman E, Damgaard J, Poulsen F, Fels JJ, Fledelius
Funding
C, Hansen BF, Pollak M. Treatment with insulin analog X10 and IGF-1
This study was funded by Novo Nordisk A/S.
increases growth of colon cancer allografts. PLoS ONE. 2013;8:e79710.
21. Metzinger MN, Miramontes B, Zhou P, Liu Y, Chapman S, Sun L, Sasser
Publisher’s Note TA, Duffield GE, Stack MS, Leevy WM. Correlation of X-ray computed
Springer Nature remains neutral with regard to jurisdictional claims in pub- tomography with quantitative nuclear magnetic resonance methods for
lished maps and institutional affiliations. pre-clinical measurement of adipose and lean tissues in living mice. Sen-
sors (Basel). 2014;14:18526–42.
Jensen et al. Diabetol Metab Syndr (2018) 10:4 Page 13 of 13

22. Ipsen DH, Tveden-Nyborg P, Rolin B, Rakipovski G, Beck M, Mortensen LW, 39. Zavaroni I, Chen YD, Reaven GM. Studies of the mechanism of fructose-
Faerk L, Heegaard PM, Moller P, Lykkesfeldt J. High-fat but not sucrose induced hypertriglyceridemia in the rat. Metabolism. 1982;31:1077–83.
intake is essential for induction of dyslipidemia and non-alcoholic steato- 40. Kelley GL, Allan G, Azhar S. High dietary fructose induces a hepatic stress
hepatitis in guinea pigs. Nutr Metab (Lond). 2016;13:51. response resulting in cholesterol and lipid dysregulation. Endocrinology.
23. Brown MS, Goldstein JL. Selective versus total insulin resistance: a patho- 2004;145:548–55.
genic paradox. Cell Metab. 2008;7:95–6. 41. Park OJ, Cesar D, Faix D, Wu K, Shackleton CH, Hellerstein MK. Mecha-
24. Sanyal AJ, Campbell-Sargent C, Mirshahi F, Rizzo WB, Contos MJ, Sterling nisms of fructose-induced hypertriglyceridaemia in the rat. Activation
RK, Luketic VA, Shiffman ML, Clore JN. Nonalcoholic steatohepatitis: asso- of hepatic pyruvate dehydrogenase through inhibition of pyruvate
ciation of insulin resistance and mitochondrial abnormalities. Gastroen- dehydrogenase kinase. Biochem J. 1992;282(Pt 3):753–7.
terology. 2001;120:1183–92. 42. Safwat GM, Pisano S, D’Amore E, Borioni G, Napolitano M, Kamal AA, Bal-
25. Bugianesi E, Gastaldelli A, Vanni E, Gambino R, Cassader M, Baldi S, Ponti lanti P, Botham KM, Bravo E. Induction of non-alcoholic fatty liver disease
V, Pagano G, Ferrannini E, Rizzetto M. Insulin resistance in non-diabetic and insulin resistance by feeding a high-fat diet in rats: does coenzyme Q
patients with non-alcoholic fatty liver disease: sites and mechanisms. monomethyl ether have a modulatory effect? Nutrition. 2009;25:1157–68.
Diabetologia. 2005;48:634–42. 43. Ota T, Gayet C, Ginsberg HN. Inhibition of apolipoprotein B100 secretion
26. Ishimoto T, Lanaspa MA, Rivard CJ, Roncal-Jimenez CA, Orlicky DJ, by lipid-induced hepatic endoplasmic reticulum stress in rodents. J Clin
Cicerchi C, McMahan RH, Abdelmalek MF, Rosen HR, Jackman MR, et al. Invest. 2008;118:316–32.
High-fat and high-sucrose (western) diet induces steatohepatitis that is 44. Li H, You H, Fan X, Jia J. Hepatic macrophages in liver fibrosis: patho-
dependent on fructokinase. Hepatology. 2013;58:1632–43. genesis and potential therapeutic targets. BMJ Open Gastroenterol.
27. Luo Y, Burrington CM, Graff EC, Zhang J, Judd RL, Suksaranjit P, Kaew- 2016;3:e000079.
poowat Q, Davenport SK, O’Neill AM, Greene MW. Metabolic pheno- 45. Tacke F, Zimmermann HW. Macrophage heterogeneity in liver injury and
type and adipose and liver features in a high-fat western diet-induced fibrosis. J Hepatol. 2014;60:1090–6.
mouse model of obesity-linked NAFLD. Am J Physiol Endocrinol Metab. 46. Bergheim I, Weber S, Vos M, Kramer S, Volynets V, Kaserouni S, McClain
2016;310:E418–39. CJ, Bischoff SC. Antibiotics protect against fructose-induced hepatic lipid
28. Sobrecases H, Le KA, Bortolotti M, Schneiter P, Ith M, Kreis R, Boesch C, accumulation in mice: role of endotoxin. J Hepatol. 2008;48:983–92.
Tappy L. Effects of short-term overfeeding with fructose, fat and fructose 47. Cani PD, Amar J, Iglesias MA, Poggi M, Knauf C, Bastelica D, Neyrinck AM,
plus fat on plasma and hepatic lipids in healthy men. Diabetes Metab. Fava F, Tuohy KM, Chabo C, et al. Metabolic endotoxemia initiates obesity
2010;36:244–6. and insulin resistance. Diabetes. 2007;56:1761–72.
29. Savard C, Tartaglione EV, Kuver R, Haigh WG, Farrell GC, Subramanian S, 48. Shoelson SE, Lee J, Goldfine AB. Inflammation and insulin resistance. J
Chait A, Yeh MM, Quinn LS, Ioannou GN. Synergistic interaction of dietary Clin Invest. 2006;116:1793–801.
cholesterol and dietary fat in inducing experimental steatohepatitis. 49. Chusyd DE, Wang D, Huffman DM, Nagy TR. Relationships between
Hepatology. 2013;57:81–92. rodent white adipose fat pads and human white adipose fat depots.
30. Wouters K, van Gorp PJ, Bieghs V, Gijbels MJ, Duimel H, Lutjohann D, Front Nutr. 2016;3:10.
Kerksiek A, van Kruchten R, Maeda N, Staels B, et al. Dietary cholesterol, 50. Rinella ME. Nonalcoholic fatty liver disease: a systematic review. JAMA.
rather than liver steatosis, leads to hepatic inflammation in hyperlipi- 2015;313:2263–73.
demic mouse models of nonalcoholic steatohepatitis. Hepatology. 51. Larter CZ, Yeh MM. Animal models of NASH: getting both pathology and
2008;48:474–86. metabolic context right. J Gastroenterol Hepatol. 2008;23:1635–48.
31. Mari M, Caballero F, Colell A, Morales A, Caballeria J, Fernandez A, Enrich 52. Lieber CS, Leo MA, Mak KM, Xu Y, Cao Q, Ren C, Ponomarenko A, DeCarli
C, Fernandez-Checa JC, Garcia-Ruiz C. Mitochondrial free cholesterol LM. Model of nonalcoholic steatohepatitis. Am J Clin Nutr. 2004;79:502–9.
loading sensitizes to TNF- and Fas-mediated steatohepatitis. Cell Metab. 53. Ito M, Suzuki J, Tsujioka S, Sasaki M, Gomori A, Shirakura T, Hirose H, Ito
2006;4:185–98. M, Ishihara A, Iwaasa H, Kanatani A. Longitudinal analysis of murine stea-
32. Ackerman Z, Oron-Herman M, Grozovski M, Rosenthal T, Pappo O, tohepatitis model induced by chronic exposure to high-fat diet. Hepatol
Link G, Sela BA. Fructose-induced fatty liver disease: hepatic effects Res. 2007;37:50–7.
of blood pressure and plasma triglyceride reduction. Hypertension. 54. Ichimura M, Kawase M, Masuzumi M, Sakaki M, Nagata Y, Tanaka K, Suruga
2005;45:1012–8. K, Tamaru S, Kato S, Tsuneyama K, Omagari K. High-fat and high-choles-
33. Armutcu F, Coskun O, Gurel A, Kanter M, Can M, Ucar F, Unalacak M. terol diet rapidly induces non-alcoholic steatohepatitis with advanced
Thymosin alpha 1 attenuates lipid peroxidation and improves fructose- fibrosis in Sprague–Dawley rats. Hepatol Res. 2015;45:458–69.
induced steatohepatitis in rats. Clin Biochem. 2005;38:540–7. 55. Vergnes L, Phan J, Strauss M, Tafuri S, Reue K. Cholesterol and cholate
34. Kawasaki T, Igarashi K, Koeda T, Sugimoto K, Nakagawa K, Hayashi S, components of an atherogenic diet induce distinct stages of hepatic
Yamaji R, Inui H, Fukusato T, Yamanouchi T. Rats fed fructose-enriched inflammatory gene expression. J Biol Chem. 2003;278:42774–84.
diets have characteristics of nonalcoholic hepatic steatosis. J Nutr.
2009;139:2067–71.
35. Merat S, Sotoudehmanesh R, Nouraie M, Peikan-Heirati M, Sepanlou SG,
Malekzadeh R, Sotoudeh M. Sampling error in histopathology findings of
nonalcoholic fatty liver disease: a post mortem liver histology study. Arch
Iran Med. 2012;15:418–21.
36. Zhang QQ, Lu LG. Nonalcoholic fatty liver disease: dyslipidemia, risk for Submit your next manuscript to BioMed Central
cardiovascular complications, and treatment strategy. J Clin Transl Hepa-
tol. 2015;3:78–84. and we will help you at every step:
37. Chatrath H, Vuppalanchi R, Chalasani N. Dyslipidemia in patients with • We accept pre-submission inquiries
nonalcoholic fatty liver disease. Semin Liver Dis. 2012;32:22–9.
38. Spruss A, Kanuri G, Wagnerberger S, Haub S, Bischoff SC, Bergheim I. • Our selector tool helps you to find the most relevant journal
Toll-like receptor 4 is involved in the development of fructose-induced • We provide round the clock customer support
hepatic steatosis in mice. Hepatology. 2009;50:1094–104. • Convenient online submission
• Thorough peer review
• Inclusion in PubMed and all major indexing services
• Maximum visibility for your research

Submit your manuscript at

www.biomedcentral.com/submit