Life Sciences 82 (2008) 983–990

Contents lists available at ScienceDirect

Life Sciences
j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / l i f e s c i e

The establishment of a novel non-alcoholic steatohepatitis model accompanied with

obesity and insulin resistance in mice
Wei-Na Cong, Rong-Ya Tao, Jin-Ying Tian, Geng-Tao Liu, Fei Ye ⁎
Institute of Materia Medica, Chinese Academy of Medical Sciences & Peking Union Medical College, Beijing 100050, China

a r t i c l e i n f o a b s t r a c t

Article history: Non-alcoholic steatohepatitis (NASH) is a hepatic manifestation of the metabolic syndrome that can progress
Received 22 September 2007 to liver cirrhosis. The major aim of this study was to establish a novel NASH mouse model accompanied by
Accepted 26 January 2008 obesity and insulin resistance, then explore the molecular mechanisms of NASH and evaluate the effects of
both the peroxisome proliferator-activated receptor α (PPARα) agonist fenofibrate and the PPARγ agonist
Keywords: rosiglitazone in this established NASH model. The novel model was induced in C57BL/6 mice by 23 weeks of ad
Dyslipidemia libitum feeding of a modified high-fat diet (mHFD), with lower methinione and choline and higher fat content.
Fenofibrate
In comparison to the controls, the model animals developed pronounced obesity, dyslipidemia and insulin
Insulin resistance
resistance. Marked liver lesions characterized by severe steatosis, inflammation, fibrosis, increased hepatic
Non-alcoholic steatohepatitis
Obesity triglyceride content, and elevated serum alanine aminotransferase (ALT) levels were observed in the models.
Rosiglitazone In this novel model, treatment with fenofibrate or rosiglitazone significantly improved insulin sensitivity and
corrected dyslipidemia; however, fenofibrate was more effective than rosiglitazone in improving hepatic
morphology and ALT levels. Further study showed that long-term feeding of mHFD significantly increased
expression of mRNA for hepatic PPARγ, adipose fatty acid binding protein (ap2) and CD36 and suppressed
expression of mRNA for hepatic PPARα and carnitine palmitoyl transferase-1a (CPT-1a). These results showed
the successful establishment of the combined NASH and obese-insulin resistance mouse model. Additionally,
aberrant expressions of hepatic PPARα and PPARγ may play a major role in the pathogenesis of NASH by
affecting hepatic lipogenesis and fatty acid oxidation in this novel model.
© 2008 Elsevier Inc. All rights reserved.

Introduction human steatohepatitis; however, their model hardly developed insulin

resistance and obesity. Some other animal models involving ad libitum
Non-alcoholic fatty liver disease (NAFLD) is defined as a range of liver feeding of the general high-fat diet usually developed obesity and
abnormalities including simple steatosis, non-alcoholic steatohepatitis insulin resistance, but could not effectively develop noticeable
(NASH) and liver cirrhosis (Neuschwander-Tetr and Caldwell, 2003). steatohepatitis (Collins et al., 2004). All these led to our consideration
NAFLD is prevalent in obese individuals and in patients with type 2 that a modified diet with higher levels of fat, but lower levels of
diabetes (Angulo, 2002) and is regarded as a hepatic manifestation of methionine and choline may be able to induce the combined features
metabolic syndrome. As a crucial stage of NAFLD, NASH is characterized of both NASH and its associated metabolic disorders.
by steatosis, inflammation and fibrosis. The pathogenesis of NASH is Both agonists of peroxisome proliferator-activated receptor α
associated with disorders in energy metabolism including obesity, (PPARα) and peroxisome proliferator-activated receptor γ (PPARγ) are
insulin resistance and dyslipidemia (Bugianesi et al., 2005). The precise clinically widely used as lipid-lowering drugs and insulin sensitizers,
mechanisms leading to NASH are still unclear. respectively. In animal experiments, the PPARα agonist ameliorated
The study of the pathophysiologic process and molecular mechan- steatohepatitis and fibrosis induced by MCD diet (Ip et al., 2003, 2004);
isms of NASH is limited by the lack of appropriate animal models that and the PPARγ agonist ameliorated insulin resistance and dyslipidemia
can depict the combined features of steatohepatitis and the metabolic (Yang et al., 2004) induced by high-fat diet. The therapeutic effects of
abnormalities associated with NASH. Rinella and Green (2004) these two agonists on metabolic syndrome-associated steatohepatitis
reported that the methionine and choline-deficient (MCD) diet- that is accompanied by obesity, insulin resistance and dyslipidemia, are
induced model reproduced most of the histological features of rarely reported. The present study therefore sought to establish an
effective rodent model of NASH that depicts the features of steatohe-
patitis as well as obesity and insulin resistance observed frequently in
⁎ Corresponding author. Institute of Materia Medica, Chinese Academy of Medical
Sciences & Peking Union Medical College, 1, Xian Nong Tan Street, Beijing 100050, China.
NASH patients, and to further investigate the effects of agonists of PPARα
Tel.: +86 10 83150495; fax: +86 10 63017757. (fenofibrate) and PPARγ (rosiglitazone) in the treatment of NASH in this
E-mail address: yefei@imm.ac.cn (F. Ye). model.

0024-3205/$ – see front matter © 2008 Elsevier Inc. All rights reserved.
doi:10.1016/j.lfs.2008.01.022
984 W.-N. Cong et al. / Life Sciences 82 (2008) 983–990

Materials and methods Evaluation of insulin sensitivity

Reagents The insulin sensitivity was estimated by insulin tolerance test (ITT),
oral glucose tolerance test (OGTT) in vivo and the homeostasis model
Commercial kits for triglyceride (TG), total cholesterol (TC) and assessment of insulin resistance (HOMA-IR) index. Food was removed
alanine aminotransferase (ALT) were obtained from the Jian Cheng for 2 h before the tests. Insulin (0.4 IU/kg bodyweight) was
Biotechnology Company (Nanjing, China). The non-esterified fatty administered by subcutaneous injection, and blood samples were
acid (NEFA) kit was purchased from Daiichi Pure Chemicals (Japan). taken from tail vein droplets at time points of 0, 30, 60, and 120 min
The insulin radioimmunoassay kit was obtained from the Northern after insulin loading. The blood glucose concentration was immedi-
Bioengineering Institute (Beijing, China). The Trizol reagent and the ately determined. The area under the glucose–time curve (AUC) was
Superscript First-Strand Synthesis System were obtained from also calculated. In the OGTT, glucose load (2 g/kg bodyweight) was
Invitrogen Corp (CA, USA). The primary rabbit polyclonal antibodies given by gavage, and the blood samples were collected and de-
of mouse PPARα (No.sc-9000) and mouse PPARγ (No.sc-7196) were termined as above. HOMA-IR index was calculated by the formula
purchased from Santa Cruz Biotechnology (CA, USA). (HOMA-IR = FINS × FBG/22.5) as reported previously (FBG: fasting
Fenofibrate and rosiglitazone were obtained from Laboratoires blood glucose, FINS: fasting insulin) (Matthews et al., 1985).
Fournier S.A. (Chenove, France) and GlaxoSmithKline (USA), respec-
tively. Insulin (Humulin R) was purchased from Lilly (France). All other Histological analysis of liver tissues
reagents used in the study were of analytical grade.
Liver tissues were fixed in formaldehyde saline (4%) solution, and
Animals processed on slides for hematoxylin-eosin (H&E) and masson
trichrome staining. Histological changes were semi-quantitated by
Four-week-old male C57BL/6 mice were obtained from the Animal two independent pathologists according to the degrees of the macro
Center of the Institute of Laboratory Animal Sciences, Chinese and micro-vacuolar steatosis and the corresponding zonal distribu-
Academy of Medical Sciences & Peking Union Medical College. Mice tion, lobular inflammation and hepatic fibrosis (De Oliveira et al.,
were housed 4 per cage and maintained under standardized conditions 2006). Two independent pathologists, who were blinded to the
of temperature (21–22 °C) and humidity (40–60%), with light from treatment status, computed the results of histological changes and
0600 h to 1800 h. One group of mice used for control (Con) was allowed calculated the overall steatohepatitis scores.
to eat normal rodent chow. The rest of the mice were fed with a
modified high-fat diet (mHFD) ad libitum for 23 weeks. The normal Reverse transcription polymerase chain reaction (RT-PCR) analysis of
rodent chow contained (in energy %) 12% fat, 62% carbohydrate, and hepatic genes related to fatty acid metabolism
26% protein, with a total energy content of 12.6 kJ/g. The mHFD was
formulated to balance micronutrient content on a per calorie basis and Total RNA was isolated from liver with Trizol and reverse-
contained 60% fat, 14% protein, and 26% carbohydrate, with total transcribed by two-step method with the SuperScript First-Strand
energy content of 21.0 kJ/g. The mHFD contained much less choline Synthesis System. The resulting single-stranded cDNA (2 μl) was
bicitrate (0.6 g/kg) and DL-methionine (1.5 g/kg) and more lard than the denatured at 94 °C for 5 min and, after the addition of the polymerase,
general high-fat diet (choline bicitrate, 2 g/kg; DL-methionine, 3 g/kg). subjected to 28 cycles of amplification, each consisting of 30 s at 94 °C,
Fatty acid composition of the fats (mainly from lard) in mHFD was 36% 1 min at 58 °C, and 1 min at 72 °C, with a 2-minute final extension at
saturated fatty acids, 45% monounsaturated fatty acids, and 19% 72 °C during the last cycle. Each PCR reaction mixture (50 μl)
polyunsaturated fatty acids (PUFA). contained the cDNA template, 1 μM of the primers, 200 μM of dNTPs,
After 23-week feeding, the mice fed with mHFD were randomly 1.5 mM MgCl2 and 1.25 U Platinum Taq polymerase (Invitrogen Corp.,
divided into three groups, the model group (Model), rosiglitazone USA). The expressions of hepatic genes including sterol regulatory
group (Rosi) and fenofibrate group (Feno), of 10 mice per group. element binding protein-1c (SREBP-1c), acetyl-CoA carboxylase α
Fenofibrate (150 mg/kg bodyweight/day) and rosiglitazone (15 mg/kg (ACCα), fatty acid synthase (FAS), PPARγ, CD36, adipose fatty acid
bodyweight/day) were administrated by gavage for the following binding protein (ap2), PPARα, and carnitine palmitoyl transferase-1a
8 weeks. The model group and the control group received distilled (CPT-1a) were analyzed. Four to five samples were chosen randomly
water. The body weights and food uptake were recorded weekly. All
mice were sacrificed by decapitation after 4-hour food deprivation.
Blood samples were collected for the assays of glucose, insulin, lipid Table 1
Sequences of the primers used in the PCR measurements
profiles and alanine aminotransferase. Livers were rapidly dissected
and part of each liver was cut and fixed in formaldehyde saline (4%) Gene Sense Sequence (5V to 3V) GenBank no./ref.
solution for histological analysis; the rest was snap frozen in liquid
PPARα Forward CGGGAAAGACCAGCAACAAC NM011144
nitrogen, and stored at −70 °C until use. White adipose tissues were Reverse ATAGCAGCCACAAACAGGGA
carefully removed and used to determine the adiposity index, PPARγ Forward GCAAGACATAGACAAAACACCAGTGTGA NM011146
calculated as white adipose tissue weight (g) / body weight (g) × 100. Reverse AGCAACCATTGGGTCAGCTCTTGTGA
All animal experiments were approved by the Ethics Committee of SREBP-1c Forward CACTTCTGGAGACATCGCAAAC NM011480
Reverse TGGTAGACAACAGCCGCATC
Laboratory Animals of Beijing Municipality.
ACCα Forward AGGAGGACCGCATTTATCGAC NM133360
Reverse TGACCGTGGGCACAAAGTT
Biochemical assays FAS Forward CTGCGGAAACTTCAGGAAATG NM007988
Reverse GGTTCGGAATGCTATCCAGG
Serum was separated by centrifugation at 4 °C and analyzed CD36 Forward TGTACCTGGGAGTTGGCGAG NM007643
Reverse CTGCTGTTCTTTGCCACGTC
immediately or stored at −20 °C. Serum TG, TC, NEFA and ALT levels ap2 Forward TGGAAGACAGCTCCTCCTCG NM024406
were determined by spectrophotometry. Blood glucose levels were Reverse CCGCCATCTAGGGTTATGATG
determined with a glucose analyzer (EKF Diagnostic Company, Germany). CPT-1a Forward TCCACCCTGAGGCATCTATT NM013495
Insulin levels were measured by radioimmunoassay. Hepatic triglyceride Reverse ATGACCTCCTGGCATTCTCC
β-actin Forward CCCATC TACGAGG GCTAT NM007393
content was measured by spectrophotometry and expressed as TG mg/g
Reverse TGGCC AGTAA TGTCC AGG
wet liver weight according to published methods (Harrity et al., 2006).
 W.-N. Cong et al. / Life Sciences 82 (2008) 983–990 985

on a 1% agarose gel. The genes and their forward and reverse primers
are listed in Table 1.

Immunoblotting of hepatic PPARα and PPARγ protein

Hepatic nuclear protein was extracted from mouse liver as described

previously by Neish et al. (1995). Aliquots of nuclear protein (30 μg) were
subjected to 10% SDS-PAGE. Immunoblot analysis was performed with
enhanced chemiluminescence (ECL) reagent and detected by a gel
analysis system (Flurochem 5500, Alpha Innotech, USA).

Statistical analyses

All data are expressed as means ± standard error of the mean

(SEM). The one-way ANOVA analysis (SPSS version 10.0, USA) was
performed to assess data differences among various groups. P b 0.05
was considered statistically significant.

Results

Obesity, dyslipidemia and insulin resistance induced by mHFD in

C57BL/6 mice

During the 23-week feeding period, before the drug treatments,

body weights of the mice fed with mHFD were much higher than
those of mice on normal diet (P b 0.001). At the end of the study, body
weights of the model mice increased by 68% compared to the controls
(Fig. 1a). White adipose tissue weights of the model mice were about
5-fold greater than those of the control mice. After the 8-week
treatment with fenofibrate or rosiglitazone, body weights of the
fenofibrate-treated mice were markedly reduced by 29%, while body
weights of the rosiglitazone-treated mice were not significantly
changed. Adiposity index of the fenofibrate-treated mice was also
decreased by 43% as shown in Table 2. There was, however, no
significant difference in food intake among all mHFD-fed mice with or
without drug treatment (data not shown).
After 4-h fasting, the model mice demonstrated hypercholester-
olemia, but not hypertriglyceridemia, and even lower NEFA levels
as compared with the normal control mice. Compared to the model
mice, serum TC, TG and NEFA were significantly decreased in both
fenofibrate and rosiglitazone groups (Table 2).
Fasting blood glucose and insulin levels were measured at the end
of the experiment. A 6-fold increase in insulin levels was induced by
the mHFD feeding in the model mice as compared with the control
mice (Table 2). Compared to the model mice, insulin levels were
reduced by 57% and 53% in fenofibrate-treated and rosiglitazone-
treated mice, respectively. HOMA-IR index of the model mice was 10-
fold higher than that of the controls. Compared to the model mice,

Table 2
Characteristics of mice fed normal chow or modified high-fat diet

Item Con Model Feno Rosi

White adipose tissue (g) 0.62 ± 0.1 3.16 ± 0.1⁎⁎⁎ 1.30 ± 0.1### 3.19 ± 0.3
Adiposity index (%) 2.14 ± 0.1 6.50 ± 0.2⁎⁎⁎ 3.72 ± 0.3### 6.68 ± 0.4
TC (mg/dl) 75.9 ± 2.9 171.9 ± 7.9⁎⁎⁎ 102.9 ± 7.1### 113.3 ± 11.2###
Fig. 1. Obesity and insulin resistance induced by mHFD in C57BL/6 mice. (a) Changes in TG (mg/dl) 96.7 ± 4.4 99.8 ± 3.8 62.0 ± 3.4### 79.9 ± 4.1###
body weights. (b) Insulin sensitivity assessed by insulin tolerance test. (c) Glucose NEFA (uEq/L) 2520 ± 77 2120 ± 73⁎⁎⁎ 1433 ± 88### 1254 ± 51###
tolerability assessed by oral glucose tolerance test. Con = control, Model = mHFD Insulin (mIU/L) 10.9 ± 0.7 70.2 ± 6.1⁎⁎⁎ 29.9 ± 2.4### 33.2 ± 2.4###
without drug treatment, Feno = mHFD with fenofibrate treatment, Rosi = mHFD with HOMA-IR (mmol/L mIU/L) 2.06 ± 0.1 21.5 ± 2.2⁎⁎⁎ 5.23 ± 0.6### 6.23 ± 0.6###
rosiglitazone treatment. Values are expressed as mean ± SEM (10 mice in each group).
White adipose tissue included epididymal fat pad and abdominal adipose tissue.
⁎⁎⁎: P b 0.001 vs. Con. ##, ###: P b 0.01, 0.001 vs. Model.
Adiposity index was calculated as white adipose tissue weight (g) / body weight (g) ×
100. Serum parameters were measured in the fasting state of the mice. Con = control,
Model = modified high-fat diet without drug treatment, Feno = modified high-fat diet
from each group and the assay of every gene in each sample was with fenofibrate treatment, Rosi = modified high-fat diet with rosiglitazone treatment,
replicated three times. The mouse β-actin gene was amplified as a TC = total cholesterol, TG = triglyceride. Values are expressed as mean ± SEM (10 mice
loading control. The PCR products were separated by electrophoresis in each group). ⁎⁎⁎: P b 0.001 vs. Con. ###: P b 0.001 vs. Model.
986 W.-N. Cong et al. / Life Sciences 82 (2008) 983–990

HOMA-IR values were decreased by 76% and 71% after the treatment mation and fibrosis, with non significant improvement in steatosis
with fenofibrate or rosiglitazone. (Fig. 2d, h). Consequently, the steatohepatitis scores were significantly
Compared to the controls, the model mice displayed reduced re- lowered in the fenofibrate group and slightly reduced without
sponse to exogenous insulin in the insulin tolerance test (ITT). Addi- significance in the rosiglitazone group.
tionally, a 27% increase in the AUC was seen in the models. Treatment Consistent with the results of morphological observations, the
with either fenofibrate or rosiglitazone significantly improved the insulin hepatic TG content was 5-fold increased in the model mice in
sensitivity and decreased the AUC by 31% and 50% respectively (Fig. 1b). comparison with the controls. The hepatic fat accumulation was
Compared to the controls, mice in the model displayed impaired oral decreased in the fenofibrate group (P b 0.001), but not significantly
glucose tolerance (IGT), with 108% increase in AUC. Treatment with either reduced in the rosiglitazone group (Table 3).
fenofibrate or rosiglitazone significantly improved glucose tolerability, To further evaluate the degree of liver injury and the therapeutic
and decreased their AUC by 56% and 60% respectively (Fig. 1c). potential of these two agonists in this novel model, serum ALT levels
were assayed. As shown in Table 3, serum ALT levels of the model mice
Pronounced NASH caused by mHFD in C57BL/6 mice increased by 99%. The elevated ALT levels were decreased by about
31% in fenofibrate group of mice, but not markedly reduced in the
H&E staining revealed severe steatohepatitis in the livers from rosiglitazone-treated mice.
model mice (Fig. 2b), compared to livers from the controls (Fig. 2a).
Furthermore, livers of the model mice showed severe macro and Molecular changes of hepatic genes involved in fatty acid metabolism
micro-vesicular steatosis around the pericentral zone and scattered induced by mHFD
foci of lobular inflammation. Masson's trichrome staining revealed
mild fibrosis in the livers of the model mice (Fig. 2f), compared to the SREBP-1c is a key hepatic transcriptional factor regulating de novo
control mice (Fig. 2e). The overall steatohepatitis scores of the model lipogenesis in liver (Jump et al., 2001; Shimano et al., 1999; Brown and
mice were much higher than the controls (Fig. 2i). Administration of Goldstein, 1998). Compared to the controls, hepatic SREBP-1c mRNA
fenofibrate significantly attenuated steatosis, inflammation and levels of the model group were reduced by 32% (P b 0.05). The mRNA
fibrosis (Fig. 2c, g). Rosiglitazone only mildly decreased the inflam- levels of hepatic ACCα, the target gene of SREBP-1c and the key

Fig. 2. Non-alcoholic steatohepatitis caused by mHFD in C57BL/6 mice. (a, b, c, d) Hematoxylin and eosin (H&E) staining of livers from each group; original magnifications ×200.
(a) Normal liver histology of control mouse. (b) Extensive intracellular vacuolization and foci of inflammation (arrows) in the model mouse. (c) Reduced lipid accumulation and
inflammatory focus in the fenofibrate-treated mouse. (d) Mild reduction of inflammatory focus in the rosiglitazone-treated mouse. (e, f, g, h) Masson staining of livers from another
mouse in each group; original magnifications ×200. (e) Normal liver histology of the control mouse. (f) Mild fibrosis in the model mouse. (g) Attenuated fibrosis in the fenofibrate-
treated mouse. (h) Attenuated fibrosis in the rosiglitazone-treated mouse. (i) Overall steatohepatitis scores. Con = control, Model = mHFD without drug treatment, Feno = mHFD with
fenofibrate treatment, Rosi = mHFD with rosiglitazone treatment. Values are expressed as mean ± SEM (10 mice in each group). ⁎⁎: P b 0.01 vs. Con. ##: P b 0.01 vs. Model.
 W.-N. Cong et al. / Life Sciences 82 (2008) 983–990 987

Table 3 The nuclear receptor PPARγ is regarded as another regulator of

Hepatic triglyceride content and serum alanine aminotransferase levels lipogenesis in the liver (Schadinger et al., 2005). As shown in Fig. 3b, an
Item Con Model Feno Rosi
approximately 5-fold increase in hepatic PPARγ mRNA levels was induced
by mHFD feeding in the model mice. Both CD36 and ap2, which are
Hepatic TG content 2.1 ± 0.1 10.6 ± 1.0⁎⁎⁎ 3.0 ± 0.8 ###
8.4 ± 0.7
regulated by PPARγ (Yu et al., 2003), were also induced by 45% and 18%
(mg/g liver)
Serum ALT levels (U/L) 99.7 ± 3.4 198.7 ± 15.6⁎⁎⁎ 136.5 ± 5.6 ###
205.4 ± 12.6 respectively in the model mice. Up-regulation of PPARγ was significantly
attenuated by about 79% and 22%, respectively, in the fenofibrate and
Hepatic TG content and serum ALT levels were measured in the fasting state of the mice.
Con = control, Model = modified high-fat diet without drug treatment, Feno = modified rosiglitazone groups. The mildly elevated ap2 mRNA levels were
high-fat diet with fenofibrate treatment, Rosi = modified high-fat diet with rosiglitazone significantly decreased by 21% in rosiglitazone-treated mice. The
treatment, TG = triglyceride, ALT = alanine aminotransferase. Values are expressed as enhanced expression of CD36 was not inhibited by either drug treatment.
mean ± SEM (n = 10 in each group). ⁎⁎⁎: P b 0.001 vs. Con. ###: P b 0.001 vs. Model. Contrary to changes in hepatic PPARγ gene expression, hepatic PPARα
mRNA levels of the model mice were about 3-fold reduced (Fig. 3c).
Gene expression levels of hepatic CPT-1a, the rate-limiting enzyme
enzyme of de novo lipogenesis (Towle et al., 1997; Foufelle et al., 1996), of β-oxidation of fatty acid in mitochondria (McGarry and Foster, 1980;
were also down regulated by 35% in the model mice (P b 0.05). Hepatic McGarry and Brown, 1997; Ryu et al., 2005) were reduced by 32% in
gene expression of FAS, the rate-limiting enzyme of de novo the model mice (P b 0.001). The down-regulated PPARα mRNA
lipogenesis (Fukuda et al., 1999), was not markedly changed in the expressions were restored by 38% in the fenofibrate group but further
model mice (Fig. 3a). Compared to the models, SREBP-1c mRNA levels depressed by 62% in the rosiglitazone group. Treatment with
were further decreased by 48% and 73% in fenofibrate or rosiglitazone- fenofibrate but not with rosiglitazone restored the decreased mRNA
treated mice. Overall, treatment with any of these drugs did not expression of CPT-1a by 33% (P b 0.001). Although we chose more
significantly affect the expression levels of ACCα and FAS. samples (n = 4–5) in each group as shown in Fig. 3d and did more

Fig. 3. Changes in hepatic expression of genes involved in fatty acid metabolism. Gene expressions of (a) SREBP-1c and the target genes of hepatic de novo lipogenesis. (b) PPARγ and
the target genes involved with hepatic lipogenesis. (c) PPARα and CPT-1a mediate fatty acid oxidation. (d) The gel of control samples for each gene. Con = control, Model = mHFD
without drug treatment, Feno = mHFD with fenofibrate treatment, Rosi = mHFD with rosiglitazone treatment. RT-PCR reactions of each sample were performed in three independent
experiments, and 4–5 samples were chosen at random from each group. Relative mRNA levels are presented as mean ± SEM. ⁎, ⁎⁎⁎: P b 0.05, 0.001 vs. Con. #, ##, ###: P b 0.05, 0.01,
0.001 vs. Model.
988 W.-N. Cong et al. / Life Sciences 82 (2008) 983–990

Fig. 4. Aberrant hepatic PPARα and PPARγ nuclear protein levels. Nuclear PPARα and PPARγ protein levels in all groups were assessed by Western blots. Relative protein expression
levels are presented as mean ± SEM of 10 mice from three independent experiments. ⁎, ⁎⁎, P b 0.05, 0.01, vs. Con. #, ##, P b 0.05, 0.01, vs. Model.

repeats to make our experiments sufficient for showing changes, we characteristics of severe steatohepatitis, insulin resistance and
are still aware that the RT-PCR method we used is not strictly dyslipidemia that are similar to the general clinical features of
quantitative. The results are therefore tentative and need to be human NASH, and could therefore be a valuable tool in studying the
confirmed using more quantitative methods. pathophysiology of NASH as well as for the development of appro-
priate pharmacotherapy and preventative measures for NASH.
The alterations of hepatic PPARα and PPARγ protein levels induced by De novo lipogenesis plays a key role in hepatic lipid metabolism
mHFD and is modulated by SREBP-1c (Farrell and Larter, 2006). In one
mechanistic study, increased expression of SREBP-1c and the key
Western blot analysis revealed a profound up-regulation of hepatic enzymes of de novo lipogenesis were identified as key mediators of
nuclear protein levels of PPARγ and PPARα induced by the mHFD (Fig. 4) hepatic steatosis in a high-fat diet-induced animal model (Osei-
compared to the controls. Treatments with fenofibrate and rosiglitazone Hyiaman et al., 2005). Excessive fat uptake induces the expression
markedly attenuated the enhanced PPARγ protein levels; whereas of SREBP-1c, which then mediates transcription of the two key
the protein levels of hepatic PPARα were not affected by either drug enzymes in de novo lipogenesis (ACCα and FAS), and results in a
treatment. pronounced increase in hepatic lipid accumulation (Farrell and
Larter, 2006). The results of the present study indicated that, in our
Discussion mHFD-induced NASH model, hepatic mRNA levels of SREBP-1c and
ACCα were both decreased, with only a slight increase in FAS levels.
Several studies have demonstrated the complex relationship It is likely that PUFA from the soybean oil in the mHFD may have
between NASH and metabolic disorders such as central obesity, triggered the suppression of SREBP-1c in the model mice (Xu et al.,
dyslipidemia and insulin resistance in both humans and rodents 1999). Considering the fact that both treatments with fenofibrate
(Samuel et al., 2004). Further research in this field has been hampered and rosiglitazone further suppressed SREBP-1c mRNA levels, we
by the lack of an appropriate animal model of NASH that can also propose that the development of steatohepatitis and associated
display the features of metabolic syndrome with obesity, insulin insulin resistance could not be explained by de novo lipogenesis
resistance and dyslipidemia simultaneously, in a manner that alone.
approximates the human condition more closely. Although animal Hepatic PPARγ plays a crucial role in the development of hepatic
models force-fed with high-fat diet or high-fat emulsion successfully steatosis (Rahimian et al., 2001). In our novel NASH model, the
developed steatohepatitis and insulin resistance, the manner of significant up-regulation of both PPARγ mRNA and protein expression
feeding and the mild degree of hepatic injury do not reflect well the was observed. In addition, the mRNA levels of ap2 and CD36, which
clinical cases and are unsuitable for studying the natural history of are the target genes of PPARγ and responsible for the transport of fatty
NASH (Zou et al., 2006; Deng et al., 2005). acid (Yu et al., 2003), were also significantly enhanced. The increased
In the present study, we have established a novel mouse model for expression of hepatic ap2 and CD36 is suggestive of a functional role
NASH in C57BL/6 mice using 23-week ad libitum feeding of mHFD: for PPARγ signaling in the combined development of NASH, obesity
The mHFD contained much more saturated fatty acid from lard, and and insulin resistance in the model mice (Memon et al., 2000;
had less methionine and choline. This model exhibited the typical Matsusue et al., 2003). It is also possible that both fenofibrate and
hepatic NASH lesions, such as steatosis, inflammation and fibrosis. rosiglitazone may directly suppress hepatic PPARγ protein levels as
Most importantly, this model also developed significant energy opposed to inhibition of its transcriptional activity. These results
metabolism abnormalities, such as obesity, insulin resistance, and suggest that lipogenesis mediated by aberrant PPARγ signaling may
dyslipidemia. This NASH model in the present study clearly displayed underlie the onset of hepatic steatosis in this NASH model.
 W.-N. Cong et al. / Life Sciences 82 (2008) 983–990 989

Another important factor in the development of fatty liver could be Chinese Medicine, People's Republic of China (Guo Zhong Yi Yao Ke
altered hepatic lipid oxidation, which is always correlated with the 02-03 ZP11).
expression and activation of PPARα. Evidence from PPARα knockout
studies indicated that deficiency of PPARα could lead to non-alcoholic Appendix A. Supplementary data
steatohepatitis and disorders of energy metabolism (Reddy, 2001; Yu
et al., 2007). CPT-1a is the rate-limiting enzyme for fatty acid β- Supplementary data associated with this article can be found, in
oxidation and is also a target gene for PPARα. As previously reported, the online version, at doi:10.1016/j.lfs.2008.01.022.
the disturbance in mitochondrial fatty acid β-oxidation due to down-
regulation of hepatic CPT-1a is a crucial mechanism in pathogenesis of References
hepatic steatosis (Den Boer et al., 2004; Nagasawa et al., 2006). In
present study, the significant decrease in both PPARα and CPT-1a Angulo, P., 2002. Nonalcoholic fatty liver disease. The New England Journal of Medicine
mRNA levels also suggests altered hepatic mitochondrial fatty acid 346, 1221–1231.
Brown, M.S., Goldstein, J.L., 1998. Sterol regulatory element binding proteins (SREBPs):
oxidation as results of long-term mHFD feeding, and treatment with controllers of lipid synthesis and cellular uptake. Nutrition Review 56, S1–S3.
fenofibrate normalized the expression of CPT-1a and PPARα: These Bugianesi, E., McCullough, A.J., Marchesini, G., 2005. Insulin resistance: a metabolic
results are in line with the favorable effects of fenofibrate on NASH; pathway to chronic liver disease. Hepatology 42, 987–1000.
Chao, L., Marcus-Samuels, B., Mason, M.M., Moitra, J., Vinson, C., Arioglu, E., Gavrilova,
while the lower effectiveness of rosiglitazone on NASH might O., Reitman, M.L., 2000. Adipose tissue is required for the antidiabetic, but not for
correlate with its further inhibition of PPARα and CPT-1a at the the hypolipidemic, effect of thiazolidinediones. The Journal of Clinical Investigation
transcriptional level. It is reasonable to postulate that the reduction in 106, 1221–1228.
Collins, S., Martin, T.L., Surwit, R.S., Robidoux, J., 2004. Genetic vulnerability to diet-
hepatic lipid oxidation mediated by PPARα and CPT-1a might be
induced obesity in the C57BL/6J mouse: physiological and molecular characteristics.
another important mechanism of steatohepatitis in this novel NASH Physiology & Behavior 81, 243–248.
model. De Oliveira, C.P., Stefano, J.T., de Lima, V.M., de Sa, S.V., Simplicio, F.I., de Mello, E.S.,
Earlier studies have reported that in rodents fed normal chow, Correa-Giannella, M.L., Alves, V.A., Laurindo, F.R., de Oliveira, M.G., Giannella-Neto,
D., Carrilho, F.J., 2006. Hepatic gene expression profile associated with non-
treatment with fenofibrate significantly reduced body weight, fat pad alcoholic steatohepatitis protection by S-nitroso-N-acetylcysteine in ob/ob mice.
weight, blood glucose, triglyceride, and free fatty acid levels (Haluzik Journal of Hepatology 45, 725–733.
et al., 2006), and that rosiglitazone showed only little effect on these Den Boer, M., Voshol, P.J., Kuipers, F., Havekes, L.M., Romijn, J.A., 2004. Hepatic steatosis:
a mediator of the metabolic syndrome. Lessons from animal models. Arterio-
parameters and histology of mouse liver (Chao et al., 2000; Sanchez sclerosis, Thrombosis, and Vascular Biology 24, 644–649.
et al., 2002). In present study, both rosiglitazone and fenofibrate treat- Deng, Q.G., She, H., Cheng, J.H., French, S.W., Koop, D.R., Xiong, S., Tsukamoto, H., 2005.
ments effectively improved insulin resistance and dyslipidemia, but only Steatohepatitis induced by intragastric overfeeding in mice. Hepatology 42, 905–914.
Farrell, G.C., Larter, C.Z., 2006. Nonalcoholic fatty liver disease: from steatosis to cirrhosis.
fenofibrate significantly reversed the morphological changes of NASH, Hepatology 43, S99–S112.
reduced hepatic TG content and decreased ALT levels. In the further Foufelle, F., Girard, J., Ferre, P., 1996. Regulation of lipogenic enzyme expression by
study, the up-regulations of PPARγ and ap2 were reversed by the glucose in liver and adipose tissue: a review of the potential cellular and molecular
mechanisms. Advances in Enzyme Regulation 36, 199–226.
treatments of both rosiglitazone and fenofibrate, whereas only Fukuda, H., Iritani, N., Sugimoto, T., Ikeda, H., 1999. Transcriptional regulation of fatty
fenofibrate treatment recovered the reduced mRNA expressions of acid synthase gene by insulin/glucose, polyunsaturated fatty acid and leptin in
PPARα and CPT-1a. Thus, the therapeutic effects of fenofibrate on NASH hepatocytes and adipocytes in normal and genetically obese rats. European Journal
of Biochemistry 260, 505–511.
might mainly result from its differential effects on both of PPARα and
Haluzik, M.M., Lacinova, Z., Dolinkova, M., Haluzikova, D., Housa, D., Horinek, A.,
PPARγ signaling pathways. Furthermore, treatment with rosiglitazone Vernerova, Z., Kumstyrova, T., Haluzik, M., 2006. Improvement of insulin sensitivity
or fenofibrate did not affect food intake, but fenofibrate significantly after peroxisome proliferator-activated receptor-{alpha} agonist treatment is
reduced the final body weights of mice. Although de novo lipogenesis accompanied by paradoxical increase of circulating resistin levels. Endocrinology
147, 4517–4524.
might not be the major cause of the pathogenesis in the established Harrity, T., Farrelly, D., Tieman, A., Chu, C., Kunselman, L., Gu, L., Ponticiello, R., Cap, M.,
NASH model, fenofibrate treatment markedly inhibited de novo Qu, F., Shao, C., Wang, W., Zhang, H., Fenderson, W., Chen, S., Devasthale, P., Jeon, Y.,
lipogenesis. These results suggest that, at least in this study, the Seethala, R., Yang, W.P., Ren, J., Zhou, M., Ryono, D., Biller, S., Mookhtiar, K.A.,
Wetterau, J., Gregg, R., Cheng, P.T., Hariharan, N., 2006. Muraglitazar, a novel dual
therapeutic effect of fenofibrate on NASH may be mediated through ({alpha}/{gamma}) peroxisome proliferator-activated receptor activator, improves
multiple pathways, besides the actions on PPARα and PPARγ in this diabetes and other metabolic abnormalities and preserves {beta}-cell function in
novel NASH model accompanied with obesity and insulin resistance. db/db mice. Diabetes 55, 240–248.
Ip, E., Farrell, G., Hall, P., Robertson, G., Leclercq, I., 2004. Administration of the potent
Further research is needed to clarify the multiple mechanisms of PPARalpha agonist, Wy-14,643, reverses nutritional fibrosis and steatohepatitis in
fenofibrate effects on NASH. In addition, other paradoxical results such mice. Hepatology 39, 1286–1296.
as the lack of reconciliation in PPARα mRNA and protein levels also Ip, E., Farrell, G.C., Robertson, G., Hall, P., Kirsch, R., Leclercq, I., 2003. Central role of
PPARalpha-dependent hepatic lipid turnover in dietary steatohepatitis in mice.
require further studies.
Hepatology 38, 123–132.
Jump, D.B., Thelen, A.P., Mater, M.K., 2001. Functional interaction between sterol regulatory
Conclusion element-binding protein-1c, nuclear factor Y, and 3,5,3′-triiodothyronine
nuclear receptors. Journal of Biological Chemistry 276, 34419–34427.
Matsusue, K., Haluzik, M., Lambert, G., Yim, S.H., Gavrilova, O., Ward, J.M., Brewer Jr., B.,
In summary, we have successfully established a chronic mHFD- Reitman, M.L., Gonzalez, F.J., 2003. Liver-specific disruption of PPAR {gamma} in
induced novel NASH model with obesity and insulin resistance. This leptin-deficient mice improves fatty liver but aggravates diabetic phenotypes. The
model could be a valuable tool to study the mechanisms and the Journal of Clinical Investigation 111, 737–747.
Matthews, D.R., Hosker, J.P., Rudenski, A.S., Naylor, B.A., Treacher, D.F., Turner, R.C., 1985.
pharmacotherapy of not just NASH but also of energy metabolic Homeostasis model assessment: insulin resistance and beta-cell function from
disorders. We conclude that a combination of aberrant expression of plasma fasting glucose and insulin concentrations in man. Diabetologia 28, 412–419.
hepatic PPARα and PPARγ and their target genes appears to play an McGarry, J.D., Brown, N.F., 1997. The mitochondrial carnitine palmitoyltransferase
system. from concept to molecular analysis. European Journal of Biochemistry 244,
important role in the pathogenesis of NASH, and that the PPARα 1–14.
agonist fenofibrate may be a potent therapy for insulin resistance- McGarry, J.D., Foster, D.W., 1980. Regulation of hepatic fatty acid oxidation and ketone
obesity associated NASH. body production. Annual Review of Biochemistry 49, 395–420.
Memon, R.A., Tecott, L.H., Nonogaki, K., Beigneux, A., Moser, A.H., Grunfeld, C., Feingold,
K.R., 2000. Up-regulation of peroxisome proliferator-activated receptors (PPAR-
Acknowledgements {alpha}) and PPAR-{gamma} messenger ribonucleic acid expression in the liver in
murine obesity: troglitazone induces expression of PPAR-{gamma}-responsive
adipose tissue-specific genes in the liver of obese diabetic mice. Endocrinology 141,
This study was supported by the Natural Sciences Funds of China
4021–4031.
(No.30572215), National High-tech R&D Program (863 Program, Nagasawa, T., Inada, Y., Nakano, S., Tamura, T., Takahashi, T., Maruyama, K., Yamazaki, Y.,
No.2006AA09Z446) and the State Administration of Traditional Kuroda, J., Shibata, N., 2006. Effects of bezafibrate, PPAR pan-agonist, and
990 W.-N. Cong et al. / Life Sciences 82 (2008) 983–990

GW501516, PPAR [delta] agonist, on development of steatohepatitis in mice fed a Schadinger, S.E., Bucher, N.L.R., Schreiber, B.M., Farmer, S.R., 2005. PPAR {gamma} 2
methionine- and choline-deficient diet. European Journal of Pharmacology 536, regulates lipogenesis and lipid accumulation in steatotic hepatocytes. American
182–191. Journal of Physiology, Endocrinology and Metabolism 288, E1195–E1205.
Neish, A.S., Khachigian, L.M., Park, A., Baichwal, V.R., Collins, T., 1995. Sp1 is a component Shimano, H., Yahagi, N., Amemiya-Kudo, M., Hasty, A.H., Osuga, J.I., Tamura, Y., Shionoiri, F.,
of the cytokine-inducible enhancer in the promoter of vascular cell adhesion Iizuka, Y., Ohashi, K., Harada, K., Gotoda, T., Ishibashi, S., Yamada, N., 1999. Sterol
molecule-1. Journal of Biological Chemistry 270, 28903–28909. regulatory element-binding protein-1 as a key transcription factor for nutritional
Neuschwander-Tetr, B.A., Caldwell, S.H., 2003. Nonalcoholic steatohepatitis: summary induction of lipogenic enzyme genes. Journal of Biological Chemistry 274, 35832–35839.
of an AASLD single topic conference. Hepatology 37, 1202–1219. Towle, H.C., Kaytor, E.N., Shih, H.M., 1997. Regulation of the expression of lipogenic
Osei-Hyiaman, D., DePetrillo, M., Pacher, P., Liu, J., Radaeva, S., Batkai, S., Harvey-White, enzyme genes by carbohydrate. Annual Review of Nutrition 17, 405–433.
J., Mackie, K., Offertaler, L., Wang, L., Kunos, G., 2005. Endocannabinoid activation at Xu, J., Nakamura, M.T., Cho, H.P., Clarke, S.D., 1999. Sterol regulatory element binding
hepatic CB1 receptors stimulates fatty acid synthesis and contributes to diet- protein-1 expression is suppressed by dietary polyunsaturated fatty acids. a
induced obesity. The Journal of Clinical Investigation 115, 1298–1305. mechanism for the coordinate suppression of lipogenic genes by polyunsaturated
Rahimian, R., Masih-Khan, E., Lo, M., van Breemen, C., McManus, B.M., Dube, G.P., 2001. fats. Journal of Biological Chemistry 274, 23577–23583.
Hepatic over-expression of peroxisome proliferator activated receptor gamma2 in Yang, B., Brown, K.K., Chen, L., Carrick, K.M., Clifton, L.G., McNulty, J.A., Winegar, D.A.,
the ob/ob mouse model of non-insulin dependent diabetes mellitus. Molecular and Strum, J.C., Stimpson, S.A., Pahel, G.L., 2004. Serum adiponectin as a biomarker for in
Cellular Biochemistry 224, 29–37. vivo PPARgamma activation and PPARgamma agonist-induced efficacy on insulin
Reddy, J.K., 2001. Nonalcoholic steatosis and steatohepatitis: III. Peroxisomal beta- sensitization/lipid lowering in rats. BMC Pharmacology 4, 23.
oxidation, PPARalpha, and steatohepatitis. The American Journal of Physiology – Yu, J., Chu, E.S.H., Hui, A.Y., Cheung, K.F., Chan, H.L.Y., Leung, W.K., Farrell, G.C., Sung, J.J.Y.,
Gastrointestinal and Liver Physiology 281, G1333–G1339. 2007. Lipoprotein lipase activator ameliorates the severity of dietary steatohepa-
Rinella, M.E., Green, R.M., 2004. The methionine-choline deficient dietary model of titis. Biochemical and Biophysical Research Communications 356, 53–59.
steatohepatitis does not exhibit insulin resistance. Journal of Hepatology 40, 47–51. Yu, S., Matsusue, K., Kashireddy, P., Cao, W.Q., Yeldandi, V., Yeldandi, A.V., Rao, M.S.,
Ryu, M.H., Daily Iii, J.W., Cha, Y.S., 2005. Effect of starvation on hepatic acyl-CoA Gonzalez, F.J., Reddy, J.K., 2003. Adipocyte-specific gene expression and adipogenic
synthetase, carnitine palmitoyltransferase-I, and acetyl-CoA carboxylase mRNA steatosis in the mouse liver due to peroxisome proliferator-activated receptor
levels in rats. Nutrition 21, 537–542. gamma 1 (PPARgamma 1) overexpression. Journal of Biological Chemistry 278,
Samuel, V.T., Liu, Z.X., Qu, X., Elder, B.D., Bilz, S., Befroy, D., Romanelli, A.J., Shulman, G.I., 498–505.
2004. Mechanism of hepatic insulin resistance in non-alcoholic fatty liver disease. Zou, Y., Li, J., Lu, C., Wang, J., Ge, J., Huang, Y., Zhang, L., Wang, Y., 2006. High-fat
Journal of Biological Chemistry 279, 32345–32353. emulsion-induced rat model of nonalcoholic steatohepatitis. Life Sciences 79,
Sanchez, J.C., Converset, V., Nolan, A., Schmid, G., Wang, S., Heller, M., Sennitt, M.V., 1100–1107.
Hochstrasser, D.F., Cawthorne, M.A., 2002. Effect of rosiglitazone on the differential
expression of diabetes-associated proteins in pancreatic islets of C57Bl/6 lep/lep
mice. Molecular and Cellular Proteomics 1, 509–516.