Lipid-induced oxidative stress causes steatohepatitis in mice fed an atherogenic diet

Authors

  • Naoto Matsuzawa,

    1. Department of Disease Control and Homeostasis, Graduate School of Medical Science, Kanazawa University, Kanazawa, Japan
    2. Department of Pharmacy and Health Science, Graduate School of Natural Science and Technology, Kanazawa University, Kanazawa, Japan
    Search for more papers by this author
  • Toshinari Takamura,

    Corresponding author
    1. Department of Disease Control and Homeostasis, Graduate School of Medical Science, Kanazawa University, Kanazawa, Japan
    • Department of Disease Control and Homeostasis, Graduate School of Medical Science, Kanazawa University, 13-1 Takara-machi, Kanazawa, Ishikawa 920-8641, Japan===

    Search for more papers by this author
    • fax: (81) 76-234-4250

  • Seiichiro Kurita,

    1. Department of Disease Control and Homeostasis, Graduate School of Medical Science, Kanazawa University, Kanazawa, Japan
    Search for more papers by this author
  • Hirofumi Misu,

    1. Department of Disease Control and Homeostasis, Graduate School of Medical Science, Kanazawa University, Kanazawa, Japan
    Search for more papers by this author
  • Tsuguhito Ota,

    1. Department of Disease Control and Homeostasis, Graduate School of Medical Science, Kanazawa University, Kanazawa, Japan
    Search for more papers by this author
  • Hitoshi Ando,

    1. Department of Disease Control and Homeostasis, Graduate School of Medical Science, Kanazawa University, Kanazawa, Japan
    Search for more papers by this author
  • Masayoshi Yokoyama,

    1. Department of Disease Control and Homeostasis, Graduate School of Medical Science, Kanazawa University, Kanazawa, Japan
    Search for more papers by this author
  • Masao Honda,

    1. Department of Disease Control and Homeostasis, Graduate School of Medical Science, Kanazawa University, Kanazawa, Japan
    Search for more papers by this author
  • Yoh Zen,

    1. Department of Human Pathology, Graduate School of Medical Science, Kanazawa University, Kanazawa, Japan
    Search for more papers by this author
  • Yasuni Nakanuma,

    1. Department of Human Pathology, Graduate School of Medical Science, Kanazawa University, Kanazawa, Japan
    Search for more papers by this author
  • Ken-ichi Miyamoto,

    1. Department of Pharmacy and Health Science, Graduate School of Natural Science and Technology, Kanazawa University, Kanazawa, Japan
    Search for more papers by this author
  • Shuichi Kaneko

    1. Department of Disease Control and Homeostasis, Graduate School of Medical Science, Kanazawa University, Kanazawa, Japan
    Search for more papers by this author

  • Potential conflict of interest: Nothing to report.

Abstract

Recently, nonalcoholic steatohepatitis (NASH) was found to be correlated with cardiovascular disease events independently of the metabolic syndrome. The aim of this study was to investigate whether an atherogenic (Ath) diet induces the pathology of steatohepatitis necessary for the diagnosis of human NASH and how cholesterol and triglyceride alter the hepatic gene expression profiles responsible for oxidative stress. We investigated the liver pathology and plasma and hepatic lipids of mice fed the Ath diet. The hepatic gene expression profile was examined with microarrays and real-time polymerase chain reactions. The Ath diet induced dyslipidemia, lipid peroxidation, and stellate cell activation in the liver and finally caused precirrhotic steatohepatitis after 24 weeks. Cellular ballooning, a necessary histological feature defining human NASH, was observed in contrast to existing animal models. The addition of a high-fat component to the Ath diet caused hepatic insulin resistance and further accelerated the pathology of steatohepatitis. A global gene expression analysis revealed that the Ath diet up-regulated the hepatic expression levels of genes for fatty acid synthesis, oxidative stress, inflammation, and fibrogenesis, which were further accelerated by the addition of a high-fat component. Conversely, the high-fat component down-regulated the hepatic gene expression of antioxidant enzymes and might have increased oxidative stress. Conclusion: The Ath diet induces oxidative stress and steatohepatitis with cellular ballooning. The high-fat component induces insulin resistance, down-regulates genes for antioxidant enzymes, and further aggravates the steatohepatitis. This model suggests the critical role of lipids in causing oxidative stress and insulin resistance leading to steatohepatitis. (HEPATOLOGY 2007.)

Nonalcoholic fatty liver disease (NAFLD) is currently the most common chronic liver condition in the Western world. Clinical, epidemiological, and biochemical data strongly support the concept that NAFLD is the hepatic manifestation of the metabolic syndrome, the constellation of metabolic abnormalities including obesity, dyslipidemia, and insulin resistance.1 NAFLD includes not only steatosis (without other injury) but also various degrees of inflammation and fibrosis.2 Simple steatosis is usually considered benign, but the development of inflammatory changes in the liver [called nonalcoholic steatohepatitis (NASH)] is recognized as a precursor to more severe liver disease and sometimes evolves into cryptogenic cirrhosis.3 It has been recently proposed that NASH is strongly correlated with cardiovascular disease events independently of the metabolic syndrome.4 Therefore, further investigations of NASH are required to elucidate the pathogenesis of this process and to develop treatments.

To date, however, studies of NASH have been hampered by the absence of a suitable experimental model. The use of genetic defects or targeted overexpression to produce obesity5 or impaired hepatic lipid metabolism6 in rodents has been used as an NAFLD model. Although these genetic manipulations can assess the biological importance of each gene in vivo, they might not reflect the natural etiology of NAFLD in patients and rarely lead to the pathology of NASH. The other models frequently used are based on nutritional manipulations. Natural nutritional models have been described, including the use of a sucrose-rich and fat-rich diet.7 However, in these models, rodents accumulate minimal fat and develop subtle inflammation of the liver. The methionine- and choline-deficient (MCD) model, which is frequently used to produce more progressive liver pathology, leads to the development of steatosis with lobular inflammation and with perisinusoidal and pericentral fibrosis.8, 9 However, this model lacks lipotrophic factors, insulin resistance,10 or the cellular ballooning that is observed only with the addition of a high-fat component to the MCD diet.11

In this study, we focused on an atherogenic (Ath) diet, which contains cholesterol and cholic acid. Because the diet produces not only an Ath lipoprotein profile but also vascular fatty streak lesions, it has been widely used to study atherosclerosis in animals, including mice.12 Although the Ath diet has recently been reported to induce liver steatosis, inflammation, and fibrosis,13 lipid metabolism, insulin resistance, and hepatic gene expression profiles responsible for liver pathology remain to be determined in this model. To address this issue, we investigated the time course of the pathological changes and gene expression profiles of the liver in mice fed the Ath diet. In addition, by adding a high-fat component to the Ath diet, we elucidated the impact of insulin resistance, which is commonly observed in NASH patients, on the development of oxidative stress in the liver and pathology of steatohepatitis.

Abbreviations

4-HNE, 4-hydroxy-2-nonenal; α-SMA, α-smooth muscle actin; ALT, alanine aminotransferase; Ath, atherogenic; Ath+HF, atherogenic and high-fat; AUC, area under the curve; BW, body weight; Col1a1, procollagen type I alpha 1; Col1a2, procollagen type I alpha 2; Col4a1, procollagen type IV alpha 1; CPT-1a, carnitine palmitoyltransferase 1a; FAS, fatty acid synthase; FFA, free fatty acid; GPCR, G protein-coupled receptor; GPCRDB, G Protein-Coupled Receptor Database; GTT, glucose tolerance test; H&E, hematoxylin-eosin; HDL-C, high-density lipoprotein-cholesterol; HOMA-IR, homeostasis model assessment of insulin resistance; HPLC, high-performance liquid chromatography; HSC, hepatic stellate cell; IRS, insulin receptor substrate; ITT, insulin tolerance test; LDL, low-density lipoprotein; LDL-C; low-density lipoprotein-cholesterol; MAPK, mitogen-activated protein kinase; MCD, methionine- and choline-deficient; mRNA, messenger RNA; NADPH, reduced-form nicotinamide adenine dinucleotide phosphate; NAFLD, nonalcoholic fatty liver disease; NASH, nonalcoholic steatohepatitis; ND, not determined; PAI-1, plasminogen activator inhibitor 1; PCR, polymerase chain reaction; PPARα, peroxisome proliferator-activated receptor α; ROS, reactive oxygen species; SEM, standard error of the mean; SREBP-1c, sterol regulatory element binding protein 1c; TBS-T, trishydroxymethylaminomethane-buffered saline Tween 20; TCA, tricarboxylic acid cycle; TG, triglyceride; TGF-β, transforming growth factor β; TNF-α, tumor necrosis factor α; VLDL-C, very low density lipoprotein-cholesterol.

Materials and Methods

Animals and Experimental Design.

Male C57Bl/6J mice were purchased from Charles River Laboratories Japan (Yokohama, Japan) at 6 weeks of age. After 2 weeks of acclimation, the mice were divided into the following 3 groups: (1) control mice given a standard chow (CRF-1, Charles River Laboratories Japan), (2) mice given an Ath diet, and (3) mice fed an atherogenic and high-fat (Ath+HF) diet. The Ath and Ath+HF diets were prepared by the addition of cocoa butter, cholesterol, and cholate to CRF-1. These diets were prepared by Oriental Yeast (Tokyo, Japan). The compositions of each diet are shown in Table 1 and Supplementary Table 1. At 6 weeks of age, the mice were housed in colony cages with a 12-hour light/12-hour dark cycle, and they were given food and water ad libitum. All animal procedures were in accordance with the standards set forth in the Guidelines for the Care and Use of Laboratory Animals at the Takara-Machi campus of Kanazawa University (Japan).

Table 1. The Composition of the 3 Diets
CompositionControlAthAth+HF
  1. The contents of vitamins and minerals in each diet are presented in Supplementary Table 1.

CRF-1 (%)10090.7538.25
Cocoa butter (%)7.5060.0
Cholesterol (%)1.251.25
Cholate (%)0.500.50
   (wt/wt %)
Energy compositionControlAthAth+HF
Carbohydrate (g)60.955.223.3
Protein (g)22.420.38.6
Fat (g)6.014.060.0
Total calorie (kcal)363411669
   (/100 g)

Blood Sampling and Analysis.

At 6, 12, or 24 weeks, blood samples were obtained from the tail vein following a 12-hour fast. Enzymatic assays for the total cholesterol, free cholesterol, free fatty acids (FFAs), triglyceride (TG), and alanine aminotransferase (ALT) were performed with kits purchased from Wako Pure Chemical Industries (Osaka, Japan). The cholesterol and TG profiles in plasma lipoproteins were analyzed with a dual-detection high-performance liquid chromatography (HPLC) system with 2 tandem connected TSKgel LipopropakXL columns (300 × 7.8 mm; Tosoh, Japan) by Skylight Biotech (Akita, Japan).14

Glucose Tolerance Tests (GTTs) and Insulin Tolerance Tests (ITTs).

At 12 weeks, GTTs and ITTs were conducted. For GTTs, glucose was administered (1.5 g/kg body weight) following a 12-hour fast. For ITTs, mice were injected intraperitoneally with insulin (0.5 U/kg of body weight; Humulin R, Eli Lilly, Indianapolis, IN) following a 4-hour fast. The glucose values were measured from whole venous blood with a blood glucose monitoring system (FreeStyle, Kissei, Matsumoto, Japan) 0, 15, 30, 60, and 120 minutes after the administration of glucose or insulin.

Pyruvate Challenge Test.

At 6 weeks, we conducted the pyruvate challenge test.15, 16 The mice, deprived of food for 16 hours, were injected intraperitoneally with pyruvate dissolved in saline (2 g/kg). The blood glucose values were measured 0, 15, 30, 60, and 90 minutes after the injection of pyruvate.

Tissue Preparation and Histological Examination.

At 6, 12, or 24 weeks, the mice were killed by cervical dislocation under diethyl ether anesthesia following a 12-hour fast. The livers were immediately removed and weighed. A large portion of each liver was snap-frozen in liquid nitrogen for later RNA studies. The remaining tissue was fixed in 10% buffered formalin, processed, and embedded in paraffin for hematoxylin-eosin (H&E), Azan, and Sirius red staining and was blindly scored by a single pathologist. Steatosis, fibrosis, and acinar inflammation were semiquantitatively evaluated according to the standard criteria of grading and staging for NASH, with minor modifications.17 To evaluate steatosis, we used the absolute percentage of the macrovesicular fat droplet area in the section area (that is, 8 × 105 hepatocytes in 4 mm2). For inflammation, 0 was defined as no hepatocyte injury or inflammation, 1 was defined as mild focal injury, 2 was defined as noticeable injury, and 3 was defined as severe zone 3 hepatocyte injury or inflammation. For fibrosis, 0 was defined as no fibrosis, 1 was defined as pericellular and perivenular fibrosis, 2 was defined as focal bridging fibrosis, 3 was defined as much bridging fibrosis with lobular distortion, and 4 was defined as cirrhosis.

Slides were immunostained with monoclonal mouse anti-human α-smooth muscle actin (α-SMA; Dako Japan, Kyoto, Japan). This was followed by the application of the immunoperoxidase technique with an Envision kit (Dako Japan). The peroxidase activity was identified by a reaction with 3′,3′-diaminobenzidine (Sigma, St Louis, MO). Areas staining for α-SMA were quantified morphometrically with WinROOF version 5.7 (Mitani Shoji, Fukui, Japan) and expressed as percentages of the field area.

Measuring the Hepatic Lipid Content.

Hepatic lipids were extracted with chloroform/methanol (2:1) according to a published method.18 With a kit (Wako), the extract was dissolved in water and subsequently analyzed for TG, total cholesterol, free cholesterol, and FFAs.

Measuring the Hepatic Hydroxyproline Content.

The hydroxyproline content in liver samples was quantified colorimetrically according to a published method.19 Briefly, a 0.2-g liver sample was homogenized in 6 N HCl and hydrolyzed at 110°C for 16 hours. The hydrolysate was filtered, aliquots were evaporated under a vacuum, and the sediment was redissolved in 50% isopropanol. Then, the samples were incubated in a solution containing 0.84% chloramine-T, 42 mM sodium acetate, 2.6 mM citric acid, and 39.5% (vol/vol) isopropanol (pH 6.0) for 10 minutes at room temperature. Next, the samples were incubated in a solution containing 0.248 g of p-dimethylaminobenzaldehyde dissolved in 0.27 mL of 60% perchloric acid and 0.73 mL of isopropanol for 90 minutes at 50°C. The hydroxyproline content was quantified photometrically at 558 nm.

Measuring Hepatic Protein Carbonyls.

The concentration of hepatic proteins containing carbonyl groups (those that react with 2,4-dinitrophenylhydrazine to form the corresponding hydrazone) was determined spectrophotometrically according to the instructions with a protein carbonyl assay kit (Cayman Chemical, Ann Arbor, MI).

RNA Preparation for the Microarray Analysis.

Total RNA was isolated from the frozen liver with the ToTALLY RNA kit (Applied Biosystems, Foster City, CA). Each sample was prepared by equal amounts of total RNA being pooled from 3 mice in the same group. Three micrograms of total RNA was used to synthesize antisense RNA with the AminoAllyl MessageAmp II antisense RNA kit (Applied Biosystems) for oligo-microarrays (AceGene Mouse Oligo Chip 30K, DNA Chip Research, Yokohama, Japan). Each microarray hybridization sample and the reference amino allyl antisense RNA were labeled with Cy5 and Cy3, respectively. Hybridization and washing were performed according to the manufacturer's instructions; this was followed by scanning with a G2505B microarray scanner (Agilent Technologies, Palo Alto, CA) and then image analysis with GenePix Pro 4.1 software (Axon Instruments, Union City, CA). Microarray data were normalized (LOWESS [locally weighted polynomial regression] method) with GeneSpring version 7.2 software (Agilent Technologies). For the pathway analysis, we used the GenMAPP and MAPPFinder software package.20, 21 The GenMAPP program contains many pathway maps that can be associated with imported microarray data. The MAPPFinder program, which links gene expression data to the pathway maps, can calculate the z score (standardized difference score) and the percentage of genes measured that meet user-defined criteria (±25% in the change fold in our analysis). With the z score and the percentage, the pathways were ranked according to the relative change in the gene expression. The microarray data sets have been submitted to the Genome Expression Omnibus Database (available at http://www.ncbi.nlm.nih.gov/geo/) under series GSE5852.

Quantitative Real-Time Polymerase Chain Reaction (PCR).

The reverse transcription of 100 ng of total RNA (the same sample used for the microarray analysis) was performed with Oligo(dT)12–18 primer and SuperScript III reverse transcriptase (Invitrogen, Carlsbad, CA). PCR was performed on an ABI-Prism 7900HT (Applied Biosystems). The specific PCR primers and TaqMan probe used in this study were obtained from Applied Biosystems. The PCR conditions were 1 cycle at 50°C for 2 minutes and at 95°C for 10 minutes followed by 40 cycles at 95°C for 15 s and at 60°C for 1 minute.

Western Blot Analysis.

Livers were homogenized in a buffer containing 20 mM trishydroxymethylaminomethane-HCl (pH 7.5), 5 mM ethylene diamine tetraacetic acid, 1% NP-40, and a protease inhibitor cocktail (Pierce, Rockford, IL). Homogenated proteins (30 μg/lane) were separated by 4%–20% gradient sodium dodecyl sulfate–polyacrylamide gels (Daiichi Chemicals, Tokyo, Japan) and resolved with 130 V over 2 hours. Proteins were transferred to polyvinylidene difluoride membranes (Millipore, Billerica, MA) with a Transblot apparatus (Bio-Rad, Hercules, CA). The membranes were blocked in a buffer containing 5% nonfat milk, 50 mM trishydroxymethylaminomethane (pH 7.6), 150 mM NaCl, and 0.1% Tween 20 [trishydroxymethylaminomethane-buffered saline Tween 20 (TBS-T)] for 12 hours at 4°C. They were then probed with the monoclonal anti–4-hydroxy-2-nonenal (4-HNE) antibody (NOF, Tokyo, Japan) at a 1:200 dilution, with the polyclonal anti–insulin receptor substrate 2 (IRS-2) antibody (Millipore) at a 1:500 dilution, or with the polyclonal anti–glyceraldehyde 3- phosphate dehydrogenase antibody (Santa Cruz Biotechnology, Santa Cruz, CA) at a 1:3000 dilution in 5% bovine serum albumin TBS-T for 12 hours at 4°C. After the membranes had been washed in TBS-T, the blots were incubated with the horseradish peroxidase–linked secondary antibody (Cell Signaling Technology, Beverly, MA). Signals were detected with a chemiluminescence detection system (ECL Plus, GE Healthcare Bio-Sciences, Piscataway, NJ) and exposure to X-ray film. The hepatic 4-HNE contents were quantified with WinROOF version 5.7 (Mitani Shoji).

Statistical Analysis.

The results are shown as the means ± the standard error of the mean (SEM). The data were analyzed with a 1-way analysis of variance to compare the means of all groups. The Bonferroni multiple-comparison procedure was used to determine which pairs of means were different. Differences in the histological scores between the Ath and Ath+HF groups were compared with the Mann-Whitney U test. All calculations were performed with SPSS version 12.0 software for Windows (SPSS, Chicago, IL).

Results

Ath Diet Causes Hepatomegaly and Liver Injury.

Hepatomegaly was observed in the Ath and Ath+HF groups (Fig. 1A). As shown in Fig. 1B, the livers of mice fed the Ath diet were grossly enlarged and pale in color. The serum ALT level was also elevated in the Ath and Ath+HF groups (Fig. 1C). Splenomegaly, frequently associated with cirrhosis, was detected in the Ath and Ath+HF groups at 24 weeks.

Figure 1.

Effects of 3 diets on the liver weight and morphology and serum ALT. (A) Liver weight with respect to the body weight (BW) of control mice fed standard chow (white bars), the Ath diet (gray bars), or the Ath+HF diet (black bars) after 6 or 24 weeks. (B) Photograph of livers after 12 weeks of feeding with the standard chow, Ath diet, or Ath+HF diet (scale bars: 10 mm). (C) Serum ALT levels after 6 or 24 weeks. The values represent the means ± the SEM. The number of animals per group is indicated in or just above the bars. *P < 0.05 and **P < 0.01 versus the control group.

Effect of the Ath Diet on the Plasma Lipid Levels and Hepatic Lipid Content.

As shown in Table 2, the plasma cholesterol levels were significantly elevated in the Ath diet group after both 6 and 24 weeks. An HPLC analysis revealed that the Ath and Ath+HF diets markedly increased the very low density lipoprotein-cholesterol (VLDL-C), low-density lipoprotein-cholesterol (LDL-C), and small dense LDL-C fractions, whereas they lowered high-density lipoprotein-cholesterol (HDL-C) in comparison with the controls (Fig. 2). As reported previously, we also confirmed atherosclerotic lesions in the mice fed the Ath and Ath+HF diets but not in the mice fed normal chow (data not shown).

Table 2. Effects of the 3 Diets on Body Weight and Lipid Levels at 6 or 24 Weeks of Feeding
Diet type6 weeks24 weeks
Control (n = 3)Ath (n = 3)Ath+HF (n =3)Control (n = 4)Ath (n = 6)Ath+HF (n = 6)
  • Data are means ± SEM. Significantly different from control value:

  • *

    P < 0.05;

  • **, *

    P < 0.01. Abbreviations: FFA, free fatty acid; HOMA-IR, homeostasis model assessment insulin resistance; N.D., Not determined.

Body weight (g)24.7 ± 0.524.9 ± 0.423.2 ± 0.429.0 ± 0.728.1 ± 1.826.4 ± 1.1**
Epididymal fat pad weight (g)0.14 ± 0.010.15 ± 0.010.15 ± 0.020.25 ± 0.010.09 ± 0.01**0.17 ± 0.01**
Plasma triglycerides (mg/dL)68.0 ± 5.254.6 ± 10.124.0 ± 3.6**41.5 ± 4.633.8 ± 3.520.8 ± 1.7*
Plasma total cholesterol (mg/dL)85.0 ± 8.5173.6 ± 5.3**168.1 ± 5.3**84.6 ± 3.8257.1 ± 10.8**204.4 ± 8.8**
Plasma free cholesterol (mg/dL)23.9 ± 2.245.9 ± 2.7**40.2 ± 0.7**17.3 ± 0.347.5 ± 5.6**31.3 ± 2.3*
Plasma FFA (mEq/L)0.58 ± 0.090.75 ± 0.100.46 ± 0.070.48 ± 0.040.48 ± 0.040.24 ± 0.06*
Plasma insulin (μU/mL)N.D.N.D.N.D.6.2 ± 0.311.2 ± 1.713.8 ± 2.9
Fasting blood glucose (mg/dL)N.D.N.D.N.D.93 ± 485 ± 493 ± 8
HOMA-IRN.D.N.D.N.D.1.4 ± 0.12.3 ± 0.33.1 ± 0.4*
Hepatic triglycerides (μg/mg protein)67.2 ± 10.489.3 ± 19.7150.8 ± 21.6*148.6 ± 20.952.8 ± 17.4*64.5 ± 9.2*
Hepatic total cholesterol (μg/mg protein)42.0 ± 2.8206.8 ± 22.5**342.8 ± 40.8**34.3 ± 2.2143.5 ± 24.1*192.8 ± 25.0**
Hepatic free cholesterol (μg/mg protein)22.1 ± 3.930.4 ± 4.252.6 ± 6.6*19.9 ± 2.333.0 ± 2.6*30.9 ± 5.9
Hepatic FFA (μEq/mg protein)45.6 ± 4.052.6 ± 3.363.0 ± 2.8*53.1 ± 1.881.1 ± 6.3*83.6 ± 8.7*
Figure 2.

HPLC analysis of plasma lipoproteins: fractionation by HPLC of cholesterol and free TG from mouse plasma after 24 weeks on the diet. The chromatograms for 1 representative sample are presented. The chylomicron, VLDL-C, LDL-C, and HDL-C fractions are labeled C, V, L, and H, respectively. The shaded fractions correspond to the level of small dense LDL-C.

The Ath and Ath+HF diets accumulated cholesterol in the liver after both 6 and 24 weeks. In addition to cholesterol, TG and FFA also accumulated with the Ath+HF diet. In comparison with hepatic lipid levels after 6 weeks, cholesterol and TG decreased in the livers of mice fed the Ath and Ath+HF diets after 24 weeks, and this indicated the progression of extensive hepatic fibrosis and impaired hepatocellular function. As is often found in patients with advanced liver disease, the serum ALT levels decreased with the progression of hepatic fibrosis, probably because of the impaired regeneration of hepatocytes and the production of ubiquitous liver enzymes.22

Effects of the Ath Diet on Systemic or Hepatic Insulin Resistance.

GTT and ITT after 12 weeks showed that the mice fed the Ath diet were remarkably sensitive to insulin (Fig. 3A,B). This ameliorating effect on the glucose tolerance and insulin sensitivity may be attributable to decreased adipose tissue in the mice fed the Ath or Ath+HF diet (Table 2). Therefore, we next evaluated the hepatic insulin sensitivity. For this purpose, we performed the pyruvate challenge test, an established method for evaluating hepatic insulin sensitivity,15, 16 by investigating the rise in blood glucose in response to the administration of pyruvate, a precursor for gluconeogenesis. The mice fed the Ath+HF diet showed an increased rise in the blood glucose concentration after pyruvate injection (Fig. 3C) compared with the mice fed the standard chow, and this suggested that the Ath+HF diet causes hepatic insulin resistance. Furthermore, as shown in Table 2, the homeostasis model assessment of insulin resistance (HOMA-IR) was significantly higher in the mice fed the Ath+HF diet than in the control mice. The expression of messenger RNA (mRNA) for phosphoenolpyruvate carboxykinase, the rate-controlling enzyme of gluconeogenesis for which the expression is negatively regulated by insulin, was significantly higher in the mice fed the Ath+HF diet than in the control mice (Fig. 3D). These results suggest that the Ath+HF diet causes hepatic insulin resistance.

Figure 3.

Evaluation of glucose tolerance and insulin sensitivity. (A) GTT and (B) ITT after 12 weeks on standard chow (n = 4), the Ath diet (n = 5), or the Ath+HF diet (n = 5). (C) Pyruvate challenge test after 6 weeks on standard chow (n = 4), the Ath diet (n = 4), or the Ath+HF diet (n = 4). The area under the curve (AUC) of the blood glucose levels during the pyruvate challenge test was calculated. (D) mRNA levels of phosphoenolpyruvate carboxykinase genes in the livers of mice fed standard chow (white bar; n = 3), the Ath diet (gray bar; n = 3), or the Ath+HF diet (black bar; n = 3) after 12 weeks. The gene expression was normalized with eukaryotic 18S ribosomal RNA. The degree of change in the gene expression was based on the mean expression levels in control mice. The values represent the means ± the SEM. *P < 0.05 and **P < 0.01 versus the control group. #P < 0.05 versus the Ath group.

Ath Diet Induces Steatosis, Fibrosis, and Cellular Ballooning of the Liver.

Figure 4 shows the time course of histological changes in the livers of mice fed the Ath or Ath+HF diet. The Ath diet induced progressive steatosis, inflammation, and fibrosis in a time-dependent manner from 6–24 weeks. Moreover, cellular ballooning, an important histological feature for the diagnosis of human NASH, was observed in the Ath group after 24 weeks. The addition of a high-fat component to the Ath diet accelerated the development of steatosis, inflammation, and fibrosis. Furthermore, before the Ath group, cellular ballooning was already observed in the Ath+HF group after 12 weeks. The hepatic hydroxyproline content, an indicator of collagen accumulation in the liver, increased significantly in the mice fed the Ath diet and increased further in the mice fed the Ath+HF diet (Fig. 4C). Therefore, the Ath diet induces steatohepatitis, and the addition of a high-fat component exacerbates the histological severity of steatohepatitis and hepatic insulin resistance.

Figure 4.

Representative liver histology, scoring, and occurrence of hepatocyte ballooning. (A) Liver sections were stained with H&E, Azan, and Sirius red after 6, 12, and 24 weeks. The arrows indicate infiltration of the inflammatory cells in the hepatic parenchyma. The characteristic initial pattern of fibrosis in steatohepatitis is collagen deposition, as identified by blue and red staining. The original magnification was ×200. The scale bars represent 10 μm. Ballooning hepatocytes were seen only in the Ath and Ath+HF groups (shown in the inset). (B) The absolute percentage of the macrovesicular fat droplet area in the H&E-stained area was determined to evaluate steatosis. The values represent the means ± the SEM. The number of animals per group is indicated just above the points. ##P < 0.01 versus the Ath group. Inflammation and fibrosis scores were assigned in a blinded fashion to H&E-stained samples for inflammation and to Azan-stained samples for fibrosis. The criteria for each score are described in the Materials and Methods section. Differences in the inflammation and fibrosis histological scores between the Ath and Ath+HF groups were compared with the Mann-Whitney U test. The control, Ath, and Ath+HF groups are labeled C (white circles), A (gray triangles), and A+H (black squares), respectively. (C) The hydroxyproline content was determined in the livers of mice fed standard chow (white bars; n = 3), the Ath diet (gray bars; n = 3), or the Ath+HF diet (black bars; n = 3) at 6, 12, and 24 weeks. The values represent the means ± the SEM. **P < 0.01 versus the control group. ##P < 0.01 versus the Ath group.

High-Fat Component Further Enhances the Activation of Hepatic Stellate Cells (HSCs) with the Ath Diet.

The major sources of collagen and other extracellular matrix proteins in liver fibrosis are HSCs.23 In response to stimuli such as oxidative stress and inflammatory cytokines, HSCs become activated and transform into proliferative fibrogenic cells.24 We performed an immunohistochemical analysis of α-SMA, an activated HSC marker, at different times. Representative photomicrographs of liver sections stained with the anti–α-SMA antibody are shown in Fig. 5A. We quantified the areas in the liver sections positive for α-SMA morphometrically in the 3 groups at different times as described (Fig. 5A, lower panel). The activation of HSCs was promoted in the livers of mice fed the Ath diet in a time-dependent manner from 6–24 weeks and was further accelerated by the addition of a high-fat component to the Ath diet.

Figure 5.

Activation of HSCs and oxidative stress in the livers of mice fed the Ath or Ath+HF diet. (A) Hepatic α-SMA–positive cells (indicated by arrows) were detected by immunohistochemical staining at 6, 12, or 24 weeks. The original magnification was ×200. The scale bars represent 10 μm. The α-SMA–positive area was quantified morphometrically in the liver sections of mice fed standard chow (white bar; n = 3), the Ath diet (gray bar; n = 3), or the Ath+HF diet (black bar; n = 3) at different times, as described in the Materials and Methods section. (B) Western blot of 4-HNE–modified proteins in the liver after 24 weeks. The hepatic content of 4-HNE–modified proteins was quantified in mice fed standard chow (white bar; n = 4), the Ath diet (gray bar; n = 4), or the Ath+HF diet (black bar; n = 4), as described in the Materials and Methods section. (C) Hepatic protein carbonyls were determined in the mice fed standard chow (white bar; n = 3), the Ath diet (gray bar; n = 4), or the Ath+HF diet (black bar; n = 4) after 24 weeks, as described in the Materials and Methods section. The values represent the means ± the SEM. *P < 0.05 and **P < 0.01 versus the control group. #P < 0.05 and ##P < 0.01 versus the Ath group.

To evaluate oxidative stress causing HSC activation, we assayed proteins modified with 4-HNE, which is a major aldehyde end product of membrane lipid peroxidation due to oxidative stress (Fig. 5B). In concert with the increase in α-SMA–positive cells, 4-HNE–modified proteins accumulated in the livers of mice fed the Ath diet and further accumulated in those of mice fed the Ath+HF diet. In addition to 4-HNE–modified proteins, hepatic protein carbonyls, another marker of oxidative stress, also increased with the Ath and Ath+HF diets (Fig. 5C). These results are consistent with the observation that the Ath+HF diet induced more severe inflammation and fibrosis than the Ath diet.

Gene Expression in the Livers of Mice Fed the Ath Diet.

To address the molecular basis of Ath diet–induced steatohepatitis, we performed a microarray analysis, using livers at early (6 weeks) and precirrhosis stages (24 weeks) in the development of steatohepatitis. We screened 103 pathways determined with GenMAPP and extracted the metabolic pathways significantly altered in the livers of the mice fed the Ath and Ath+HF diets (Table 3). In the livers of the mice fed the Ath diet, genes involved in the inflammatory response and p38 mitogen-activated protein kinase (MAPK) signaling pathway were up-regulated significantly, whereas genes involved in fatty acid β-oxidation were down-regulated significantly in the early stage (6 weeks), and this was followed by coordinated up-regulation of the genes involved in fibrogenesis, such as the transforming growth factor β (TGF-β) signaling pathway, in the late stage (24 weeks). Adding the high-fat component to the Ath diet accelerated the up-regulation of the genes involved in inflammation (electron-transport chain, p38 MAPK signaling pathway, and Fas pathway and stress induction) and fibrogenesis (TGF-β signaling pathway and matrix metalloproteinases). Of these pathways altered in the models, we present the expression levels of representative genes involved in lipid metabolism, inflammation, oxidative stress, and fibrogenesis in Fig. 6 and Supplementary Table 2.

Table 3. Biological Pathways of Liver Genes Regulated by the Ath or Ath+HF Diets After 6 or 24 Weeks
Pathway NameNumber of Genes ChangedNumber of Genes MeasuredZ ScorePermuted P Value
Ath diet    
Up-regulated at 6 weeks    
Inflammatory Response23413.22< 0.01
DNA replication Reactome21412.560.010
Cell Cycle-G1 to S control Reactome32682.560.016
G1 to S cell cycle Reactome32682.560.016
RNA transcription Reactome20402.360.036
p38 MAPK signaling15282.380.037
Down-regulated at 6 weeks    
Amino Acid Metabolism23452.94< 0.01
Cholesterol Biosynthesis11153.56< 0.01
Complement and Coagulation Cascades29593.04< 0.01
Mitochondrial fatty acid betaoxidation11163.28< 0.01
Blood Clotting Cascade11182.770.012
Unsaturated Fatty Acid Beta Oxidation562.780.014
Biogenic Amine Synthesis8142.130.042
Krebs-TCA Cycle14292.030.045
Up-regulated at 24 weeks    
mRNA processing binding Reactome1964385.91< 0.01
TGF Beta Signaling Pathway621244.37< 0.01
Translation Factors27493.50< 0.01
Complement Activation Classical9152.340.021
Down-regulated at 24 weeks    
GPCRDB Other521473.58< 0.01
Small ligand GPCRs11193.61< 0.01
Synthesis and Degradation of Ketone Bodies443.66< 0.01
Mitochondrial fatty acid betaoxidation9163.16< 0.01
Cholesterol Biosynthesis8152.79< 0.01
Metabotropic glutamate pheromone6102.780.020
Ath + HF diet    
Up-regulated at 6 weeks    
Electron Transport Chain35824.93< 0.01
mRNA processing binding Reactome1204343.64< 0.01
Translation Factors20493.48< 0.01
p38 MAPK signaling pathway13283.36< 0.01
Unsaturated Fatty Acid Beta Oxidation462.780.018
Matrix Metalloproteinases9242.030.034
TGF Beta Signaling Pathway351242.080.039
Fas pathway and stress induction411492.070.042
Down-regulated at 6 weeks    
Focal adhesion561863.51< 0.01
Steroid Biosynthesis8124.06< 0.01
Complement and Coagulation Cascades20592.70< 0.01
G Protein Signaling26832.620.013
Calcium regulation in cardiac cells411452.540.014
Cholesterol Biosynthesis7152.600.016
Up-regulated at 24 weeks    
Translation Factors21493.993< 0.01
mRNA processing binding Reactome1214374.055< 0.01
p38 MAPK signaling pathway12283.016< 0.01
TGF Beta signaling pathway351242.0770.039
Down-regulated at 24 weeks    
Amino Acid Metabolism19453.891< 0.01
Urea cycle and metabolism of amino groups10203.472< 0.01
Striated muscle contraction16423.080< 0.01
Steroid Biosynthesis6122.6890.015
Small ligand GPCRs8192.5130.020
Glycolysis and Gluconeogenesis14412.4020.023
Figure 6.

Quantitative real-time PCR for representative genes involved in steatohepatitis. The mRNA levels of genes for SREBP-1c, FAS, PPARα, CPT-1a, TNF-α, p22phox, p47phox, gp91phox, TGF-β1, procollagen type I alpha 1 (Col1a1), procollagen type I alpha 2 (Col1a2), procollagen type IV alpha 1 (Col4a1), and PAI-1 in the livers of mice fed standard chow (n = 3), the Ath diet (n = 3), or the Ath+HF diet (n = 3) were quantified with real-time PCR after 6 and 24 weeks. The RNA samples used for real-time PCR were the same as those used for the microarray analysis. The gene expression was normalized with eukaryotic 18S ribosomal RNA. The degree of change in the gene expression was based on the mean expression levels of control mice at 6 weeks. The values represent the means ± SEM. *P < 0.05 and **P < 0.01 versus the control group. #P < 0.05 and ##P < 0.01 versus the Ath group.

In the livers of mice fed the Ath diet, the expression of genes for fatty acid synthesis, such as sterol regulatory element binding protein 1c (SREBP-1c), a transcriptional regulator of fatty acid synthesis,25 and fatty acid synthase (FAS), was coordinately up-regulated. In contrast, the expression levels of genes for the mitochondrial fatty acid β-oxidation pathway were coordinately repressed in concert with a decrease in the expression of peroxisome proliferator-activated receptor α (PPARα), a transcriptional up-regulator of fatty acid β-oxidation in the liver.26 It is recognized that mitochondrial β-oxidation and the levels of carnitine palmitoyltransferase 1a (CPT-1a) and PPARα expression are increased compensatively in the livers of patients with NAFLD27, 28 and obese-diabetic (ob/ob) mice with severe steatosis of the liver.29 Therefore, although the levels of PPARα and CPT-1a mRNA expression in the Ath+HF group were higher than those in the Ath group, it may not have been enough to metabolize the excessive fatty acids from the high-fat component and intrahepatic fatty acid synthesis.

It is believed that oxidative stress due to the generation of reactive oxygen species (ROS) or decreased antioxidant defenses is directly involved in the development of steatohepatitis.30 The expression levels of genes for the reduced-form nicotinamide adenine dinucleotide phosphate (NADPH) oxidase complex, an important source of ROS,31 were coordinately elevated in mice fed the Ath diet and further up-regulated in mice fed the Ath+HF diet.

The Ath diet has previously been reported to induce the expression of genes for inflammation.32, 33 Our results further demonstrate that inflammatory cytokines, such as tumor necrosis factor α (TNF-α), chemokines, and their receptors, are up-regulated in mice fed the Ath diet.

The Ath diet also induced genes involved in collagen accumulation, especially after 24 weeks. At 6 weeks, the expression levels of collagen genes were higher in the Ath+HF group than in the Ath group (Fig. 6). In addition, the expression levels of genes for TGF-β and plasminogen activator inhibitor 1 (PAI-1), key inducers of fibrogenesis, were dramatically up-regulated in the Ath+HF group compared with the Ath group at 24 weeks. These results support the finding that the Ath+HF diet induces more rapid progression of steatohepatitis than the Ath diet.

Discussion

Whether cholesterol, TG, or FFA contributes to the development of NASH remains controversial.34 Because the feeding of cholesterol and cholic acid, which are the main components of the Ath diet, leads to the additive accumulation of cholesterol in the liver, the main pathology in Ath diet–induced steatohepatitis is caused by cholesterol-induced toxicity.35

In this study, we have shown that Ath diet–induced steatohepatitis with atherosclerosis is a better experimental model of human NASH for the following reasons: (1) this model seems to be a more physiological dietary model of NASH than existing animal models, which require genetic defects, chemical agents such as carbon tetrachloride, or the depletion of nutrients, such as the MCD diet–induced model; (2) the liver pathology involves steatohepatitis with cellular ballooning, a necessary histological feature defining human NASH; (3) the addition of a high-fat component to the Ath diet causes hepatic insulin resistance and promotes oxidative stress, the activation of HSCs, and all components of the liver pathology of NASH (steatosis, inflammation, fibrosis, and cellular ballooning); and (4) there is a molecular signature indicative of lipid-induced oxidative stress in the liver, which may play a causal role in the development of steatohepatitis.

To diagnose human NASH, cellular ballooning, in addition to simple steatosis and inflammatory cell infiltration, is one of the most important pathological features.36 However, ballooning degeneration has scarcely been determined in the existing animal models, including mice fed the MCD diet. We believe that our study is the first to report that cellular ballooning is frequently induced in the livers of mice fed the Ath diet.

Recently, we proved that insulin resistance accelerates the pathological development of steatohepatitis experimentally.11 In this study, on the basis of the results of the pyruvate challenge test and HOMA-IR, we concluded that the Ath+HF diet causes hepatic insulin resistance. It is known that the excessive accumulation of FFAs caused by the overexpression of lipoprotein lipase37 and an increase in SREBP-1c–regulated lipogenesis38 leads to impaired tyrosine phosphorylation of IRS-1 and IRS-2, resulting in hepatic insulin resistance. Furthermore, the up-regulation of SREBP-1c–regulated lipogenesis contributes to the development of insulin resistance via the down-regulation of IRS-2 in the liver.39, 40 Indeed, in our study, the induction of lipoprotein lipase and SREBP-1c and the repression of IRS-2 were detected in the livers of mice fed the Ath diet (Fig. 7). Moreover, the up-regulation of stearoyl–coenzyme A desaturase 1, an enzyme that catalyzes the synthesis of monounsaturated fatty acids, might contribute to lipid accumulation and insulin resistance in the liver, as reported in skeletal muscle.41 Therefore, the cholesterol-induced and TG-induced alteration of fatty acid metabolism may cause hepatic insulin resistance in this model of steatohepatitis.

Figure 7.

The Ath and Ath+HF diets decreased the mRNA and protein levels of IRS-2 in the liver. (A) mRNA levels of the IRS-2 genes in the livers of mice fed standard chow (white bar; n = 3), the Ath diet (gray bar; n = 3), or the Ath+HF diet (black bar; n = 3) after 12 weeks. The values represent the means ± the SEM. *P < 0.05 versus the control group. #P < 0.05 versus the Ath group. (B) Western blot of IRS-2 in the livers of mice fed the standard chow, Ath diet, or Ath+HF diet after 12 weeks.

Another possible cause of the liver pathology in our model is lipid-induced oxidative stress and its downstream events, as we identified an accumulation of 4-HNE and protein carbonyls, the activation of stellate cells, and hepatic inflammation with cell ballooning. In this study, in addition to cholesterol, the accumulation of TG and FFAs by the addition of a high-fat component accelerated oxidative stress, possibly via the up-regulation of genes involved in the generation of ROS, such as the NADPH oxidase complex, and the down-regulation of genes for antioxidant enzymes. While we were preparing this article, Mari et al.35 reported that the mitochondrial loading of free cholesterol, but not TG and FFA, decreases mitochondrial glutathione and sensitizes it to the TNF-α–mediated apoptosis of hepatocytes. Therefore, the different kinds of accumulated lipids may cause oxidative stress in the liver additively in different ways. In patients with NASH, impaired glutathione metabolism and antioxidant enzyme activity probably cause an increase in oxidative stress.42

The Ath diet induces an Ath lipid profile, including an increase in small dense LDL-C, which is highly susceptible to oxidation, and then leads to oxidized low-density lipoprotein (LDL), which induces an inflammatory response in endothelial cells.43 In the livers of mice fed the Ath diet, the expression levels of genes for CD36 antigen and scavenger receptor type B member 1, which are receptors for oxidized LDL,44 tended to be up-regulated (Supplementary Table 2). Therefore, it might be possible that up-regulated receptors for oxidized LDL enhance the uptake of increasing levels of small dense LDL-C and contribute to inflammation in the liver.

In response to the lipid-induced oxidative stress, genes involved in fibrogenesis were coordinately up-regulated. Indeed, the hepatic expression of TNF-α and NADPH oxidase complex genes preceded that of fibrogenic genes, and this suggested that inflammation precedes the fibrogenic process in our models. The expression of TGF-β and PAI-1 genes was up-regulated dramatically, especially in the Ath+HF group. PAI-1 is a key factor in matrix remodeling, and the gene is highly induced in response to TGF-β.45 Urokinase plasminogen activator generates plasmin, and this process is inhibited by PAI-1. Plasmin degrades the extracellular matrix both directly and by activating matrix metalloproteinases.46 Therefore, PAI-1 inhibits collagenolysis by inhibiting the generation of plasmin in the liver. Consequently, the inhibition of collagenolysis, in addition to the overall up-regulation of collagen genes, might contribute to hepatic fibrosis in this model.

In summary, we report that the Ath diet induces steatohepatitis with cellular ballooning via cholesterol-induced oxidative stress and hepatic insulin resistance. Adding a high-fat component further aggravates oxidative stress and steatohepatitis, possibly by inducing insulin resistance and down-regulating genes for antioxidant enzymes. This model suggests the critical and different roles of cholesterol, TG, and FFAs in causing oxidative stress and insulin resistance leading to steatohepatitis and provides a system for screening therapeutic targets to treat NASH and atherosclerosis.

Acknowledgements

We thank Dr. Isao Usui, Dr. Hajime Ishihara, and Professor Toshiyasu Sasaoka of Toyama University for their technical advice concerning western blot analyses.

Ancillary