Hepatology Research 34 (2006) 92–98

Development of nonalcoholic steatohepatitis model throughcombination of high-fat diet and tetracycline with morbid obesity in mice

Makoto Itoa, Jun Suzukia, Minoru Sasakib, Keiko Watanabea,Shigeharu Tsujiokab, Yuki Takahashia, Akira Gomoria, Hiroyasu Hirosea,

Akane Ishiharaa, Hisashi Iwaasaa,∗, Akio Kanatania

a Tsukuba Research Institute, Banyu Pharmaceutical Co. Ltd., Tsukuba, Japanb Tsukuba Safety Assessment Laboratories, Banyu Pharmaceutical Co. Ltd., Tsukuba, Japan

Received 30 September 2005; received in revised form 2 December 2005; accepted 5 December 2005

Abstract

Nonalcoholic steatohepatitis (NASH) is a chronic liver disease closely associated with obesity, type 2 diabetes and hyperlipidemia. Forf important.B achievem ycline wasi e contenta d by thisc ammatoryc r furtheru©

K

1

f[taagTts

cT

har--in

larlyws is

ry”rsd to

atosisides,osedm-nes

1d

urther understanding of NASH, development and characterization of appropriate animal models with metabolic abnormalities isased on the “two hit theory”, we tried to develop a new murine model of NASH with metabolic abnormalities. For the first hit toetabolic abnormalities, a high-fat diet (HFD: 60 cal% fat) was fed to C57BL/6 mice for 10 weeks. For the second hit, 30 mg/kg tetrac

njected intraperitoneally once daily for 10 days. The HFD-fed mice treated with tetracycline showed robust increases in triglyceridnd expression levels of proinflammatory cytokine mRNAs in the liver. In addition, plasma ALT levels were significantly elevateombinational treatment. Furthermore, histological examination demonstrated that combinational treatment induced multifocal inflellular infiltration in the livers of all mice, and thus caused mild steatohepatitis. The HFD–tetracycline model could be useful fonderstanding NASH.2005 Elsevier Ireland Ltd. All rights reserved.

eywords: NASH; Metabolic abnormality; Fatty liver; Animal models

. Introduction

Several lines of evidence indicate that nonalcoholicatty liver disease (NAFLD) is a metabolic disorder31–33,41,51,54]. Nonalcoholic steatohepatitis (NASH) is aype of NAFLD that critically requires medical treatment,s recent investigations have demonstrated that NASH isprodrome for end-stage liver diseases, such as crypto-

enic cirrhosis and hepatocellular carcinoma[4,5,21,43,48].o understand the pathogenic mechanisms and to developherapeutic methods, several animal models of NASH,uch as rodents fed methionine–choline-deficient (MCD)

∗ Corresponding author at: Tsukuba Research Institute, Banyu Pharma-eutical Co. Ltd., 3 Okubo, Tsukuba, Ibaraki 300-2611, Japan.el.: +81 29 877 2000; fax: +81 29 877 2027.

E-mail address: hisashiiwaasa@merck.com (H. Iwaasa).

diet or genetically manipulated mice, have been cacterized[9,17,18,22,23,26,47,52]. However, NASH models with metabolic abnormalities similar to thosehumans have rarely been reported to date, particuin mice [2,6,9,13,30,42]. Therefore, development of neanimal models that accurately mimic human patientimportant.

As a pathogenic mechanism of NASH, the “two hit theohas generally prevailed[11,39]. The first hit refers to factothat promote hepatic steatosis; the fatty liver is believehave increased vulnerability to oxidative stress[14,53]. Thesecond hit refers to factors that aggravate hepatic steto steatohepatitis. To date, factors such as lipid peroxlipopolysaccharides and ATP depletion have been propas second hits[16,20,28]. These factors accelerate inflamation via hepatic production of proinflammatory cytoki[36,46,50].

386-6346/$ – see front matter © 2005 Elsevier Ireland Ltd. All rights reserved.oi:10.1016/j.hepres.2005.12.001

M. Ito et al. / Hepatology Research 34 (2006) 92–98 93

In this study, we attempted to develop a novel animalmodel of NASH that mimics the clinical features of NASHpatients, i.e. morbid obesity, based on the two hit theory. Forthe first hit, we employed a high-fat diet (HFD) to induce hep-atic steatosis and metabolic abnormalities in mice. Mice werethen treated with tetracycline, which is reported to induceoxidative stress in the liver, in order to stimulate inflamma-tion [3,15,28,29]. Herein, we describe the phenotypes of theHFD–tetracycline model as a new animal model of NASH.

2. Materials and methods

2.1. Animals

Male C57BL/6J mice (Clea Japan, Tokyo, Japan) wereused. Mice were housed individually in plastic cages thatwere kept at 23± 2◦C at a relative humidity of 55± 15%and a 12:12-h light–dark cycle. Mice had ad libitum accessto food and tap water. All experimental procedures wereconducted in accordance with the Japanese PharmacologicalSociety Guidelines for Animal Use. Experimental protocolswere approved by the in-house institutional animal care anduse committee.

2.2. Experimental designs

eref igh-f for1 car-b wasf odyw lyi t 0( htsw suredu erO thed plesw venac xsan-g ces.P pt in1 form ozeni

2

andt ter-m n),r kits( erem hi,T

2.4. Measurement of hepatic biochemical parameters

Liver specimens were homogenized with a Poly-tron homogenizer (Kinematica, Littau-Lucerne, Switzer-land) in 50 mM HEPES buffer, pH 7.4, containing5 mM MgCl2, 1 mM EDTA, 100 mM KCl and Proteaseinhibitor cocktail (Roche Diagnostics, Mannheim, Ger-many). Neutral lipids in the homogenates were extracted withchloroform–methanol (2:1, vol/vol), and triglycerides (TGs)and total cholesterols (T. Chol) were measured by Deter-miner L TGII and L TCII (Kyowa Medix, Tokyo, Japan),respectively. Glutathione (GSH) content was measuredusing a BIOXYTECH GSH-400 kit (Oxis Research, OR,USA).

2.5. Measurement of mRNA levels in the liver

Real-time quantitative RT-PCR using an ABI PRISM7900HT Sequence Detection system (Applied Biosystems,Foster City, CA) was performed in order to determinethe levels of tumor necrosis factor (TNF�), IL-1�, �-smooth muscle actin (�-SMA) and uncoupling protein-2(UCP2) mRNA in the liver. Amount of 18S rRNA wasmeasured as an internal control. Total RNA was puri-fied using Isogen (Nippon Gene, Tokyo, Japan) and itsquantity and purity were determined by absorbance at2t ents( obeswCATS -3 ,5ACAACTT -wCGm Fa ems( anda rer’sp

2

ed-d in

After an acclimation period, mice at 6 weeks of age wed either a regular diet (RD: CE2, Clea Japan) or a hat diet (HFD: cat No. D12492, Research diet, NJ, USA)0 weeks. HFD contained 60 cal% fat (lard), 20 cal%ohydrates and 20 cal% protein. Each group of mice

urther divided into two groups matched for average beight. Mice in these groups (n = 5) were intraperitoneal

njected with tetracycline (Sigma–Aldrich, MO, USA) avehicle) or 30 mg/kg once daily for 10 days. Body weigere monitored during the study. Fat mass was measing the Minispec Fat & Lean Live Mice Analyzer (Brukptic, Ettilingen, Germany) with NMR technology. Onay after the 10-day tetracycline treatment, blood samere collected under isofluran anesthesia via the inferiorava for measurement of plasma parameters. After euination, livers were excised and divided into four pieieces for pathological observation were fixed and ke0% neutral buffered formalin. The remaining sampleseasurement of hepatic lipids and mRNA were snap-fr

n liquid nitrogen and stored at−80◦C.

.3. Measurement of plasma biochemical parameters

After plasma collection by centrifugation, glucoseotal cholesterol (T. Chol) levels were measured by Deiner GL-E and L TCII kits (Kyowa Medix, Tokyo, Japa

espectively. Insulin levels were measured with ELISAMorinaga, Kanagawa, Japan). ALT, AST and LDH weasured using a HITACHI Clinical Analyzer 7070 (Hitac

okyo, Japan).

60 and 280 nm. cDNA was synthesized from 1�g ofotal RNA using TaqMan Reverse Transcription ReagApplied Biosystems). Sequences of primers and prere as follows: for detection of IL-1� (forward, 5′-AACCAACAAGTGATATTCTCCATG-3′; reverse, 5′-G-TCCACACTCTCCAGCTGCA-3′; probe, 5′-FAM-CTG-GTAATGAAAGACGGCACACCCACC-TAMRA-3′), �-MA (forward, 5′-GAGCGTGAGATTGTCCGTCCGTGA′; reverse, 5′-AGCGTTCGTTTCCAATGGTG-3′; probe′-FAM-CCCTGGAGAAGAGCTACGAACTGCCTGA-T-MRA-3′), CYP2E1 (forward, 5′-TGACTTTGGCCGAC-TGTTC-3′; reverse, 5′- GAATCAGGAGCCCATATCTC-GA-3′; probe, 5′-FAM-TTGCAGGAACAGAGACCACC-GCACA-TAMRA-3′), CYP4A14 (forward, 5′-GAGGCT-TATCCACCAGTAATATCTGT-3′; reverse, 5′-GGGTA-GGAGCGTCCATCTG-3′; probe, 5′-FAM-TCGAGAGC-CAGCTCACCTGTCACCTT-TAMRA-3′) and UCP2 (forard, 5′-TCCTGAAAGCCAACCTCATGA-3′; reverse, 5′-GATGACGGTGGTGCAGA-3′; probe, 5′-FAM-AGAT-ACCTCCCTTGCCACTTCACTTCTG-TAMRA-3′). Theixture of primers and probes for measurement of TN�nd 18S rRNA was purchased from Applied BiosystTaqMan Gene Expression Assays). PCR reactionsnalyses were performed according to the manufacturotocols.

.6. Histology

Liver tissues from all animals in each group were embed in paraffin, sectioned at 3�m, stained with hematoxyl

94 M. Ito et al. / Hepatology Research 34 (2006) 92–98

Fig. 1. Changes in body weight (A) and fat mass (B). (A) Body weights were monitored during high-fat diet (HFD) (circle) or regular diet (RD) (triangle)administration with (closed) or without (open) tetracycline treatment. The 10-day tetracycline treatment was started on day 62. (B) Fat mass was measuredafter treatment with vehicle or tetracycline (TC).*** P < 0.001 vs. RD and vehicle treatment group.

and eosin (HE) and Azan stain, and examined microscopi-cally.

2.7. Statistical analysis

Data are expressed as mean values± S.E.M. Data wereanalyzed by one-way analysis of variance (ANOVA), fol-lowed by Bonferroni test for multiple comparisons. Differ-ences were considered significant whenP values were lessthan 0.05.

3. Results

3.1. Body weight and biochemical plasma parameters

Mice fed HFD showed body weight increases of over 7 gwhen compared with mice fed regular diet after 10 weeks(35.3± 1.4 g versus 27.8± 0.3 g) (Fig. 1A). The increasein body weight was mainly caused by fat accumulation(Fig. 1B). Injections of tetracycline into mice fed HFD sus-tained obese phenotypes, such as overweight and high-fatmass during the treatment period as compared with controlmice fed regular diet, although some body weight reductionwas observed in the mice treated with tetracycline (Fig. 1andTable 1).

With regard to blood chemistry, plasma glucose levelswere slightly, but significantly higher in mice fed HFD whencompared with those of the regular diet-fed group (Table 1).HFD exposure also doubled plasma cholesterol levels whencompared with regular diet-fed groups. Mice fed HFD exhib-ited hyperinsulinemia, while plasma insulin concentrationstended to decrease to normal levels following injection oftetracycline.

Interestingly, combinational treatment with HFD andtetracycline clearly elevated plasma ALT levels (Fig. 2A).Plasma AST and LDH were also increased in mice treatedwith a combination of HFD and tetracycline, while theother groups did not show any significant increase in theseenzymes, as compared to the regular diet-fed group with vehi-cle treatment (Fig. 2B andTable 1).

3.2. Hepatic parameters

Treatment with HFD or tetracycline alone produced amaximum three-fold increase in liver triglycerides (Fig. 2C).The HFD-fed condition with subsequent injections oftetracycline synergistically increased liver triglycerides by10-fold. Accumulation of hepatic cholesterol was similarlypronounced in the combinational treatment group (Fig. 2D).In contrast, hepatic GSH content was significantly decreasedby combinational treatment (Table 1).

TB conte -fa

Tetra

B 27.3±I 0.30±G 207±T 81 ±L 143±L 4.8±D

*

able 1ody weight, plasma biochemical parameters and hepatic glutathione

Regular diet

Vehicle

ody weight (g) 27.8± 0.3nsulin (ng/mL) 0.39± 0.04lucose (mg/dL) 197± 11otal cholesterol (mg/dL) 72± 6DH (IU/L) 121 ± 10iver GSH (�mol/g) 4.4± 0.2

ata are means± S.E.M. (n = 5).* P < 0.05 vs. regular diet and vehicle treatment group.

** P < 0.01 vs. regular diet and vehicle treatment group.** P < 0.001 vs. regular diet and vehicle treatment group.††† P < 0.001 vs. HFD and vehicle treatment group.

nts in mice treated with vehicle or tetracycline and fed regular or hight diet

High-fat diet

cycline Vehicle Tetracycline

0.1 35.3± 1.4*** 32.7± 1.3**

0.05 1.52± 0.21*** 0.60± 0.10†††

5 229± 5* 231 ± 12*

3 158± 5*** 127 ± 4††† ,***

13 126± 9 218± 13††† ,***

0.3 3.8± 0.2 3.6± 0.2**

M. Ito et al. / Hepatology Research 34 (2006) 92–98 95

Fig. 2. Plasma and hepatic parameters of mice treated with vehicle or tetracycline (TC) on regular diet (RD) or high-fat diet (HFD). Plasma ALT (A) and AST(B) levels, and hepatic TG (C) and T. Chol (D) content were measured as described in Section2. *** P < 0.001 vs. RD and vehicle treatment group;###P < 0.001vs. HFD and vehicle treatment group.

Expression levels of TNF� and IL-1� mRNA were quan-tified in order to address the inflammatory response in theliver (Fig. 3A and B). No changes in their expression wereobserved after treatment with HFD or tetracycline alone.Consistent with plasma ALT levels, injections of tetracyclinein HFD-fed mice synergistically increased the expression

levels of TNF� and IL-1� mRNA in the liver. Expression of�-SMA was also up-regulated by combinational treatment(Fig. 3C). These results suggest that inflammation accompa-nied by fibrogenesis was accelerated by the combination ofHFD and tetracycline. As often observed in human NASH,induction of CYP2E1 mRNA in the liver was observed in

F D), CY cline( P < 0.0v

ig. 3. Hepatic expression of TNF� (A), IL-1� (B), �-SMA (C), CYP2E1 (TC) on regular diet (RD) or high-fat diet (HFD).* P < 0.05,** P < 0.01,***

s. HFD and vehicle treatment group.

P4A14 (E) and UCP2 (F) mRNA in mice treated with vehicle or tetracy01 vs. RD and vehicle treatment group;#P < 0.05,##P < 0.01,###P < 0.001

96 M. Ito et al. / Hepatology Research 34 (2006) 92–98

Fig. 4. Representative photomicrographs of liver showing the effects of high-fat diet or/and tetracycline treatment. Hematoxylin and eosin-stained liver sectionsfrom regular diet and vehicle treatment (A), regular diet and tetracycline treatment (B), high-fat diet and vehicle treatment (C), and high-fat dietand tetracyclinetreatment (D) (scale bar = 50�m).

both HFD treatment groups, while induction of CYP4A14expression was not observed in the liver (Fig. 3D andE). Interestingly, the expression levels of hepatic UCP2mRNA were significantly increased in the combinationaltreatment group, whereas treatment with HFD or tetra-cycline alone increased UCP2 mRNA expression slightly(Fig. 3F).

3.3. Histology

Hepatocellular vacuolation consistent with hepatic steato-sis was observed in mice treated with HFD and tetracycline(Fig. 4D). The histomorphological changes were well corre-lated with hepatic TG content (Fig. 4). Tetracycline treatmentin HFD-fed mice also induced inflammatory cellular infil-tration in the livers of all mice tested here (Fig. 4D). Suchchanges were not observed in the livers of mice fed HFD with-out tetracycline treatment. Taken together, these observationssuggest that steatohepatitis was induced by the combinationof HFD and tetracycline. Apparent fibrosis was not micro-scopically detectable on Azan-stained sections of the liverin the HFD–tetracycline treatment group (data not shown),even though there was indicative of increased level of�-SMAmRNA. Mild capsulitis around the liver was also observed inmice treated with tetracycline regardless of HFD treatment(data not shown). In contrast, increases in plasma ALT ande ina-t toryp thert

4. Discussion

In the present study, we employed the “two hit theory”to develop a novel NASH model with obesity. We foundthat combinational treatment with HFD and tetracycline ledto steatohepatitis, as confirmed by histological observation.Plasma ALT and AST levels, and hepatic TG content,also supported the development of steatohepatitis. Hepaticexpression of TNF� and IL-1� as well as�-SMA wasnot induced by HFD exposure itself, whereas subsequentinjection of tetracycline stimulated expression of these keyfactors. The increases in these hepatic mRNA expressionsfurther supported the status of steatohepatitis accompaniedby fibrogenesis, even though apparent fibrosis was notobserved microscopically. Although elevation of ALTand hepatic cytokine expressions were relatively mildwhen compared to reported MCD-induced NASH models,the combinational treatment with HFD and tetracyclineproduced mild steatohepatitis with metabolic abnormalities,as is the case in NASH patients[23,51].

Under the experimental conditions employed here, tetra-cycline induced steatohepatitis in mice fed HFD but not inmice fed regular diet, thus suggesting the presence of hep-atic vulnerability in mice fed HFD[14,23]. Interestingly,hepatic GSH content was likely to be reduced in the HFD-vehicle treatment group and significantly reduced in theH SHi ther un-t evels

xpression of cytokines were observed only in the combional treatment group. Thus, the increases in inflammaarameters such as ALT were derived from NASH, ra

han capsulitis.

FD–tetracycline treatment group. It is well known that Gs vital in eliminating peroxidized materials. Therefore,eduction in GSH might be a result of consumption to coeract increased lipid peroxides. In support, expression l

M. Ito et al. / Hepatology Research 34 (2006) 92–98 97

of CYP2E1, which is a key enzyme to produce lipid per-oxides, was significantly increased in the HFD–tetracyclinetreatment group, like the case of patients[52]. The reductionsin GSH observed in the HFD-treated animals indicate hepaticvulnerability or intolerability to oxidative stress.

Tetracycline is known to induce oxidative stress by impair-ing mitochondrial function[3,15,28,29]. Development ofsteatohepatitis after combinational treatment suggested a keyrole of oxidative stress induced by tetracycline in this model.Interestingly, we confirmed that tetracycline induced expres-sion of UCP2 mRNA in the liver. Recent investigations haveshown that induction of UCP2 can cause mitochondrial dys-function and ATP depletion via dissipation of mitochondrialmembrane potential[7,19,24]. Based on these results, it ispossible that overexpression of hepatic UCP2 is a key medi-ator in the accumulation of oxidative stress induced by tetra-cycline. Further investigation remains necessary in order toreveal the role of UCP2 in NAFLD[2,10,19,35].

Another curious aspect of this model was the fact thattetracycline treatment synergistically enhanced hepatic lipidaccumulation in HFD-fed mice. It was recently reportedthat tetracycline inhibited VLDL release from the liver[29].Impairment of VLDL release by tetracycline may have a rolein the marked hepatic TG accumulation. Therefore, in addi-tion to inducing oxidative stress, tetracycline may stimulateTG accumulation and synergistically induce severe steatosisi les-t ightb lsor ctly[ ho-l tly.

n theM rest -m asesi tualN sedA ionsim ohepa ow-e dela ella ]I arev m-b ectso

stingat usess ns ear,t si-

tion in each study. In this aspect, the model reported byLieber et al. (2004) is unique[30]. Three-week exposure toHFD in rats caused severe steatohepatitis without significantchanges in body weight, as compared with the standard dietgroup. Considering the diverse etiology of human NASH, thedirect impact of diet composition as well as species differ-ences in the development of NASH are critical issues to beaddressed.

In summary, we confirmed that long-term exposure toHFD enhanced hepatic vulnerability to oxidative stress andthat subsequent short-term treatment with tetracycline stimu-lated development of NASH in mice. In addition, mice receiv-ing the combinational treatment exhibited obese phenotypesthat are commonly observed in NASH patients. Althoughsome issues remain to be overcome, the HFD–tetracyclinemodel will assist in advancing our understanding of NASH.

Acknowledgments

We thank T. Shirakura, Y. Tokushima and K. Taguchi fortechnical assistance.

References

oflin

ttyling

ofacol

cir-se. J

nd.

adsbesej.04-

n ofeple-

etesn the

itrict diet.

[ AM,lco-

[ arty?

[ b/ob)of

A

n the liver in the presence of HFD. The low plasma choerol levels observed in the HFD–tetracycline model me indicative of inhibited VLDL release. However, it was aeported that tetracycline inhibited phospholipase A2 dire37,38]. Therefore, the involvement of accumulated phospipids in this model is remaining to be addressed correc

A number of significant studies have been conducted iCD diet model, which exhibits similar histological featu

o those observed in NASH patients[18]. However, treatent with MCD diet decreases body weight and incre

nsulin sensitivity in rodents, which are rarely seen in acASH patients [25,32,40]. Therefore, genetically obeb/db mice on MCD diet provide an interesting model[42].lthough MCD diet exposure causes significant reduct

n body weight in this model, the MCD-db/db mice exhibitedetabolic abnormalities and demonstrated severe steattitis, as compared with the HFD–tetracycline model. Hver, db/db mice represent a leptin-signal deficient mond leptin itself is involved in fat accumulation as ws in immune responses in the liver[1,8,12,27,44,45,51.

n contrast, obese patients with lacking leptin signalsery rare[34]. Thus, HFD-fed models more closely resele NASH patients and provide insight into other aspf NASH.

Recent investigations have revealed several interespects of HFD-fed models[2,6,9,13,14,30,49]. Among

hese, only short-term HFD exposure (3–12 weeks) cateatohepatitis in rats[6,13,30]. Although the difference iusceptibility to NASH between rats and mice is unclhis may simply result from differences in HFD compo

-

[1] Asilmaz E, Cohen P, Miyazaki M, et al. Site and mechanismleptin action in a rodent form of congenital lipodystrophy. J CInvest 2004;113:414–24.

[2] Baffy G, Zhang CY, Glickman JN, Lowell BB. Obesity-related faliver is unchanged in mice deficient for mitochondrial uncoupprotein 2. Hepatology 2002;35:753–61.

[3] Bocker R, Strobel B, Hopf G, Estler CJ. Effect of toxic dosesprogesterone on hepatotoxic effects of tetracycline. Acta PharmToxicol 1985;56:327–30.

[4] Caldwell SH, Crespo DM. The spectrum expanded: cryptogenicrhosis and the natural history of non-alcoholic fatty liver diseaHepatol 2004;40:578–84.

[5] Caldwell SH, Crespo DM, Kang HS, Al-Osaimi AMS. Obesity ahepatocellular carcinoma. Gastroenterology 2004;127:S97–103

[6] Carmiel-Haggai M, Cederbaum AI, Nieto N, A high-fat diet leto the progression of non-alcoholic fatty liver disease in orats. FASEB J, Published on line November 2 2004 10.1096/f2291file.

[7] Chavin KD, Yang S, Lin HZ, et al. Obesity induces expressiouncoupling protein-2 in hepatocytes and promotes liver ATP dtion. J Biol Chem 1999;274:5692–700.

[8] Chen H, Charlat O, Tarraglia LA, et al. Evidence that the diabgene encodes the leptin receptor: identification of a mutation ileptin receptor gene indb/db mice. Cell 1996;84:491–5.

[9] Chen Y, Hozawa S, Sawamura S, et al. Deficiency of inducible noxide synthase exacerbates hepatic fibrosis in mice fed high-faBiochem Biophys Res Commun 2004;326:45–51.

10] Cortez-Pinto HM, Chatham JP, Chacko P, Arnold CR, RashidDiehl AM. Alterations in liver ATP homeostasis in human nonaholic steatohepatitis: a pilot study. JAMA 1999;282:1659–64.

11] Day C, James O. Hepatic steatosis: inocent bystander or guilty pHepatology 1998;27:1463–6.

12] Faggioni R, Jones-Carson J, Reed DA, et al. Leptin-deficient (omice are protected from T cell-mediated hepatotoxicity: roletumor necrosis factor� and IL-18. Proc Natl Acad Sci US2000;97:2367–72.

98 M. Ito et al. / Hepatology Research 34 (2006) 92–98

[13] Fan J-G, Zhong L, Xu Z-J, et al. Effects of low-calorie diet onsteatohepatitis in rats with obesity and hyperlipidemia. World J Gas-troenterol 2003;9:2045–9.

[14] Feldstein AE, Canbay A, Guicciardi ME, Higuchi H, Bronk SF,Gores GJ. Diet associated hepatic steatosis sensitizes to fas mediatedliver injury in mice. J Hepatol 2003;39:978–83.

[15] Freneaux E, Labbe G, Letteron P, et al. Inhibition of the mitochon-drial oxidation of fatty acids by tetracycline in mice and in man:possible role in microvesicular steatosis induced by this antibiotic.Hepatology 1988;8:1056–62.

[16] George D, Goldwurm S, Macdonald G, et al. Increased hepaticiron concentration in nonalcoholic steatohepatitis is associated withincreased fibrosis. Gastroenterology 1998;114:311–8.

[17] George J, Pera N, Phung N, Leclercq I, Hou JY, Farrell G. Lipidperoxidation, stellate cell activation and hepatic fibrogenesis in a ratmodel of chronic steatohepatitis. J Hepatol 2003;39:756–64.

[18] Green RM. NASH: hepatic metabolism and not simply the metabolicsyndrome. Hepatology 2003;38:14–7.

[19] Gyorgy B. Uncoupling protein-2 and non-alcoholic fatty liver dis-ease. Front Biosci 2005;10:2082–96.

[20] Harrison SA, Kadakia S, Lang KA, Schenker S. Nonalcoholic steato-hepatitis: what we know in the new millennium. Am J Gastroenterol2002;97:2714–24.

[21] Hashimoto E, Taniai M, Kaneda H, et al. Comparison of hepatocellu-lar carcinoma patients with alcoholic liver disease and nonalcoholicsteatohepatitis. Alcohol Clin Exp Res 2004;28:164S–8S.

[22] Horie H, Suzuki A, Kataoka E, et al. Hepatocyte-specific Pten defi-ciency results in steatohepatitis and hepatocellular carcinomas. J ClinInvest 2004;113:1774–83.

[23] Ip E, Farrell GC, Robertson G, Hall P, Kirsch R, Leclercq I. Centralrole of PPAR�-dependent hepatic lipid turnover in dietary steato-

[ cedion. J

[ odelrence

[ ract

[ the

[ andice.

[ hibi-nism33–

[ to-

[ ofMed

[ attylogy

[ pati-logy

[ ty: a

[35] Perez-Carreras M, Hoyo PD, Martin MA, et al. Defective hepaticmitochondrial respiratory chain in patients with nonalcoholic steato-hepatitis. Hepatology 2003;38:999–1007.

[36] Portincasa P, Grattagliano I, Palmieri VO, Palasciano G. Nonalco-holic steatohepatitis: recent advances from experimental models toclinical management. Clin Biochem 2005;38:203–17.

[37] Pruzanski W, Greenwald RA, Street IP, et al. Inhibition of enzy-matic activity of phospholipases A2 by minocycline and doxycycline.Biochem Pharmacol 1992;44:1165–70.

[38] Pruzanski W, Stefanski E, Vadas P, et al. Chemically modified non-antimicrobial tetracyclines inhibit activity of phospholipases A2. JRheumatol 1998;25:1807–12.

[39] Reid AE. Nonalcoholic steatohepatitis. Gastroenterology2001;121:710–23.

[40] Rinella ME, Green RM. The methionine–choline deficient dietarymodel of steatohepatitis does not exhibit insulin resistance. J Hepatol2004;40:47–51.

[41] Ruhl CE, Everhart JE. Determinants of the association of overweightwith elevated serum alanine aminotransferase activity in the UnitedStates. Gastroenterology 2003;124:71–9.

[42] Sahai A, Malladi P, Pan X, et al. Obese and diabeticdb/db micedevelop marked liver fibrosis in a model of nonalcoholic steato-hepatitis: role of short-form leptin receptors and osteopontin. Am JPhysiol Gastrointest Liver Physiol 2004;287:G1035–43.

[43] Shimada M, Hashimoto E, Taniai M, et al. Hepatocellular car-cinoma in patients with non-alcoholic steatohepatitis. J Hepatol2002;37:154–60.

[44] Shimomura I, Hammer RE, Ikemoto S, Brown MS, Goldstein JL.Leptin reverses insulin resistance and diabetes mellitus in mice withcongenital lipodystrophy. Nature 1999;401:73–6.

[45] Siegmund B, Sennello JA, Jones-Carson J, et al. Leptin receptoram-

[ en-atol

[ evel-ed aiver

[ ivermale.

[ nicn andcrinol

[ atol

[ obe-logy

[ C.onal-

[ e-nsUSA

[ weene male

hepatitis in mice. Hepatology 2003;38:123–32.24] Joseph JW, Koshkin V, Saleh MC, et al. Free fatty acid-indu

�-cell defects are dependent on uncoupling protein 2 expressBiol Chem 2004;279:51049–56.

25] Kirsch R, Clarkson V, Shephard EG, et al. Rodent nutritional mof non-alcoholic steatohepatitis: species, strain and sex diffestudies. J Gastroenterol Hepatol 2003;18:1272–82.

26] Koteish A, Diehl AM. Animal models of steatohepatitis. Best PRes Clin Gastroenterol 2002;16:679–90.

27] Lee GH, Proenca R, Montez JM, et al. Abnormal splicing ofleptin receptor in diabetic mice. Nature 1996;379:632–5.

28] Letteron P, Fromenty B, Benoit T, Degott C, Pessayre D. Acutechronic hepatic steatosis lead to in vivo lipid peroxidation in mJ Hepatol 1996;24:200–8.

29] Letteron P, Sutton A, Mansouri A, Fromenty B, Pessayre D. Intion of microsomal triglyceride transfer protein: another mechafor drug-induced steatosis in mice. Hepatology 2003;38:140.

30] Lieber CS, Leo MA, Mak KM, et al. Model of nonalcoholic steahepatitis. Am J Clin Nutr 2004;79:502–9.

31] Marchesini G, Brizi M, Morselli-Labate AM, et al. Associationnonalcoholic fatty liver disease with insulin resistance. Am J1999;107:450–5.

32] Marchesini G, Bugianesi E, Forlani G, et al. Nonalcoholic fliver, steatohepatitis, and the metabolic syndrome. Hepato2003;37:917–23.

33] Neuschwander-Tetri BA, Caldwell SH. Nonalcoholic steatohetis: summary of an AASLD Single Topic Conference. Hepato2003;37:1202–19.

34] Paracchini V, Pedotti P, Taioli E. Genetics of leptin and obesiHuGE review. Am J Epidemiol 2005;162:101–14.

expression on T lympkocytes modulates chronic intestinal inflmation in mice. Gut 2004;53:965–72.

46] Solga SF, Diehl AM. Non-alcoholic fatty liver disease: lumliver interactions and possible role for probiotics. J Hep2003;38:681–7.

47] Sundaram SS, Whitington PF, Green RM. Steatohepatitis dops rapidly in transgenic mice overexpressing Abcd11 and fmethionine–choline-deficient diet. Am J Physiol Gastrointest LPhysiol 2005;288:G1321–7.

48] Suzuki D, Hashimoto E, Kaneda H, Tokushige K, Shiratori K. Lfailure caused by non-alcoholic steatohepatitis in obese youngJ Gastroenterol Hepatol 2005;20:327–9.

49] Thupari JN, Kim EK, Moran TH, Ronnett GV, Kuhajda FP. ChroC75 treatment of diet-induced obese mice increases fat oxidatioreduces food intake to reduce adipose mass. Am J Physiol EndoMetab 2004;287:E97–104.

50] Tsutsui H. TNF signaling as pleiotropic gates in the liver. HepRes 2005;31:121–3.

51] Wanless IR, Lentz JS. Fatty liver hepatitis (steatohepatitis) andsity: an autopsy study with analysis of risk factors. Hepato1990;12:1106–10.

52] Weltman MD, Farrell GC, Hall P, Ingelman-Sundberg M, LiddleHepatic cytochrome P450 2E1 is increased in patients with ncoholic steatohepatitis. Hepatology 1998;27:128–33.

53] Yang SQ, Lin HZ, Lane MD, Clemens M, Diehl AM. Obsity increases sensitivity to endotoxin liver injury: implicatiofor the pathogenesis of steatohepatitis. Proc Natl Acad Sci1997;94:2557–62.

54] Yokoyama H, Hirose H, Ohgo H, Saito I. Inverse association betserum adiponectin level and transaminase activities in Japanesworkers. J Hepatol 2004;41:19–24.