Hepatic AdipoR2 Signaling Plays a Protective RoleAgainst Progression of Nonalcoholic Steatohepatitis

in MiceKengo Tomita,1 Yuichi Oike,2 Toshiaki Teratani,1 Takashi Taguchi,1 Masaaki Noguchi,3 Takahiro Suzuki,1

Akiko Mizutani,4 Hirokazu Yokoyama,5 Rie Irie,6 Hidetoshi Sumimoto,3 Atsushi Takayanagi,7 Kiichi Miyashita,8

Masaki Akao,9 Mitsuhisa Tabata,2 Gen Tamiya,10 Tamiko Ohkura,3 and Toshifumi Hibi1

It is unclear how hepatic adiponectin resistance and sensitivity mediated by the adiponectinreceptor, AdipoR2, contributes to the progression of nonalcoholic steatohepatitis (NASH).The aim of this study was to examine the roles of hepatic AdipoR2 in NASH, using an animalmodel. We fed C57BL/6 mice a methionine-deficient and choline-deficient (MCD) diet forup to 8 weeks and analyzed changes in liver pathology caused by either an AdipoR2 shorthairpin RNA–expressing adenovirus or an AdipoR2-overexpressing adenovirus. Inhibitionof hepatic AdipoR2 expression aggravated the pathological state of NASH at all stages: fattychanges, inflammation, and fibrosis. In contrast, enhancement of AdipoR2 expression in theliver improved NASH at every stage, from the early stage to the progression of fibrosis.Inhibition of AdipoR2 signaling in the liver diminished hepatic peroxisome proliferatoractivated receptor (PPAR)-� signaling, with decreased expression of acyl-CoA oxidase(ACO) and catalase, leading to an increase in lipid peroxidation. Hepatic AdipoR2 overex-pression had the opposite effect. Reactive oxygen species (ROS) accumulation in liver in-creases hepatic production of transforming growth factor (TGF)-�1 at all stages of NASH;adiponectin/AdipoR2 signaling ameliorated TGF-�–induced ROS accumulation in primarycultured hepatocytes, by enhancing PPAR-� activity and catalase expression. Conclusion:The adiponectin resistance and sensitivity mediated by AdipoR2 in hepatocytes regulatedsteatohepatitis progression by changing PPAR-� activity and ROS accumulation, a processin which TGF-� signaling is implicated. Thus, the liver AdipoR2 signaling pathway could bea promising target in treating NASH. (HEPATOLOGY 2008;48:458-473.)

Abbreviations: �-SMA, alpha-smooth muscle actin; ACO, acyl-CoA oxidase; ALT, alanine aminotransferase; AMPK, AMP-activated protein kinase; Col1�1, �1(I)collagen; Col1�2, �2(I) collagen; HE, hematoxylin-eosin; HNE, hydroxynonenal; MCD, methionine-deficient and choline-deficient; MOI, multiplicity of infection;mRNA, messenger RNA; NAC, N-acetylcysteine; NAFLD, nonalcoholic fatty liver disease; NASH, nonalcoholic steatohepatitis; O/E-R2-adenovirus, Adenovirus producingthe mouse adipoR2; PCR, polymerase chain reaction; PPAR, peroxisome proliferator activated receptor; ROS, reactive oxygen species; SEM, standard error of the mean;ShR2-adenovirus, mouse AdipoR2 shRNA-expressing adenovirus; shRNA, short hairpin RNA; SOD, superoxide dismutase; t-BHP, tert-butyl hydroperoxide; TG,triglyceride; TGF, transforming growth factor; TIMP-1, tissue inhibitor of metalloproteinase 1; WT, wild-type.

From the 1Division of Gastroenterology and Hepatology, Department of Internal Medicine, Keio University School of Medicine, Tokyo, Japan; the 2Department ofMolecular Genetics, Graduate School of Medical Sciences, Kumamoto University, Kumamoto, Japan; the 3Institute for Advanced Medical Research, Keio University Schoolof Medicine, Tokyo, Japan; the 4Department of Molecular Life Science, School of Medicine, Tokai University, Kanagawa, Japan; the 5Health Center, Keio UniversitySchool of Medicine, Tokyo, Japan; the 6Department of Diagnostic Pathology, Keio University School of Medicine, Tokyo, Japan; the 7Department of Molecular Biology, KeioUniversity School of Medicine, Tokyo, Japan; the 8Central Research Laboratory, Keio University School of Medicine, Tokyo, Japan; the 9Department of Cell Differentiation,Keio University School of Medicine, Tokyo, Japan; and the 10Department of Neurology, Tokushima University Graduate School of Medicine, Tokushima, Japan.

Received December 11, 2007; accepted April 3, 2008.Supported by a grant from Keio University School of Medicine, Nateglinide Memorial Toyoshima Research and Education Fund, the Naito Memorial Foundation, and

the Ministry of Education, Culture, Sports, Science, and Technology of Japan.Address reprint requests to: Toshifumi Hibi, MD, Division of Gastroenterology and Hepatology, Department of Internal Medicine, Keio University School of Medicine,

35 Shinanomachi, Shinjuku-ku, Tokyo 160-8582, Japan. E-mail: thibi@sc.itc.keio.ac.jp; fax: (81) 3-3357-6156.Copyright © 2008 by the American Association for the Study of Liver Diseases.Published online in Wiley InterScience (www.interscience.wiley.com).DOI 10.1002/hep.22365Potential conflict of interest: Nothing to report.

458

According to a recent report, nonalcoholic fattyliver disease (NAFLD) afflicts as much as 20% ofthe US adult population. Nonalcoholic steato-

hepatitis (NASH), part of the spectrum of NAFLD, is themost prevalent liver disease in the United States and isthought to affect approximately 3%-4% of the popula-tion.1

In this so called “age of satiation,” NAFLD and NASHare very often accompanied by lifestyle diseases, such asdiabetes and obesity, that have insulin resistance as a pri-mary background factor. Thus, these liver diseases can beconsidered hepatic representations of metabolic syn-drome. Given the growing number of metabolic syn-drome patients in recent years, the incidences of NAFLDand NASH are expected to increase further, particularlyin North America, Europe, Asia, and the western Pacificcountries.

NASH is a progressive disease. In a study that followedNASH patients for 10 years, the disease progressed tocirrhosis in 20% of the patients and led to death caused byliver diseases in 8%.2 In a population-based cohort study,approximately 3% of the patients in the community di-agnosed with NAFLD developed cirrhosis or a liver-re-lated complication.3 Because NASH is a progressivedisease, it is necessary to develop and establish a therapeu-tic method to treat it. The details of the pathologicalmechanism of NASH, however, have not yet been clari-fied.

Obesity is a major risk factor for NASH. The primaryfactor in the mechanism of the induction of insulin resis-tance by obesity is abnormal expression and excretion ofthe adipocytokines, physiologically active substances thatare released by adipose tissue. Adiponectin, an adipocyto-kine that improves insulin resistance and enhances theburning of fatty acids, can be used as an effective therapyagainst the insulin resistance that accompanies obesity ordiabetes.4 In the NAFLD mouse model, adiponectin hasanti-inflammatory and insulin resistance-reducing effects,and improves the pathological condition of NAFLD.5

Furthermore, there is an inverse association between se-rum adiponectin level and transaminase activities in Jap-anese male workers.6 The genes of the receptors foradiponectin (AdipoR1 and AdipoR2) were recentlycloned.7 AdipoR1 is abundantly expressed in skeletalmuscles, whereas AdipoR2 is expressed predominantly inthe liver, suggesting an association with the pathology ofliver diseases. Indeed, decreased expression of AdipoR2 inthe livers of patients with NASH has recently been report-ed,8 suggesting that adiponectin signaling mediated byAdipoR2 might play an important role in the pathogen-esis of NASH.

Therefore, in the current study, we examined the roleof adiponectin/AdipoR2 signaling during NASH, usingan adenovirus and a mouse model. We found that hepaticadiponectin sensitivity and resistance mediated by adi-ponectin/AdipoR2 signaling plays a significant role in thepathogenesis of NASH. Our findings suggest that theregulation of hepatic sensitivity to adiponectin by Adi-poR2 is a promising target in the development of newtherapeutic modalities for NASH.

Materials and Methods

Preparation of Adenovirus. The nucleotide se-quence for the short hairpin RNA (shRNA) againstmouse AdipoR2 were as follows: gatccgGCTTAGAGA-CACCTGTTTGTTttcaagagaAACAAACAGGTGTCTCTAAGCcttttttctcgagg (forward) and aattcctc-gagaaaaaagGCTTAGAGACACCTGTTTGTTtctcttgaaAACAAACAGGTGTCTCTAAGCCg (reverse), aspreviously described.7 The capital letters refer to the hy-bridizing shRNA-like portion of the hairpin. Vectors thatexpress mouse AdipoR2 shRNAs under the control of theU6 promoter were constructed by inserting the pairs ofannealed DNA oligonucleotides into the pSIREN-DNRplasmid (BD Clontech). The U6-driven shRNA cassettesand the CMV-driven DsRed expression cassette werethen inserted into the BD-AdenoX expression system(Clontech) to produce the mouse AdipoR2 shRNA-ex-pressing (ShR2)-adenovirus and the DsRed adenovirus(corresponding mock adenovirus).

Adenovirus producing the mouse adipoR2 (O/E-R2-adenovirus) and the corresponding mock adenovirus wereconstructed with the AdEasy adenoviral system (Strat-agene, La Jolla, CA). Recombinant adenoviruses wereproduced in HEK293 cells and purified on a cesium chlo-ride gradient before use.

Animal Studies. Male 9-week-old to 10-week-oldC57BL/6 mice were housed in cages and maintained un-der 12-hour light-dark cycles. Mice were fed a methi-onine-deficient and choline-deficient (MCD) diet (catNO 960439; ICN, Aurora, Ohio) or standard chow(CE-2; CLEA Japan Inc.) for up to 8 weeks. At variousintervals determined by the focus of the specific experi-mental question, animals were injected intravenouslythrough the tail with either 4 � 109 plaque-forming unitsof the adenoviruses or saline. In experiments focused onthe early stage of NASH, mice were fed the MCD diet for3 weeks after receiving injections. In experiments for latestage of NASH, we injected mice with adenoviruses orsaline after feeding them the MCD diet for 5 weeks; themice were further fed the MCD diet until 3 weeks afterinfection.

HEPATOLOGY, Vol. 48, No. 2, 2008 TOMITA ET AL. 459

For tert-butyl hydroperoxide (t-BHP) treatment, ani-mals were treated intraperitoneally with 1.5 mmol t-BHP(Sigma)/kg body weight or vehicle 24 hours before thefinal sample administration. For N-acetylcysteine (NAC)treatment, animals received intraperitoneally 50 mg/kgdoses of NAC or vehicle daily for 1 week or 3 weeksduring the feeding period. For vitamin C and vitamin E

treatment, animals were treated by gavage with a com-bined dose of vitamin C (100 mg/kg) and vitamin E(40 mg/kg) or vehicle daily for 4 weeks during thefeeding period. All animals received humane care incompliance with the National Research Council’s cri-teria outlined in the “Guide for the Care and Use ofLaboratory Animals,” prepared by the US National

Fig. 1. Adenovirus-mediated expression of AdipoR2 in primary cultures of hepatocytes, and in the livers of mice fed an MCD diet. (A) Levels ofAdipoR2 mRNA in primary cultures of mice hepatocytes 48 hours after infection with 100 MOI of mock-adenovirus or ShR2-adenovirus (upper panel),and mock-adenovirus or O/E-R2-adenovirus (lower panel). (B) Levels of hepatic AdipoR2 mRNA in mice fed control diet or MCD diet for 1 week (leftfigure) or 3 weeks (right figure), with treatment of saline, mock-adenovirus, or ShR2-adenovirus. (C) Levels of hepatic AdipoR2 mRNA in mice fedcontrol diet or MCD diet for 1 week (left figure) or 3 weeks (right figure), with treatment of saline, mock-adenovirus, or O/E-R2-adenovirus. Real-timePCR analysis was used to quantitate the mRNA levels of AdipoR2. The results (mean � SEM) of four individual experiments are shown. All real-timePCRs were performed in duplicate. *P � 0.05 compared with the mock group. (D) Western blot analysis of expression of hepatic AdipoR2 in micefed control diet or MCD diet for 1 week (left figure) or 3 weeks (right figure), with treatment of mock-adenovirus or ShR2-adenovirus. (E) Westernblot analysis of expression of hepatic AdipoR2 in mice fed control diet or MCD diet for 1 week (left figure) or 3 weeks (right figure), with treatmentof mock-adenovirus or O/E-R2-adenovirus. Cont, mice fed the control diet; MCD, mice fed the MCD diet; N/S, mice that received saline; Mock, micethat received the mock adenovirus; ShR2, mice that received the ShR2-adenovirus; O/E-R2, mice that received the O/E-R2-adenovirus.

460 TOMITA ET AL. HEPATOLOGY, August 2008

Academy of Sciences and published by the US NationalInstitutes of Health.

Biochemical and Histological Analysis. Serum con-centrations of alanine aminotransferase (ALT), glucose,and insulin were measured as previously described.9-11 Acommercially available enzyme-linked immunosorbentassay kit (R&D Systems, Minneapolis, MN) was used todetermine serum adiponectin. Hepatic triglyceride (TG)content and liver hydroxyproline concentrations weremeasured as previously described.9-11 Liver tissues werefixed in 4% paraformaldehyde, embedded in paraffin, and

stained with hematoxylin-eosin (HE) and a Masson-trichrome solution. For protein or RNA analysis, tissueswere frozen in liquid nitrogen and stored at �80°C untilneeded.

Western Blot Analysis. Western blotting was per-formed as described,10 using the following antibodies:peroxisome proliferator-activated receptor (PPAR)-�(Santa Cruz Biotechnology, Santa Cruz, CA), AdipoR1and AdipoR2 (Immuno-Biological Laboratories Co.,Ltd., Gumma, Japan), catalase and acyl-CoA oxidase(ACO) (Abcam Ltd., Cambridge, United Kingdom),

Fig. 2. AdipoR2 signaling in theliver, controlled by the adenoviruses,regulates the progression of theearly stage of steatohepatitis. Theeffect of hepatic AdipoR2 down-reg-ulation on (A) serum ALT activities,(B) hepatic TG levels, and (C) HE-stained sections of representativeliver samples, in each treated groupthat received 3 weeks of MCD orcontrol feeding. Values aremeans � SEM (n � 6-9/group).*P � 0.05 compared with the con-trol diet group. **P � 0.05 com-pared with the mock group. Cont,mice fed the control diet; MCD, micefed the MCD diet; N/S, mice thatreceived saline; Mock, mice that re-ceived the mock adenovirus; ShR2,mice that received the ShR2-adeno-virus. The effects of increased he-patic AdipoR2 expression on (D)serum ALT activities, (E) hepatic TGlevels, and (F) HE-stained sectionsof representative liver samples fromeach treated group that received 3weeks of MCD or control feeding.Values are means � SEM (n �6-9/group). *P � 0.05 comparedwith the control diet group. **P �0.05 compared with the mockgroup. Cont, mice fed the controldiet; MCD, mice fed the MCD diet;N/S, mice that received saline;Mock, mice that received the mockadenovirus; O/E-R2, mice that re-ceived the O/E-R2-adenovirus.

HEPATOLOGY, Vol. 48, No. 2, 2008 TOMITA ET AL. 461

Fig. 3. Hepatic AdipoR2 signal-ing controls ROS accumulation inthe early stage of steatohepatitis,through regulation of the activities ofPPAR-� and catalase. (A) The effectof hepatic AdipoR2 down-regulationon immunohistochemical stainingfor 4-HNE in the livers of eachtreated group that received 3 weeksof MCD feeding or control feeding.(B) The effects of increased hepaticAdipoR2 expression on immunohis-tochemical staining for 4-HNE in thelivers from each treated group thatreceived 3 weeks of MCD feeding orcontrol feeding. (C, D) Real-timePCR was used to analyze eachtreated group that received 3 weeksof MCD feeding, and to quantitatethe hepatic mRNA levels of PPAR-�,ACO, catalase, SOD1, SOD2, andAdipoR1. (C) Mock-adenovirus orShR2-adenovirus treated group. (D)Mock-adenovirus or O/E-R2-adeno-virus treated group. All real-timePCR reactions were performed in du-plicate. Values are means � SEM(n � 5-7/group). *P � 0.05 com-pared with the mock group. (E, F)Western blot analysis of expressionof hepatic PPAR-�, ACO, catalase,and AdipoR1 in mice fed MCD dietfor 3 weeks. (E) Mock-adenovirus orShR2-adenovirus treated group. (F)Mock-adenovirus or O/E-R2-adeno-virus treated group. (G, H) Westernblot analysis of phosphorylated andtotal AMPK-� of liver lysates in micefed MCD diet for 3 weeks. Represen-tative blots are shown (left figure).Ratio of phosphorylated versus totalAMPK-� was calculated by densito-metric quantification, and expressedas a ratio of the values in mock-adenovirus–treated mice (right fig-ure). Results are means � SEM(n � 5-7/group). (G) Mock-adeno-virus or ShR2-adenovirus treatedgroup. (H) Mock-adenovirus or O/E-R2-adenovirus treated group. Cont,mice fed the control diet; MCD, micefed the MCD diet; Mock, mice thatreceived the mock adenovirus;ShR2, mice that received the ShR2-adenovirus; O/E-R2, mice that re-ceived the O/E-R2 adenovirus.

462 TOMITA ET AL. HEPATOLOGY, August 2008

adenosine monophosphate–activated protein kinase(AMPK)-� and phospho-AMPK-� (Thr172) (Cell Sig-naling Technology Inc., Beverly, MA), and �-actin(Sigma Aldrich Inc., St Louis, MO).

Immunohistochemistry. Alpha-smooth muscle actin(�-SMA) and 4-hydroxynonenal (4-HNE) were detectedin paraffin-embedded liver sections as previously de-scribed.9-11 For transforming growth factor (TGF)-�1immunostaining, anti–TGF-�1 antibody (Santa CruzBiotechnology) was used.

Real-Time Quantitative and Reverse TranscriptionPolymerase Chain Reaction Analysis. Total RNA wasextracted from total liver homogenates or hepatocytes us-ing Isogen (Nippon Gene, Tokyo, Japan), as previouslydescribed.9-11 Reverse transcription and real-time poly-merase chain reaction (PCR) amplification were per-formed as previously described, using Taqman assayreagents (Applied Biosystems, Foster City, CA).9-11

Hepatocyte Isolation and Cell Culture. Heterozy-gous AdipoR2 gene knockout mice (C57BL/6 back-ground) were purchased from Jackson Laboratories (BarHarbor, ME). Male wild-type (WT) and homozygous

(AdipoR2 KO) littermates were obtained by breedingheterozygous mice.

Primary cultured hepatocytes were prepared from liv-ers of WT mice or AdipoR2 KO mice, and plated onsix-well collagen-coated culture plates, as previously de-scribed.10,11 After a 16-hour culture period, the culturemedium was replaced with fresh medium with or without10 �g/mL recombinant full-length mouse adiponectin(BioVendor Laboratory Medicine Inc., Czech Republic).Six hours later, cells were exposed to 1 nM recombinantTGF-� (R & D Systems, Minneapolis, MN), or vehicle,and further cultured for 12 hours.

Measuring Hepatic Protein Carbonyls. The con-centration of hepatic proteins containing carbonyl groups(those that react with 2,4-dinitrophenylhydrazine to formthe corresponding hydrazone) was determined spectro-photometrically with a protein carbonyl assay kit (Cay-man Chemical, Ann Arbor, MI).

Statistical Analysis. All data are expressed asmeans � standard error of the mean (SEM). Statisticalanalyses were performed using the unpaired Student testor one-way analysis of variance. When the analysis of

Fig. 3 (Cont’d.)

HEPATOLOGY, Vol. 48, No. 2, 2008 TOMITA ET AL. 463

464 TOMITA ET AL. HEPATOLOGY, August 2008

COLOR

variance analyses were applied, differences in mean valuesamong groups were examined by Fisher’s multiple com-parison test.

Results

Adenovirus-Mediated Expression of AdipoR2 inPrimary Cultures of Hepatocytes, and in the Livers ofMice Fed MCD Diet. In primary cultures of hepato-cytes, the ShR2-adenovirus we created significantly inhib-ited AdipoR2 expression, and the O/E-R2-adenovirussignificantly increased it (Fig. 1A).

Mice fed an MCD diet are widely used as a model forNASH because they demonstrate similar hepatic his-tology to that observed in human cases of NASH.9,12 Inthe current study, we administered ShR2-adenovirus,O/E-R2-adenovirus, or each corresponding mock ade-novirus to male C57BL/6 mice, and then fed them anMCD diet. One week after each adenovirus adminis-tration, we found that primarily the parenchymal cellsof the liver had been infected; adenovirus infection wasconfirmed in more than 80% of parenchymal cells but

in less than 25% of the nonparenchymal cells (data notshown).

Until 3 weeks after treatment, the ShR2-adenovirussignificantly suppressed hepatic AdipoR2 messengerRNA (mRNA) and protein expression (Fig. 1B,D),whereas the O/E-R2-adenovirus significantly increasedhepatic AdipoR2 mRNA and protein expression (Fig.1C,E).

AdipoR2 Signaling in the Liver Regulates the Pro-gression of the Early Stage of Nutritional Steatohepa-titis. After 3 weeks of MCD feeding, the group receivingthe ShR2-adenovirus exhibited significant increases in ac-cumulation of lipid droplets in hepatocytes, infiltration ofinflammatory cells in the liver, and hepatic TG concen-trations, particularly in the perivenular areas, relative tothe group receiving the corresponding mock adenovirus(Fig. 2B,C). MCD feeding elevated serum ALT levels to agreater extent in the ShR2-adenovirus group than in themock adenovirus group (Fig. 2A).

Conversely, the administration of O/E-R2-adenovirusinhibited the increase in the accumulation of lipid drop-lets in hepatocytes and the infiltration of inflammatorycells, which occurred after 3 weeks of MCD feeding (Fig.2F). Consistent with the histological findings, hepatic TGconcentrations were significantly lower in the O/E-R2-adenovirus group than in the mock adenovirus group(Fig. 2E). The administration of O/E-R2-adenovirus alsosignificantly decreased the serum ALT levels (Fig. 2D).

Neither ShR2-adenovirus nor O/E-R2-adenovirustreatment had any significant effect on body weight, foodintake, and serum levels of glucose, insulin, or adiponec-tin, in comparison with the corresponding mock adeno-virus treatment (data not shown). Moreover, the effects ofadenovirus treatment were tissue specific, because therewere no differences in AdipoR2 mRNA levels in muscle,fat, kidney, or brain between the saline-treated mice andthe mice treated with ShR2-adenovirus or O/E-R2-ade-novirus (data not shown).

AdipoR2 Signaling in the Liver Regulates HepaticReactive Oxygen Species Accumulation in Early Stageof NASH, Through Activation of PPAR-� and Cata-lase. We performed immunostaining of 4-HNE–ad-ducted protein as a product of the lipid peroxidationreaction. In the early stage of nutritional steatohepatitisinduced by 3 weeks of MCD feeding, lipid peroxidationin the liver was significantly enhanced, relative to the liverof control-fed mice (Fig. 3A). The ShR2-adenovirusgroup demonstrated more intense staining of hepatocytesafter MCD feeding, centering around the centrilobularregion of the liver, than did the mock adenovirus group(Fig. 3A). Conversely, 4-HNE–adducted protein in hepa-tocytes around the centrilobular region was significantly

4™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™Fig. 4. AdipoR2 signaling in the liver regulated the progression of late

fibrotic stage of steatohepatitis. (A) The effect of hepatic AdipoR2down-regulation on liver hydroxyproline concentrations in each treatedgroup that received 8 weeks of MCD or control feeding. The data areexpressed as fold over the control diet group. Values are means � SEM(n � 5-7/group). *P � 0.05 compared with the control diet group.**P � 0.05 compared with the mock group. (B) Real-time PCR analysiswas used to quantitate hepatic mRNA levels of Col1�1, Col1�2, andTIMP-1 in each treated group, infected with mock-adenovirus or ShR2-adenovirus. Here, all animals received 8 weeks of MCD feeding. Allreal-time PCR reactions were performed in duplicate. Values aremeans � SEM (n � 5-7/group). *P � 0.05 compared with the mockgroup. (C) Immunohistochemical staining for �-SMA in the livers of eachtreated group that received 8 weeks of MCD feeding. (D) HE-stainedsections and Masson trichrome (Masson)-stained sections of represen-tative liver samples from each treated group that received 8 weeks ofMCD or control feeding. (E) The effect of increased hepatic AdipoR2expression on liver hydroxyproline concentrations in each treated groupthat received 8 weeks of MCD or control feeding. The data are expressedas fold over the control diet group. Values are means � SEM (n �5-7/group). *P � 0.05 compared with the control diet group. **P �0.05 compared with the mock group. (F) Real-time PCR analysis wasused to quantitate hepatic mRNA levels of Col1�1, Col1�2, and TIMP-1in each treated group, infected with mock-adenovirus or O/E-R2-adeno-virus. Here, all animals received 8 weeks of MCD feeding. All real-timePCR reactions were performed in duplicate. Values are means � SEM(n � 5-7/group). *P � 0.05 compared with the mock group. (G)Immunohistochemical staining for �-SMA in the livers of each treatedgroup that received 8 weeks of MCD feeding. (H) HE-stained sections andMasson trichrome (Masson)-stained sections of representative liver sam-ples from each treated group that received 8 weeks of MCD or controlfeeding. Cont, mice fed the control diet; MCD, mice fed the MCD diet;N/S, mice that received saline; Mock, mice that received the mockadenovirus; ShR2, mice that received the ShR2-adenovirus; O/E-R2,mice that received the O/E-R2-adenovirus.

HEPATOLOGY, Vol. 48, No. 2, 2008 TOMITA ET AL. 465

466 TOMITA ET AL. HEPATOLOGY, August 2008

COLOR

less intense in the livers of the O/E-R2-adenovirus groupthan in those of the mock adenovirus group (Fig. 3B).

In the livers of the ShR2-adenovirus group, the expres-sion of hepatic PPAR-� and the PPAR-� target geneACO, was significantly reduced relative to the livers of themock adenovirus group (Fig. 3C,E). The expression ofcatalase, an antioxidant enzyme, was also significantly re-duced (Fig. 3C,E). Conversely, in the livers of the O/E-R2-adenovirus group, the expression of hepatic PPAR-�,ACO, and catalase was significantly increased relative tothe livers of the mock adenovirus group (Fig. 3D,F). Thealteration of AdipoR2 expression did not significantlychange hepatic AMPK activity (Fig. 3G,H).

AdipoR2 Signaling in the Liver Plays a Key Role inthe Progression of the Late Fibrotic Stage of Nutri-tional Steatohepatitis. To examine how AdipoR2 defi-ciency in the liver affects the progression of fibrosis insteatohepatitis, we injected mice with mock or ShR2-adenovirus after feeding them an MCD diet for 5 weeks.Three weeks after infection, more extensive fibrosis wascaused by steatohepatitis in the livers of the ShR2-adeno-virus group than those of the mock adenovirus group (Fig.4D). Fibrosis was conspicuously localized around the cen-tral vein and throughout the lobule in a pericellular dis-

tribution. Extraction and quantitation of liverhydroxyproline confirmed the microscopic results. Thehydroxyproline content of the livers of the group infectedwith the ShR2-adenovirus was significantly higher thanthe content in those of the group infected with the mockadenovirus (Fig. 4A). Real-time PCR analysis of wholeliver homogenates from ShR2-adenovirus-infected miceafter MCD feeding showed a significant increase in �1(I)collagen (Col1�1), �2 (I) collagen (Col1�2), and tissueinhibitor of metalloproteinase 1 (TIMP-1) mRNA levelsrelative to mock adenovirus-infected mice (Fig. 4B). Thelivers of MCD-fed ShR2-adenovirus-infected mice alsoexhibited increased staining for �-SMA, a marker of stel-late cell activation in areas of damage, in comparison withmock adenovirus-infected mice (Fig. 4C).

Next, to examine the influence of liver AdipoR2 over-expression on the progression of fibrosis in steatohepatitis,we infected mice with mock or O/E-R2-adenovirus afterfeeding them an MCD diet for 5 weeks. As shown in Fig.4H, fibrosis progressed to a lesser extent in the livers of theO/E-R2-adenovirus group than in those of the mock ad-enovirus group, 3 weeks after infection. The hydroxypro-line content of the livers of O/E-R2-adenovirus-infectedmice was significantly lower than in the livers of mockadenovirus-infected mice (Fig. 4E). Real-time PCR anal-ysis of whole liver homogenates of O/E-R2-adenovirus-infected mice following MCD feeding revealed asignificant decrease in Col1�1, Col1�2, and TIMP-1mRNA levels relative to mock adenovirus-infected mice(Fig. 4F). The livers of MCD-fed, O/E-R2-adenovirus-infected mice also exhibited decreased staining for�-smooth muscle actin in comparison with mock adeno-virus-infected mice (Fig. 4G).

AdipoR2 Signaling in the Liver Regulates HepaticReactive Oxygen Species Accumulation in the LateFibrotic Stage of NASH, Through Activation ofPPAR-� and Catalase. In the late stage of nutritionalsteatohepatitis induced by 8-week MCD feeding, lipidperoxidation in the liver was significantly enhanced, rela-tive to the liver of control-fed mice (Fig. 5A). Immuno-staining of 4-HNE-adducted protein demonstratedincreased lipid peroxidation in the livers of ShR2-adeno-virus-infected mice relative to those of mock adenovirus-infected mice (Fig. 5A). In the livers of the ShR2-adenovirus group, the expression of hepatic PPAR-�,ACO, and catalase were significantly reduced relative tothe livers of the mock adenovirus group (Fig. 5B,C).

Lipid peroxidation in the livers of O/E-R2-adenovirus-infected mice was significantly reduced, relative to thoseof mock adenovirus-infected mice (Fig. 5D). In the liversof the O/E-R2-adenovirus group, the expression of he-patic PPAR-�, ACO, and catalase was significantly in-

4™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™Fig. 5. AdipoR2 signaling in the liver regulates hepatic ROS accumu-

lation in the late fibrotic stage of NASH, through activation of PPAR-� andcatalase. (A) The effect of hepatic AdipoR2 down-regulation on 4-��Estaining in the livers of each treated group; animals received 8 weeks ofMCD feeding or control feeding. (B) Real-time PCR was used to quan-titate the hepatic mRNA levels of PPAR-�, ACO, and catalase in eachtreated group, infected with mock-adenovirus or ShR2-adenovirus. Here,all animals received 8 weeks of MCD feeding. All real-time PCR reactionswere performed in duplicate. Values are means � SEM (n � 5-7/group).*P � 0.05 compared with the mock group. (C) Western blot analysis ofexpression of hepatic PPAR-�, ACO, and catalase in mice fed MCD dietfor 8 weeks, with treatment of mock-adenovirus or ShR2-adenovirus. (D)The effect of increased hepatic AdipoR2 expression on 4-HNE staining inthe livers of each treated group; animals received 8 weeks of MCDfeeding or control feeding. (E) Real-time PCR was used to quantitate thehepatic mRNA levels of PPAR-�, ACO, and catalase in each treatedgroup, infected with mock-adenovirus or O/E-R2-adenovirus. All animalsreceived 8 weeks of MCD feeding. All real-time PCR reactions wereperformed in duplicate. Values are means � SEM (n � 5-7/group).*P � 0.05 compared with the mock group. (F) Western blot analysis ofexpression of hepatic PPAR-�, ACO, and catalase in mice fed MCD dietfor 8 weeks, with treatment of mock-adenovirus or O/E-R2-adenovirus.(G, H) Western blot analysis of phosphorylated and total AMPK-� of liverlysates in mice fed MCD diet for 8 weeks. Representative blots are shown(left figure), and ratio of phosphorylated versus total AMPK-� wascalculated by densitometric quantification, and expressed as a ratio ofthe values in mock-adenovirus–treated mice (right figure). Results aremeans � SEM (n � 5-7/group). (G) Mock-adenovirus or ShR2-adeno-virus treated group. (H) Mock-adenovirus or O/E-R2-adenovirus treatedgroup. Cont, mice fed the control diet; MCD, mice fed the MCD diet;Mock, mice that received the mock adenovirus; ShR2, mice that re-ceived the ShR2-adenovirus; O/E-R2, mice that received the O/E-R2-adenovirus.

HEPATOLOGY, Vol. 48, No. 2, 2008 TOMITA ET AL. 467

creased relative to the livers of the mock adenovirus group(Fig. 5E,F). The alteration of AdipoR2 expression did notsignificantly change hepatic AMPK activity (Fig. 5G,H).

Both Reactive Oxygen Species Accumulation andAdipoR2 Signaling in the Liver Regulate HepaticTGF-�1 Levels at All Stages of NASH Pathology.Hepatic TGF-�1 expression level in MCD-fed mice sig-nificantly increased 1 week after treatment, and contin-ued to rise until 8 weeks after treatment, compared withthe level in control-fed mice (Fig. 6A). Co-administrationof NAC, an antioxidant agent known to eliminate reac-tive oxygen species (ROS),13 significantly reduced the el-evation of hepatic TGF-�1 expression in mice fed theMCD diet for 1 week, 3 weeks, or 8 weeks, in comparisonwith each corresponding group administered vehicle (Fig.6A). This observation was confirmed by immunostainingfor TGF-�1 (Fig. 6B,C). Co-administration of combineddoses of vitamin C and vitamin E, other antioxidantagents known to eliminate hepatic ROS accumulation,14

also significantly reduced the elevation of hepaticTGF-�1 expression in mice fed an MCD diet for 4 weeks(Fig. 6D). Acute administration of t-BHP, which is re-ported to induce hepatic ROS accumulation,15-17 also in-creased hepatic TGF-�1 expression in mice (Fig. 6E).Real-time PCR analysis of whole liver homogenates fromShR2-adenovirus-infected mice after 3 weeks’ MCDfeeding showed a significant increase in TGF-�1 mRNAlevels relative to mock adenovirus-infected mice (Fig.7A), whereas hepatic TGF-�1 mRNA levels in O/E-R2-adenovirus-infected mice after 3 weeks’ MCD feedingwas significantly lower than in the livers of mock adeno-virus-infected mice (Fig. 7B). This observation was con-firmed by immunostaining for TGF-�1 (Fig. 7C,D).Similarly, in experiments focused on the late fibrotic stageof NASH induced by 8-week MCD feeding, TGF-�1mRNA and protein expression in the livers of ShR2-ade-novirus-infected mice was significantly higher than in thelivers of mock adenovirus-infected mice (Fig. 7E,G),whereas hepatic TGF-�1 expression in the O/E-R2-ade-novirus was significantly lower than in the livers of mockadenovirus-infected mice (Fig. 7F,H).

AdipoR2 Signaling Suppresses TGF-�–InducedROS Accumulation in Hepatocytes Through Activa-tion of PPAR-� and Catalase. TGF-� treatment sig-nificantly increased ROS accumulation in both WThepatocytes and AdipoR2 KO hepatocytes (Fig. 8A). Ad-dition of adiponectin significantly suppressed TGF-�–induced ROS accumulation in WT hepatocytes, but hadno effect in AdipoR2 KO hepatocytes (Fig. 8A). At thesame time, TGF-� treatment significantly decreased themRNA level of PPAR-�, ACO, and catalase in both WT

hepatocytes and AdipoR2 KO hepatocytes (Fig. 8B). Co-treatment of adiponectin significantly reversed this effectonly in WT hepatocytes. In AdipoR2 KO hepatocytes,addition of adiponectin had no effect (Fig. 8B).

DiscussionWe have elucidated the importance of hepatic Adi-

poR2 signaling in the pathology of NASH, using a mousemodel of NASH induced by an MCD diet. At all stages ofNASH pathology in the liver (fat accumulation, inflam-mation, and fibrosis), inhibition of AdipoR2 expressionin the liver significantly exacerbated the pathological stateof NASH. Conversely, increased AdipoR2 expression inthe liver improved NASH at every stage.

Adiponectin activates adenosine monophosphate-acti-vated protein kinase and PPAR-� through its receptorsAdipoR1 and AdipoR2; regulation of these genes isthought to be a primary mechanism by which adiponectinacts against insulin resistance and diabetes.4 Recently, bycontrolling AdipoR1, AdipoR2, or both genes withknockout mice or adenovirus, Yamauchi et al.18 provedthat the two receptors differ in the details of their down-stream signaling pathways. AdipoR1 is more tightlylinked to the activation of the AMPK pathway and cou-ples the inhibition of hepatic glucose production to in-creased activation of fatty acid oxidation, whereasAdipoR2 is mainly involved in the activation of thePPAR-� pathway. In agreement with their findings, alter-ation of AdipoR2 expression did not change hepaticAMPK activity in our NASH animal model. They alsoreported that the expression levels of AdipoR1 and Adi-poR2 change according to pathology, thereby regulatingthe adiponectin sensitivity and resistance of each organ.In fact, the expression levels of AdipoR1 and AdipoR2 arereportedly controlled by insulin/PI3 kinase/Foxo1,19 sug-gesting an important role for this pathway in the controlof adiponectin sensitivity. These reports suggest that thepossibility that AdipoR2, which is expressed predomi-nantly in the liver, regulates the adiponectin sensitivityand resistance of the liver, mediating PPAR-� activity.The results of our current study support this possibility.

In this study, we first demonstrated that the inhibitionof AdipoR2 signaling in the liver aggravated NASH atevery stage, from the early stage to the progression offibrosis, by inhibiting PPAR-� activity in hepatocytes.We also proved that hepatic AdipoR2 overexpression hadthe opposite effect.

PPAR-� controls the expression of a number of genesinvolved in peroxisomal and mitochondrial �-oxidation.The inhibition of AdipoR2 signaling in the liver causedthe inhibition of PPAR-� signaling in hepatocytes, in-creasing the TG accumulation in hepatocytes in livers

468 TOMITA ET AL. HEPATOLOGY, August 2008

with steatohepatitis and also inhibiting the expression ofACO, a hepatic enzyme that contributes to fatty acidoxidation. Meanwhile, the inhibition of AdipoR2 signal-ing induced lipid peroxidation and oxidative stress in liv-ers with steatohepatitis, and it also reduced the expressionof the antioxidant enzyme catalase. Toyama et al.20 re-cently reported that PPAR-� ligand activated hepatic

catalase,20 which is also consistent with our results. Theyshowed that PPAR-� ligand also activated hepatic super-oxide dismutase (SOD), although the enhancement ofexpression was more specific and striking for catalase thanSOD. These results suggest that alteration of PPAR-�activity, regulated by AdipoR2 signaling, would not besevere enough to significantly change hepatic SOD ex-

Fig. 6. ROS accumulation in the liver enhances hepatic TGF-�1 levels. (A) Mice were fed an MCD diet or control diet for 1 week, 3 weeks, or 8weeks, and co-treated with NAC or vehicle. In the 1-week or 3-week treated groups, animals were given NAC or vehicle daily during the feeding period.In the 8-week treated group, animals were given NAC or vehicle daily for the last 3 weeks of the feeding period. Real-time PCR was used to analyzeeach group, and to quantitate the hepatic mRNA levels of TGF-�1. Values are means � SEM (n � 5-7/group). *P � 0.05 compared with thecorresponding control diet group in the same feeding period. **P � 0.05 compared with the corresponding MCD diet group in the same feedingperiod. Control: animals fed control diet with vehicle treatment. MCD: animals fed MCD diet with vehicle treatment. MCD�NAC: animals fed MCDdiet with NAC treatment. (B, C) Immunohistochemical staining for TGF-�1. Representative liver samples from mice fed control diet or MCD diet for3 weeks (B) and for 8 weeks (C), with or without co-administration of NAC. Cont: animals fed control diet with vehicle treatment; MCD: animals fedMCD diet with vehicle treatment; MCD�NAC: animals fed MCD diet with NAC treatment; (D) Mice were fed an MCD diet or control diet for 4 weeks,along with co-administration of both vitamin C and vitamin E, or vehicle. Real-time PCR was used to quantitate the hepatic mRNA levels of TGF-�1.Values are means � SEM (n � 5-7/group). *P � 0.05 compared with the control diet group. **P � 0.05 compared with the MCD diet group withvehicle treatment. Cont, animals fed control diet with vehicle treatment; MCD, animals fed MCD diet with vehicle treatment; MCD�Vit, animals fedMCD diet with combined doses of vitamin C and vitamin E. (E) The amount of hepatic TGF-�1 mRNA in mice treated with t-BHP or vehicle. Valuesare means � SEM (n � 5-7/group). *P � 0.05 compared with the control group. Cont, mice given the vehicle; t-BHP, mice given the t-BHP.

HEPATOLOGY, Vol. 48, No. 2, 2008 TOMITA ET AL. 469

pression in our NASH mice model. In our NASH model,hepatic expression of antioxidant enzymes, such as cata-lase, significantly decreased (Table 1), similar to the situ-ation in NASH patients.21 Future studies will be requiredto determine the role of regulation of hepatic antioxidantmolecules other than those in the PPAR-� and AdipoR2signaling pathways in the pathogenesis of NASH.

Accumulated ROS and TGF-� signaling in liver arethought to play a key role in the pathogenesis of nonalco-holic steatohepatitis.22 In the current study, we found thatthe ROS production and TGF-� expression in liver weresignificantly enhanced at every stage of nutritional steato-hepatitis. Accumulated ROS in liver activates hepaticnonparenchymal cells, including Kupffer cells and he-

Fig. 7. AdipoR2 signaling in theliver also regulates hepatic TGF-�1level in NASH pathology. (A, B) Theamount of hepatic TGF-�1 mRNA inmice fed control diet or MCD diet for3 weeks, with treatment of mock,ShR2, or O/E-R2 adenovirus. (A)Mock-adenovirus or ShR2-adenovi-rus treated group. (B) Mock-adeno-virus or O/E-R2-adenovirus treatedgroup. Real-time PCR was used toquantitate the hepatic mRNA levelsof TGF-�1. Values are means �SEM (n � 5-7/group). *P � 0.05compared with the control dietgroup. **P � 0.05 compared withthe mock group. (C, D) Immunohis-tochemical staining for TGF-�1.Representative liver samples frommice fed control diet or MCD diet for3 weeks, with treatment of mock,ShR2, or O/E-R2 adenovirus. (C)Mock-adenovirus or ShR2-adenovi-rus treated group. (D) Mock-adeno-virus or O/E-R2-adenovirus treatedgroup. (E, F) The amount of hepaticTGF-�1 mRNA in mice fed controldiet or MCD diet for 8 weeks, withtreatment of mock, ShR2, or O/E-R2adenovirus. (E) Mock-adenovirus orShR2-adenovirus treated group. (F)Mock-adenovirus or O/E-R2-adeno-virus treated group. Real-time PCRwas used to quantitate the hepaticmRNA levels of TGF-�1. Values aremeans � SEM (n � 5-7/group).*P � 0.05 compared with the con-trol diet group. **P � 0.05 com-pared with the mock group. (G, H)Immunohistochemical staining forTGF-�1. Representative liver sam-ples from mice fed control diet orMCD diet for 8 weeks, with treat-ment of mock, ShR2, or O/E-R2 ad-enovirus. (G) Mock-adenovirus orShR2-adenovirus treated group. (H)Mock-adenovirus or O/E-R2-adeno-virus treated group. Cont, mice fedthe control diet; MCD, mice fed theMCD diet; Mock, mice that receivedthe mock adenovirus; ShR2; micethat received the ShR2-adenovirus;O/E-R2, mice that received theO/E-R2-adenovirus.

470 TOMITA ET AL. HEPATOLOGY, August 2008

patic stellate cells.23 Activated Kupffer cells are reported torelease TGF-�, a known pro-fibrotic factor that has beenimplicated in the activation of other surrounding livercells, including hepatic stellate cells, through a paracrinemechanism.24 Hepatic stellate cells are known to be keyplayers in liver fibrosis; once activated by multiple stimulisuch as ROS and TGF-�, they produce collagen and ad-ditional TGF-�, leading to the further activation of he-patic stellate cells via a paracrine or autocrinemechanism.25 In the current study, hepatic expression ofTGF-�1 was enhanced at both the early stage and latefibrotic stage of NASH. We also found that elimination of

ROS accumulation in liver decreased hepatic TGF-�1levels, whereas enhancement of ROS in liver increasedTGF-�1 mRNA levels. Our in vitro study showed thatthe treatment of TGF-� exaggerated ROS accumulationin hepatocytes, along with decreased expression ofPPAR-�, ACO, and catalase. These results suggest that avicious cycle of ROS and TGF-� could play a key role inliver injury in the early stage of NASH as well as in the latefibrotic stage. Recent reports also showed that silencing ofTGF-� signaling in hepatocytes, rather than in nonparen-chymal cells, could be effective in treating acute liver in-jury.26 Our in vitro study also proved that AdipoR2

Fig. 8. AdipoR2 signaling suppresses TGF-�–induced ROS accumulation in primary cultured hepatocytes. (A) Protein carbonyl content in primarycultured wild-type (white bar) or AdipoR2 KO (black bar) hepatocytes, after incubation with TGF-� with addition of adiponectin, or not. Quantificationswere normalized for the content from the wild-type hepatocytes cultures without adiponectin or TGF-�. The results (mean � SEM) of four individualexperiments are shown. *P � 0.05 compared with the control group. **P � 0.05 compared with the content from the AdipoR2 KO hepatocytesafter incubation of both TGF-� and adiponectin. (B) The amount of PPAR-�, ACO, and catalase mRNA in primary cultured wild-type (white bar) orAdipoR2 KO (black bar) hepatocytes, after incubation with TGF-� with addition of adiponectin, or not. Real time PCR was used to quantitate thehepatic mRNA level, and quantifications were normalized for RNA from the corresponding control cultures without addition of TGF-� or adiponectin.All real-time PCR reactions were performed in duplicate. The results (mean � SEM) of four individual experiments are shown. *P � 0.05 comparedwith the cultures without addition of TGF-�. **P � 0.05 compared with the corresponding hepatocyte cultures with TGF-� without addition ofadiponectin.

HEPATOLOGY, Vol. 48, No. 2, 2008 TOMITA ET AL. 471

signaling suppressed TGF-�–induced ROS accumula-tion in hepatocytes, through activation of PPAR-� andcatalase. Because AdipoR2 signaling was controlledmainly in hepatocytes in our in vivo study, we hypothesizethat ROS accumulation in hepatocytes, which is con-trolled by multiple mediators including TGF-�, could beregulated by AdipoR2 signaling. This mechanism couldplay a key role in the pathogenesis of NASH. This exag-gerated accumulation of ROS led to parenchymal celldamage and nonparenchymal cell activation with en-hanced release of TGF-� in the liver, thus establishing themalignant cycle of NASH progression.

In hepatocytes in our in vitro study, adiponectin/Adi-poR2 signaling did not have any direct effect on themRNA level of PPAR-�, ACO, and catalase, but reversedthe decrease in expression levels that resulted from TGF-�treatment. These results suggest that adiponectin/Adi-poR2 signaling in hepatocytes is involved in the activationof the PPAR-� pathway, mainly by antagonizing the ac-tions of various molecules, such as TGF-�, on PPAR-�activities.

In summary, we demonstrated that an increase or de-crease in AdipoR2 signaling in hepatocytes regulates thelevel of progression of nutritional steatohepatitis. Adi-poR2 signaling plays a pivotal role in the pathology-reg-ulating mechanism, which is initiated by ROS thataccumulate due to dysregulation of PPAR-� and catalaseactivity in hepatocytes.

The two-hit theory proposed by Day and James27 iswidely advocated as a pathogenic mechanism for NASH.This theory claims that, in addition to steatosis (the firsthit), the development of steatohepatitis requires the pres-ence of some other factor or factors (the second hit). Thecurrent study revealed that decreased AdipoR2 signalingin the liver can serve as the second hit in NASH progres-sion. Because AdipoR2 genes could regulate susceptibilityto NASH, relevant genetic and proteomic analysis willprove insightful in the future.

There are no established therapeutic modalities forNASH, and it is urgent to develop such modalities. Theenhancement of the AdipoR2 signaling pathway in theliver will serve as a promising target in the treatment ofNASH.

Acknowledgment: The authors thank T. Yokoyama,S. Tanaka, J. Amano, and T. Chiyo for technical assis-tance.

References1. Neuschwander-Tetri BA. Fatty liver and the metabolic syndrome. Curr

Opin Gastroenterol 2007;23:193-198.2. Falck-Ytter Y, Younossi ZM, Marchesini G, McCullough AJ. Clinical

features and natural history of nonalcoholic steatosis syndromes. Semin LivDis 2001;21:17-26.

3. Adams LA, Lymp JF, Sauver JS, Sanderson SO, Lindor KD, Feldstein A, etal. The natural history of nonalcoholic fatty liver disease: a population-based cohort study. Gastroenterology 2005;129:113-121.

4. Kadowaki T, Yamauchi T, Kubota N, Hara K, Ueki K, Tobe K. Adiponec-tin and adiponectin receptors in insulin resistance, diabetes, and the met-abolic syndrome. J Clin Invest 2006;116:1784-1792.

5. Xu A, Wang Y, Keshaw H, Xu LY, Lam KSL, Cooper GJS. The fat-derivedhormone adiponectin alleviates alcoholic and nonalcoholic fatty liver dis-eases in mice. J Clin Invest 2003;112:91-100.

6. Yokoyama H, Hirose H, Ohgo H, Saito I. Inverse association betweenserum adiponectin level and transaminase activities in Japanese male work-ers. J Hepatol 2004;41:19-24.

7. Yamauchi T, Kamon J, Ito Y, Tsuchida A, Yokomizo T, Kita S, et al.Cloning of adiponectin receptors that mediate antidiabetic metabolic ef-fects. Nature 2003;423:762-769.

8. Kaser S, Moschen A, Cayon A, Kaser A, Crespo J, Pons-Romero F, et al.Adiponectin and its receptors in non-alcoholic steatohepatitis. Gut 2005;54:117-121.

9. Tomita K, Tamiya G, Ando S, Ohsumi K, Chiyo T, Mizutani A, et al.Tumor necrosis factor � signaling through activation of Kupffer cells playsan essential role in liver fibrosis of non-alcoholic steatohepatitis in mice.Gut 2006;55:415-424.

10. Tomita K, Azuma T, Kitamura N, Nishida J, Tamiya G, Oka A, et al.Pioglitazone prevents alcohol-induced fatty liver in rats through up-regu-lation of c-Met. Gastroenterology 2004;126:873-885.

11. Tomita K, Azuma T, Kitamura N, Tamiya G, Ando S, Nagata H, et al.Leptin deficiency enhances sensitivity of rats to alcoholic steatohepatitisthrough suppression of metallothionein. Am J Physiol Gastrointest LiverPhysiol 2004;287:1078-1085.

12. Leclercq IA, Farrell GC, Field J, Bell DR, Gonzalez FJ, Robertson GR.CYP2E1 and CYP4A as microsomal catalysts of lipid peroxides in murinenonalcoholic steatohepatitis. J Clin Invest 2000;105:1067-1075.

13. Liu Y, Zhang H, Zhang L, Zhou Q, Wang X, Long J, et al. AntioxidantN-acetylcysteine attenuates the acute liver injury caused by X-ray in mice.Eur J Pharmacol 2007;575:142-148.

14. Weng D, Lu Y, Wei Y, Liu Y, Shen P. The role of ROS in microcystin-LR-induced hepatocyte apoptosis and liver injury in mice. Toxicology2007;232:15-23.

15. Lee H-U, Bae E-A, Han MJ, Kim N-J, Kim D-H. Hepatoprotective effectof ginsenoside Rb1 and compound K on tert-butyl hydroperoxide-inducedliver injury. Liver Int 2005;25:1069-1073.

16. Dubuisson ML, de Wergifosse B, Trouet A, Baguet-F, Marchand-BrynaertJ, Rees JF. Antioxidative properties of natural coelenterazine and syntheticmethyl coelenterazine in rat hepatocytes subjected to tert-butyl hydroper-oxide-induced oxidative stress. Biochem Pharmacol 2000;60:471-478.

17. Cerutti PA. Prooxidant states and tumor promotion. Science 1985;227:375-381.

Table 1. Effects of MCD Diet on Hepatic mRNA Levels inMice

Gene MCD (3 weeks) MCD (8 weeks)

PPAR-� 1.31 � 0.18 0.46 � 0.02*ACO 0.79 � 0.12 0.69 � 0.03*Catalase 0.72 � 0.04* 0.43 � 0.04*SOD1 1.03 � 0.06 0.68 � 0.09*SOD2 1.09 � 0.12 0.62 � 0.05*AdipoR1 1.18 � 0.17 1.19 � 0.14AdipoR2 0.93 � 0.13 0.73 � 0.11

Results (and standard error of the mean) from mice fed the MCD diet (n �4-5/group). Quantifications were normalized for RNA from the liver in mice fed thecontrol diet for the same total period of time.

*P � 0.05 versus the corresponding control group.

472 TOMITA ET AL. HEPATOLOGY, August 2008

18. Yamauchi T, Nil Y, Maki T, Kobayashi M, Takazawa T, Iwabu M, et al.Targeted disruption of AdipoR1 and AdipoR2 causes abrogation ofadiponectin binding and metabolic actions. Nat Med 2007;13:332-339.

19. Tsuchida A, Yamauchi T, Ito Y, Hada Y, Maki T, Takekawa S, et al.Insulin/Foxo1 pathway regulates expression levels of adiponectin receptorsand adiponectin sensitivity. J Biol Chem 2004;279:30817-30822.

20. Toyama T, Nakamura H, Harano Y, Yamauchi N, Morita A, Kirishima T,et al. PPAR� ligands activate antioxidant enzymes and suppress hepaticfibrosis in rats. Biochem Biophys Res Commun 2004;324:697-704.

21. Sreekumar B, Rosado B, Rasmussen D, Charton M. Hepatic gene expres-sion in histologically progressive nonalcoholic steatohepatitis. HEPATOL-OGY 2003;38:244-251.

22. Day CP. Genes or environment to determine alcoholic liver disease andnon-alcoholic fatty liver disease. Liver Int 2006;26:1021-1028.

23. Poli G. Pathogenesis of liver fibrosis: role of oxidative stress. Mol AspectsMed 2000;21:49-98.

24. Matsuoka M, Tsukamoto H. Stimulation of hepatic lipocyte collagen pro-duction by Kupffer cell-derived transforming growth factor beta: implica-tion for a pathogenetic role in alcoholic liver fibrogenesis. HEPATOLOGY

1990;11:599-605.25. Galli A, Svegliati-Baroni G, Ceni E, Milani S, Ridolfi F, Salzano R, et al.

Oxidative stress stimulates proliferation and invasiveness of hepatic stellate cellsvia a MMP2-mediated mechanism. HEPATOLOGY 2005;41:1074-1084.

26. Mizuguchi Y, Yokomuro S, Mishima T, Arima Y, Shimizu T, KawahigashiY, et al. Short hairpin RNA modulates transforming growth factor � sig-naling in life-threatening liver failure in mice. Gastroenterology 2005;129:1654-1662.

27. Day CP, James OF. Steatohepatitis: a tale of two “hits”? Gastroenterology1998;114:842-845.

HEPATOLOGY, Vol. 48, No. 2, 2008 TOMITA ET AL. 473