IFN- deficiency attenuates hepatic inflammation and fibrosis in a steatohepatitis model induced by a methionine- and choline-deficient high-fat diet

IFN-� deficiency attenuates hepatic inflammation and fibrosis in a steatohepatitismodel induced by a methionine- and choline-deficient high-fat diet

Xiao-Yu Luo,1,4,8 Terumi Takahara,2 Kengo Kawai,2 Masayuki Fujino,1,5 Toshiro Sugiyama,2

Koichi Tsuneyama,3 Kazuhiro Tsukada,4 Susumu Nakae,6,7 Liang Zhong,8 and Xiao-Kang Li11Division for Transplantation Immunology, National Institute for Child Health and Development, Tokyo, Japan; 2ThirdDepartment of Internal Medicine, University of Toyama, Toyama, Japan; 3Department of Diagnostic Pathology, University ofToyama, Toyama, Japan; 4Second Department of Surgery, University of Toyama, Toyama, Japan; 5AIDS Research Center,National Institute of Infectious Diseases, Tokyo, Japan; 6Laboratory of Systems Biology, Center for Experimental Medicineand Systems Biology, The Institute of Medical Science, The University of Tokyo, Tokyo, Japan; 7Precursory Research forEmbryonic Science and Technology, Japan Science and Technology Agency, Saitama, Japan; and 8Department ofGastroenterology, Huashan Hospital, Fudan University, Shanghai, China

Submitted 25 June 2013; accepted in final form 5 October 2013

Luo X, Takahara T, Kawai K, Fujino M, Sugiyama T, TsuneyamaK, Tsukada K, Nakae S, Zhong L, Li X. IFN-� deficiency attenuateshepatic inflammation and fibrosis in a steatohepatitis model induced by amethionine- and choline-deficient high-fat diet. Am J Physiol Gastroin-test Liver Physiol 305: G891–G899, 2013. First published October 17,2013; doi:10.1152/ajpgi.00193.2013.—Cytokines play important rolesin all stages of steatohepatitis, including hepatocyte injury, the in-flammatory response, and the altered function of sinusoidal cells. Thisstudy examined the involvement of a major inflammatory cytokine,interferon-� (IFN-�), in the progression of steatohepatitis. In a ste-atohepatitis model by feeding a methionine- and choline-deficienthigh-fat (MCDHF) diet to both wild-type and IFN-�-deficient mice,the liver histology, expression of genes encoding inflammatory cyto-kines, and fibrosis-related markers were examined. To analyze theeffects of IFN-� on Kupffer cells in vitro, we examined the tumornecrosis factor-� (TNF-�) production by a mouse macrophage cellline. Forty two days of MCDHF diet resulted in weight loss, elevatedaminotransferases, liver steatosis, and inflammation in wild-typemice. However, the IFN-�-deficient mice exhibited less extensivechanges. RT-PCR revealed that the expression of tumor necrosisfactor-� (TNF-�), transforming growth factor-�, inducible nitric ox-ide synthase, interleukin-4 and osteopontin were increased in wild-type mice, although they were suppressed in IFN-�-deficient mice.Seventy days of MCDHF diet induced much more liver fibrosis inwild-type mice than in IFN-�-deficient mice. The expression levels offibrosis-related genes, �-smooth muscle actin, type I collagen, tissueinhibitor of matrix metalloproteinase-1, and matrix metalloprotei-nase-2, were dramatically increased in wild-type mice, whereas theywere significantly suppressed in IFN-�-deficient mice. Moreover, invitro experiments showed that, when RAW 264.7 macrophages weretreated with IFN-�, they produced TNF-� in a dose-dependent man-ner. The present study showed that IFN-� deficiency might inhibit theinflammatory response of macrophages cells and subsequently sup-press stellate cell activation and liver fibrosis. These findings highlightthe critical role of IFN-� in the progression of steatohepatitis.

hepatic stellate cell; interferon-�; macrophage

STEATOHEPATITIS IS ONE OF the leading causes of liver-relatedmorbidity and mortality in developed Western countries. Thefeatures of steatohepatitis, regardless of whether it is nonalco-holic (NASH) or alcoholic steatohepatitis include steatosis,

liver cellular damage, inflammation, and varying degrees offibrosis (48, 52, 70). Although the exact mechanisms that causesimple liver steatosis to progress to steatohepatitis remainpoorly understood, the “two-hit” hypothesis is the most com-monly accepted model explaining such progression. Steatosisin this model represents the “first hit” that sensitizes cellsvulnerable to subsequent stress. The “second hit” can includeoxidative stress, recruitment of inflammatory cells, and dys-regulated cytokine/adipokine production, which all synergisti-cally lead to hepatocyte death by apoptosis or necrosis, andsubsequent liver inflammation and fibrosis. In the past, it wasgenerally believed that inflammation is followed by the devel-opment of hepatic steatosis (69). Cytokines have been shownto be central mediators of inflammation in steatohepatitis (15,39, 46, 53). The effect of tumor necrosis factor-� (TNF-�) onthe pathogenesis of liver steatohepatitis has been investigatedin many studies, and it has been clearly demonstrated that theliver and adipose tissue TNF-� and TNF receptor 1 transcripts,as well as the serum TNF-� levels, were increased in patientswith steatohepatitis (4). However, the role of interferon-�(IFN-�), a critical and pleiotropic cytokine, in the developmentof liver steatohepatitis is not yet clearly understood.

The previous studies mainly focused on examining the contri-bution of IFN-� to acute liver and intestinal injuries in animalmodels, revealing that IFN-� was pivotal in aggravating the acuteliver injury induced by concanavalin A or lipopolysaccharide (10,72) and that it played a central role in the intestinal inflammationinduced by interleukins (11). In most experiments, IFN-� produc-tion is generally considered to antagonize the development of liverfibrosis. For example, IFN-�-deficient mice are more susceptibleto liver fibrosis induced by carbon tetrachloride (CCl4) (59), andthe antifibrogenic effect of IFN-� is believed to be mediated viainhibiting hepatic stellate cell (HSC) activation and TGF-� sig-naling (6, 51, 56, 57, 73). In contrast, in mice fed the MCDdiet-induced steatohepatitis, Yu et al. identified significantly in-creased proinflammatory cytokines, including IFN-�, CXCL1,CXCL10, and CCL3 (80). Furthermore, fatty liver in mice fed onhypercaloric or choline-deficient diets promotes IFN-� production(39, 46). IFN-� is pivotal for efficient innate and adaptive immuneresponses, and a detrimental role of IFN-� in the initiation and/ormaintenance of proinflammatory activation in development ofobesity-associated insulin resistance and steatohepatitis has beenreported (39, 46, 53). In addition, natural killer (NK) cells can kill

Address for reprint requests and other correspondence: X.-K. Li, Division ofTransplantation Immunology, National Research Institute for Child Health andDevelopment, 2-10-1 Okura, Setagaya-ku, Tokyo, 157-8535 Japan (e-mail: [email protected]).

Am J Physiol Gastrointest Liver Physiol 305: G891–G899, 2013.First published October 17, 2013; doi:10.1152/ajpgi.00193.2013.

0193-1857/13 Copyright © 2013 the American Physiological Societyhttp://www.ajpgi.org G891

mailto:[email protected]

mailto:[email protected]

activated HSC but not quiescent HSC, and activation of NK cellsinduced HSC death and ameliorated liver fibrosis in a mousemodel of liver fibrosis. This effect was attributed to IFN-� (55).However, these studies were performed by CCl4 or dimethylni-trosamine injection, or 3,5-diethoxycarbonyl-1,4-dihydrocollidinediet. The mechanisms of induction of fibrosis by those models aredifferent from steatohepatitis. There have also been very fewstudies of the effects of IFN-� on a liver fibrosis model induced bya methionine- and choline-deficient high-fat (MCDHF) diet, oneof the most efficient models of induced steatohepatitis in rodents,and these findings still remain controversial. The present studywas therefore undertaken to clarify the importance of IFN-� in thedevelopment of liver steatohepatitis by using IFN-�-deficientmice.

MATERIALS AND METHODS

Animal models. IFN-�-deficient mice on the C57BL/6J backgroundwere generated as described previously (64). Wild-type C57BL/6mice were purchased from Shizuoka Laboratory Animal Center (Shi-zuoka, Japan). Ten-week-old male mice were housed four per cage intemperature- and light-controlled chambers for all experiments. Ste-atohepatitis was induced by feeding mice a MCDHF diet containingcorn oil and sucrose [40% (wt/wt) fat and 40% (wt/wt) carbohydrates](40, 41) for 42 days. Liver fibrosis was induced by feeding mice theMCDHF diet for 70 days. The liver was excised and divided intoseveral parts for hematoxylin-eosin (HE) staining, immunostainingexamination, and RNA extraction. All animal experiments were reviewedand approved by the Committee on the Care and Use of LaboratoryAnimals at the National Research Institute for Child Health and Devel-opment.

Histopathological examination. Liver tissue samples were kept in10% formalin solution. Paraffin blocks were prepared as 4-�m crosssections, and HE staining and Sirius red staining were performed. Thefibrotic areas were measured in three sections per mouse using animage analyzing system (VH analyzer; KEYENCE, Osaka, Japan).

Serum biochemical detection. Whole blood was collected, andserum was then evaluated for alanine aminotransferase (ALT), aspar-tate aminotransferase (AST), triglycerides, and cholesterol, expressedas units per liter or milligram per deciliter, respectively.

RNA preparation and quantitative reverse transcriptase-polymer-ase chain reaction. The total RNA was extracted from frozen livertissue using ISOGEN (Nippon Gene, Tokyo, Japan). Each 800-ngRNA sample was reverse-transcribed to cDNA using oligo(dT) prim-ers and Super Script reverse transcriptase (Invitrogen, Life Technol-ogies Japan) according to the manufacturer’s protocol. The target-specific primers and probes were designed on the basis of the reportedcDNA sequences and were synthesized by Biosearch Technologies(Novato, CA). Quantitative RT-PCR was performed using the Taq-Man system on the Applied Biosystem PRISM7700 instrument (LifeTechnologies Japan). Quantitative RT-PCR was conducted in 0.9 mMeach primer in a 25-�l final reaction volume of Premix Ex Taq(Takara Bio, Shiga, Japan). The PCR cycling conditions were asfollows: 50°C for 2 min, 95°C for 15 min and 50 cycles of 95°C for30 s, 60°C for 1 min, and 25°C for 2 min. The data were expressed asthe comparative cycle threshold (Ct) values. The normalized Ct valueof each gene was obtained by subtracting the Ct value of 18s rRNA.The fold change vs. one sample of the control group was calculated asdescribed previously (49).

Immunohistochemical examination. Formalin-fixed and paraffin-embedded sections of the livers were used in this study. A ratmonoclonal antibody against F4/80, a surface marker of mousemonocytes/macrophages (dilution 1:200; Serotec, Oxford, UK), andhamster antimouse CD11c antibodies (dilution, 1:50; clone N418;AbD Serotec, Oxford, UK) to detect M1 macrophages were applied tothe sections, and �-smooth muscle actin (�-SMA) staining was

conducted to show HSC activation (dilution 1:100; Dako Japan,Tokyo, Japan). The sections were incubated with appropriate second-ary antibodies, and the immunoreactive products were visualizedusing a DAB reagent and counterstained with hematoxylin. Thepositive F4/80 cells were detected by using an image analyzingsystem (VH analyzer; KEYENCE). The numbers of F4/80 positivecells were counted on 10 high-power (�200) fields per slide. Theintensity of the immunostaining of CD11c was scored as 0 (none), 1(faint), 2 (moderate), or 3 (intense). The scoring of immunostainingwas performed by two independent examiners.

Determination of the TNF-� production. The mouse monocyte/macrophage cell line, RAW 264.7, and IFN-� reagent were obtainedfrom the ATCC (Rockville, MD) and Invitrogen, respectively. RAW264.7 cells were cultured in 10% FBS DMEM (GIBCO, Life Tech-nologies Japan) containing 0, 100, or 500 �g/ml IFN-� and then wereincubated to 70% confluence. The TNF-� production of the RAW cellswas analyzed by an ELISA kit (R&D Systems, Minneapolis, MN).

Statistical analysis. The data are presented as the means � SE andwere analyzed statistically using a one-way ANOVA, followed byFisher’s protected least-significance difference test or the Mann-Whitney U-test. Values of P � 0.05 were considered to be statisticallysignificant.

RESULTS

IFN-� deficiency attenuated the loss of body weight inducedby MCDHF. Initially, the average body weight was not signif-icantly different between the wild-type and IFN-�-deficientmice, but, after the mice had consumed the MCDHF diet for 42days, they showed significant differences. The wild-type micedecreased in weight by 9.9 g on average, about 38% of theirinitial body weight, whereas the weight of the IFN-�-deficientmice decreased by 7.0 g on average, about 28% of their initialbody weight (Fig. 1A).

IFN-� deficiency reduced the MCDHF-induced liver inflam-mation. As in our previous study, the MCDHF diet inducedliver steatosis (40, 41). In this study, fatty droplets formed inboth wild-type and IFN-�-deficient mice, inducing macrove-sicular and microvesicular steatosis after 42 days of theMCDHF diet. The quantitative analysis of the percentage ofsteatotic area was similar in both groups (data not shown),whereas HE staining revealed a different degree of inflamma-tory cell infiltration between the two groups. In wild-type mice,the clusters of inflammatory cells and enlarged form cells wereclear and prevalent, which were scarcely seen in IFN-�-defi-cient mice (Fig. 1B). Consistent with the HE staining, F4/80immunostaining revealed a marked increase in the number ofactivated and infiltrated Kupffer/macrophage cells in wild-typemice, which were localized at the enlarged cells detected asform cells; however, these were considerably reduced in IFN-�-deficient mice (Fig. 1C). We further analyzed the surfacemarkers of macrophages, and immunostaining for CD11c showedthat the enlarged form cells and small cells around the form cellswere strongly positive for CD11c in wild-type mice, whereas thiswas seldom seen in IFN-�-deficient mice (Fig. 1D).

Because macrophages are the primary source of inflamma-tory cytokines in this steatohepatitis model, we analyzed theliver tissue mRNA expression levels of several inflammatorycytokines, including TNF-�, interleukin (IL)-4, transforminggrowth factor (TGF)-�1, inducible nitric oxide synthase, os-teopontin (OPN), and IFN-�. In wild-type mice, the IFN-�levels were significantly increased after 42 days of the MCDHF

G892 IFN-� DEFICIENCY REDUCES STEATOHEPATITIS

AJP-Gastrointest Liver Physiol • doi:10.1152/ajpgi.00193.2013 • www.ajpgi.org

diet, whereas the level was undetectable in IFN-�-deficientmice at all of the time points examined. The fact that IFN-�was not able to be detected in IFN-�-deficient mice demon-strated the accuracy of the gene knockout animal used in thisstudy. The levels of other inflammatory cytokines were alsoobviously increased after 42 days of the MCDHF diet inwild-type mice. Of note, the gene expression of these inflam-

matory cytokines was significantly suppressed in IFN-�-deficient mice fed the MCDHF diet (Fig. 1E).

IFN-� deficiency protected the liver against MCDHF-in-duced injury. Feeding wild-type mice the MCDHF diet for 42days resulted in 15- and 4.5-fold increases in the levels of ALTand AST, respectively, whereas IFN-� deficiency significantlyinhibited the increase in serum aminotransferase (Fig. 2A). In

Wild typeBody weight (g)

IFN-gamma _/_

Body weight (g)

MCDHFDay0

25.93 ± 0.94 25.18 ± 0.84

MCDHFDay42

16 ± 1.69 18.12± 0.87**

6789

1011 12

Bod

y w

eigh

t los

s (g

) wild type

IFN-gamma -/-

A Bwild type

IFN-gamma _/_

DWild type

IFN-gamma _/_

F4/8

0 po

sitiv

e ce

lls/fi

led

0

4

8

12

16

20 **

CWild type

IFN-gamma _/_

100 µ m

100 µ m

Sco

re o

f CD

11c

posi

tive

cells **

0day 42day

TNF-

wild type IFN-gamma-/-

0day 42day

IFN

-gam

ma wild type

IFN-gamma -/-

0day 42day

IL-4


0day 42day

TGF-

β


0day 42day

iNO

S


0day 42day

OPN


**

**

*

* *

*

E

Fig. 1. The body weight loss and inflammatory cell infiltration were attenuated in interferon-� (IFN-�)-deficient mice fed the methionine- and choline-deficienthigh-fat (MCDHF) diet for 42 days. A: the body weight loss in wild-type and IFN-�-deficient groups. B: large fatty droplets, form cells (red arrows), and theinfiltration of inflammatory cells (black arrows) in wild-type mice. Fatty droplets were similarly present, but inflammatory cells were significantly decreased, inIFN-�-deficient mice. C: numerous crowded clusters of F4/80-positive macrophages were recruited in wild-type mice (arrows), which were significantly reducedin IFN-�-deficient mice. The results of quantification of F4/80-positive macrophages analyzed using an image analyzing system (VH analyzer; KEYENCE). Scalebars represent 100 �m (means � SE; **P � 0.01 vs. wild-type mice). D: CD11c was weakly detected in the form cells in the wild-type mice (arrows), whereF4/80 was coimmunolocalized, whereas CD11c was significantly reduced in IFN-�-deficient mice. Small cells around the form cells were clearly positive forCD11c in the wild-type mice (arrowheads). Scale bars represent 100 �m. E: the expression of inflammatory cytokine genes was downregulated in IFN-�-deficientmice after they were fed the MCDHF diet for 42 days. TNF-�, tumor necrosis factor-�; IL-4, interleukin-4; TGF-�, transforming growth factor-�; iNOS,inducible nitric oxide synthase; OPN, osteopontin. The data are representative of 4–5 independent experiments and indicate the mean ratio of triplicate resultsfrom each experiment (arbitrary/unit, means � SE; *P � 0.05 and **P � 0.01 vs. wild-type mice).

G893IFN-� DEFICIENCY REDUCES STEATOHEPATITIS


both wild-type and IFN-�-deficient mice that had not fasted,the MCDHF diet induced decreased serum triglyceride andcholesterol levels; however, the change in the cholesterollevels in wild-type mice was slightly larger (Fig. 2B).

IFN-� deficiency reduced the MCDHF-induced liver fibrosis.Sirius red staining showed a moderate perisinusoidal collagendeposition starting from the central area and extending into thehepatic lobules, which represented 13% of the liver area inwild-type mice after 70 days of MCDHF diet (Fig. 3A). Thisamount of collagen was effectively reduced in IFN-�-deficientmice fed the MCDHF diet, accounting for 3% of the liver area(Fig. 3A), which demonstrated that the collagen deposition waslargely suppressed by IFN-� deficiency.

In the �-SMA immunostaining, numerous �-SMA-positivecells were located around the central areas and infiltrated intothe middle part of the lobules around the form cells afterwild-type mice were fed the MCDHF diet for 70 days (Fig.3B). However, there were very few �-SMA-positive cells,except in some vessel areas, in the IFN-�-deficient mice after70 days of the MCDHF diet (Fig. 3B). This indicated thatactivation of HSCs into �-SMA-positive myofibroblasts waslargely prevented by IFN-� deficiency. We also analyzed theliver tissue fibrosis-related gene expression levels, includingthose of �-SMA, type I collagen, tissue inhibitor of matrixmetalloproteinase-1 (TIMP-1), and matrix metalloproteinase-2(MMP-2), which were obviously increased in wild-type micebut were significantly decreased in IFN-�-deficient mice com-pared with wild-type mice (Fig. 3C). Together, these resultsshowed that IFN-� deficiency inhibited the progression offibrosis in response to MCDHF-induced steatohepatitis.

IFN-� promoted TNF-� production of macrophages in vitro.Although we have demonstrated that IFN-� deficiency sup-pressed the infiltration of activated Kupffer/macrophage cells

and prevented the steatohepatitis induced by the MCDHF diet,to further evaluate the influence of IFN-� on Kupffer cells invitro, we examined the TNF-� production by a mouse macro-phage cell line (RAW 264.7). We found that, when the RAW264.7 cells were treated with different concentrations of IFN-�in the culture medium, there was a dose-dependent increase inTNF-� production compared with the control group (Fig. 4).

DISCUSSION

In our study, the MCDHF diet induced serum biochemical,liver histological, and molecular changes in wild-type mice,whereas these did not occur, or were much less severe, inIFN-�-deficient mice. Of interest, the mRNA levels of IFN-�were increased in the liver tissue of wild-type mice fed theMCDHF diet (Fig. 1E). This result was supported by a studyrevealing that the IFN-� expression was enhanced in steato-hepatitis patients with hepatitis C virus infection (23). Theincrease of the IFN-� levels in different types of steatohepatitisconfirms its importance in steatohepatitis arising because ofvarious pathogenic processes. In the present study, the levels ofother inflammatory cytokines also simultaneously increased inthe wild-type mice but were largely suppressed in IFN-�-deficient mice (Fig. 1E). More importantly, this phenomenonwas consistent with hepatic F4/80 immunostaining, a marker ofmacrophage/Kupffer cells, which showed an obvious increasein wild-type mice that was inhibited in IFN-�-deficient micefed the same MCDHF diet (Fig. 1C). In addition, CD11c, amarker of proinflammatory M1 macrophages, was also prom-inently expressed in wild-type mice and was significantlysuppressed in IFN-�-deficient mice (Fig. 1D). To further esti-mate the influence of IFN-� on macrophages/Kupffer cells invitro, the TNF-� production by RAW 264.7 cells was con-

A B

0day 42day

Ser

um T

rigly

cerid

e (m

g/dl

)


0day 42day

Ser

um C

hole

ster

ol (m

g/dl

) wild type IFN-gamma-/-

0

50

100

150

200

250

0day 42day

Ser

um A

LT (U

/L)


0

50

100

150

200

0day 42day

Ser

um A

ST

(U/L

)


**

**

*

Fig. 2. The serum alanine aminotransferase(ALT) and aspartate aminotransferase (AST)levels (A) and triglyceride and cholesterol levels(B) in wild-type and IFN-�-deficient mice fedthe MCDHF diet for 42 days (means � SE;*P � 0.05 and **P � 0.01 vs. wild-type mice).



firmed to dose-dependently increase by IFN-� (Fig. 4), con-sistent with previous reports (18, 74). Considering the inter-connection between resident Kupffer cells, infiltrated macro-phages, and various cytokines, as well as their pivotal roles inthe progression of steatohepatitis reported in previous studies(7, 12, 13, 26, 60, 71), we suggest that decreased macrophagerecruitment or reduced Kupffer cells in the environment in-duced by IFN-� deficiency could account for the decreasedmRNA levels of the above inflammatory cytokines as de-

scribed previously using a T cell-mediated hepatitis rodentmodel (31).

In our study, although IFN-� appeared to contribute exten-sively to the inflammatory response, it did not appear to havea role in the initiation or aggravation in steatosis, as seen in theliver histology and serum biochemistry examinations. Interest-ingly, IL-6-deficient mice fed the MCDHF diet also did notshow any major change in the progression of steatosis in theirstudy (52). In the steatohepatitis model induced by MCDHF,

A

B wild type IFN-gamma / **

0

5

10

15

20

25

αS

MA

posi

tive

cells

/fiel

d

IFN-gamma /wild type

0

4

8

12

16

Siri

usre

d st

aini

ng a

rea

(%) **

0 day 70 day 0 day 70 day

0 day 70 day

0 day 70 day

0 day 70 day 0 day 70 day

wild type

IFN-gamma-/-

α-S

MA

MM

P-13MM

P-9

MM

P-2

colla

gen-

1

TIM

P

*

*

* *C

Fig. 3. A: Sirius red staining in wild-type and IFN-�-deficient mice. Scale bars represent 200 �m. The results of quantification of the areas positive for Siriusred staining (means � SE; **P � 0.01 vs. wild-type mice). B: �-smooth muscle actin (�-SMA) immunostaining in wild-type and IFN-�-deficient mice. Scalebars represent 100 �m. The results of the quantification of �-SMA-positive cells (means � SE; **P � 0.01 vs. wild-type mice). C: liver fibrosis-related geneexpression was downregulated in IFN-�-deficient mice after they were fed the MCDHF diet for 70 days. TIMP, tissue inhibitor of matrix metalooproteinase;MMP, matrix metalloproteinase. The data are representative of 4–5 independent experiments and indicate the mean ratio of triplicate results from eachexperiment (arbitrary/unit, means � SE; *P � 0.05 vs. wild-type mice).



our findings therefore suggest that the development of liversteatosis is mainly the result of the diet itself, rather than IFN-�or other inflammatory cytokines.

In a previous study, IFN-� was demonstrated to be centrallyinvolved in acute liver cell injury as assessed by the serumALT and AST levels (29). In our present study, IFN-� defi-ciency also effectively attenuated the MCDHF diet-inducedaugmentation of the ALT and AST levels (Fig. 2A). This wasprobably related to the fact that the decreased levels of cyto-kines such as TNF-�, a death factor for hepatocytes, lessenedthe damage to hepatocytes. Taken together, our data suggestthat IFN-� was not only involved in the liver inflammation butalso in the hepatocyte injury throughout the progression ofsteatohepatitis. In fact, IFN-� induces hepatocyte apoptosis(25), likely because inflammatory cytokines and cell injury areusually interconnected.

In human NASH, in general, patients gain body weight.Unlike with human NASH, murine NASH by MCDHF dietresults in body weight loss (Fig. 1A). Similar findings were alsoreported by another group who demonstrated that IFN-�-deficient mice showed an obviously attenuated weight losscompared with wild-type mice when they received the sameintraperitoneal injections of IL-12 and IL-18 (11). We proposethat IFN-� deficiency is beneficial for protecting mice againstthe weight loss induced by MCDHF, even though some weightloss still occurred in the IFN-�-deficient mice.

In wild-type mice fed the MCDHF diet for 70 days, theSirius red staining and �-SMA-immunopositive cells weresignificantly increased, and the liver mRNA expression levelsof type I collagen, �-SMA, TIMP-1, and MMP-2 were alsodramatically elevated, which demonstrated liver fibrosis. Theincrease of MMP-2 was likely the result of the inflammation ofthe liver, as we reported previously (66). Compared withwild-type mice, the Sirius red staining and �-SMA-positivecell immunostaining both decreased (Fig. 3, A and B), and thegene expression levels of type I collagen, �-SMA, TIMP-1,and MMP-2 also significantly decreased (Fig. 3C) in IFN-�-deficient mice, which indicated a reduced liver collagen con-tent and decreased HSC activity resulting from the MCDHF

diet. These results provided apparent evidence of an IFN-�dependence of the liver fibrosis. Interestingly, IFN-� has beenshown in numerous previous studies to be antifibrogenic, byboth inhibiting the activation of HSCs in culture as well asattenuating fibrosis in models such as the CCl4 and dimethyl-nitrosamine, and IFN-� treatment ameliorated liver fibrosis(56). IFN-� displays antifibrotic effects in HSCs via the im-pairment of TGF-� signaling (77), the inhibition of collagenproduction (8), and the suppression of SMA expression (56,58). Furthermore, IFN-� activation of signal transducer andactivator of transcription 1 is effective in ameliorating liverfibrosis in animal models, and IFN-� treatment improves thefibrosis scores in patients with chronic hepatitis B virus infec-tion (78, 79). In addition, IFN-� inhibits extracellular matrix/collagen expression in stellate cells by virtue of a number ofeffects on stellate cells (5, 6, 27, 28, 57). On the other hand,IFN-� is a very potent proinflammatory cytokine with a ubiq-uitous receptor expression, and therefore IFN-�-based experi-mental therapies are associated with side effects like severeflu-like symptoms, systemic endothelial and immune cell acti-vation, neurotropic effects, and hyperlipidemia (36, 54). IFN-�also works in accumulation of neutrophils and macrophages inthe liver (42) and is known to be a key molecule in theinduction of type I polarization (34, 43). Given that, in thepresent study, the antifibrotic effect by IFN-� attenuationmight be because of the reduction of inflammation and not adirect effect of IFN-� to HSCs. In concordance, caspase-1, animportant component of inflammation, knockout mice on theMCD diet showed marked reduction in mRNA expression ofgenes involved in inflammation and fibrogenesis with signifi-cant reduction of hepatic collagen deposition (14). In contrast,some studies reported a different result. For example, IFN-�therapy was not able to attenuate or reverse liver fibrosis in adouble-blind clinical trial including 502 patients (54), andIFN-� itself promoted the hepatic progenitor cell response andexacerbated fibrosis in a chronic liver injury model (38).Considering the unexpected result of attenuated liver fibrosisresulting from IFN-� deficiency in our present study, we madethe following assumptions. There might have been a priorevent that ameliorated the liver injury and inflammation in-duced by decreasing the macrophages/Kupffer cell infiltrationand suppressing the inflammatory response seen in IFN-�-deficient mice, and this would probably underlie the attenuatedliver fibrosis. A number of previous studies demonstrated thatthe severity of liver fibrosis was closely related to inflamma-tory cytokines, because they triggered stellate cell activationinto �-SMA-positive myofibroblasts, orchestrating a cross talkbetween different cell types and different stages of steatohepa-titis (17, 71). In fact, IFN-� induces the accumulation ofneutrophils and macrophages in the liver (42) and also inducestype I polarization (34, 43). Hepatic accumulation of inflam-matory cells is generally greater in NASH than in steatosis,suggesting that activation of the immune system may contrib-ute to progression of fatty liver damage. The liver harborsresident populations of cells that regulate innate immune re-sponses (22). The mRNA expression and histological analysisrevealed significantly higher expression of IFN-� and cellularinfiltration in liver by MCHD diet (Fig. 1, B and E). These datasuggested the possibility that the infiltrated cells might expressabundant IFN-�. Several cell populations have been known toexpress IFN-�, in not only CD4� T cells, CD8� T cells, � T

TNF-

α p

rodu

ctio

n (p

g/m

l) **

** **

Fig. 4. IFN-� dose-dependently stimulated the tumor necrosis factor-�(TNF-�) production of RAW 264.7 cells in vitro (means � SE; *P � 0.01 vs.each value).



cells, natural killer T (NKT) cells, and NK cells but also inmacrophages, dendritic cells, and B cells (20). For instance,NF-B1 deficiency stimulates the progression of NASH inassociation with the MCD diet in mice by promoting NKTcell-mediated responses with an upregulation in the productionof IFN-� and OPN (47). Additionally, many previous studieshave suggested that the accumulation of infiltrated NKT cellsin the liver is involved in pathogenesis of steatohepatisis, andIFN-� may involve it (1, 30, 47, 62, 63, 65).

In addition to NKT cells, a previous study on a pediatricNASH patient described that the hepatic microenvironment isdominated by IFN-� but not IL-4, and it is infiltrated by ahigher number of CD8� cells. The number of infiltratingneutrophils positively correlated with reactive oxygen speciesgeneration by peripheral polymorphonuclear cells. A distinc-tive increase in CD8� CD45RO and CD8� CD45RA subpopu-lations and an increased production of IFN-� by CD4� andCD8� cells was also demonstrated (16). The patients withNASH demonstrated a significantly higher ratio of IFN-��

CD4 T cells in the liver (32). A NASH murine model that wasfed atherogenic high-fat diet demonstrated an increased CD8�/CD4� T cell ratio that is also comparable with the clinicalNASH patient pathology (61). Helper T cell activation, whichinduces the production of Th1 cytokines, is thus considered tobe a pathogenic finding in steatohepatitis (16, 53). In addition,in patients with NASH, the ratio of neutrophils to lymphocytesincreases (2), suggesting that granulocytes are involved in thepathogenesis of NASH. Furthermore, many of the recent datashowed that innate immune processes both within and outsidethe liver are involved in NASH (50, 81). Both the previousstudies and our findings suggest that a decrease in the IFN-�levels might therefore inhibit the infiltration of inflammatorycells.

Among these cytokines, TNF-� signaling is considered to beespecially important for liver fibrosis (3, 21, 71, 75), andTGF-�1 plays a critical role in liver fibrosis (9, 19, 33).Furthermore, OPN, which is an activator of HSC (62, 63) thatis induced by IFN-�, is upregulated by the MCDHF diet (Fig.1E). Both mRNA and protein of OPN were abundantly ex-pressed in Kupffer cells, macrophages, and stellate cells acti-vated in the liver of rats given CCl4 or heat-killed Propionibacterium acnes (35, 76). In an in vitro study, OPN wasinduced in mouse macrophages (RAW 264.7 and ANA-1 cells)by treatment with IFN-� (24, 67, 68). Further IFN-� treatmentinduced OPN expression in both THP-1 monocytes and pri-mary human blood monocytes (45). In the present study, themRNA expression levels of TNF-�, TGF-�1, and OPN wereeffectively reduced in IFN-�-deficient mice compared withwild-type mice (Fig. 1E). The existence of IFN-� failed toprevent HSC activation, whereas a deficiency of IFN-� didprevent such activation. Therefore, considering that a defi-ciency of IFN-� could suppress the macrophage/Kupffer cellactivation and infiltration, and the subsequent inflammatoryresponse, further suppression of HSC activation may have ledto the attenuated fibrosis, not turn off the direct effect of IFN-�to HSC. Although the phenomenon was not consistent with thefindings of previous studies using other steatohepatitis models,it guided us to a hypothesis that the pathway mediating HSCactivation and function in MCDHF-induced steatohepatitis wasdifferent from those involved in other liver injury models, suchas CCl4 intoxication. Furthermore, Knight et al. reported a

similar finding in their recent study in which they used IFN-�-deficient mice to generate a steatohepatitis model using acholine-deficient, ethionine-supplemented diet (37, 38). Moreresearch regarding the role of IFN-� in related liver diseases istherefore needed.

In summary, in the present study, we discovered that IFN-�appears to modulate the inflammatory response in a mousesteatohepatitis model, whereas the improvement of inflamma-tion and fibrosis in IFN-�-deficient mice was linked to theinhibition of macrophages/Kupffer cell infiltration, the inflam-matory response, and HSC activation. We also suggest that theMCDHF-induced steatohepatitis model might have a distinctpathway mediating the fibrosis involving myofibroblast activ-ity, which led us to reconsider the influence of IFN-� �n theprogression of liver fibrosis because IFN-� was previouslyexamined in clinical trials as an antifibrogenic therapeuticagent for chronic liver disease.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

Author contributions: X.-Y.L. and K.K. performed experiments; T.T., T.S.,K. Tsuneyama, K. Tsukada, S.N., L.Z., and X.-K.L. conception and design ofresearch; T.T. and X.-K.L. interpreted results of experiments; T.T. and X.-K.L.drafted manuscript; T.T., M.F., and X.-K.L. edited and revised manuscript;X.-K.L. analyzed data; X.-K.L. prepared figures; X.-K.L. approved finalversion of manuscript.

REFERENCES

1. Adler M, Taylor S, Okebugwu K, Yee H, Fielding C, Fielding G, PolesM. Intrahepatic natural killer T cell populations are increased in humanhepatic steatosis. World J Gastroenterol 17: 1725–1731, 2011.

2. Alkhouri N, Morris-Stiff G, Campbell C, Lopez R, Tamimi TA,Yerian L, Zein NN, Feldstein AE. Neutrophil to lymphocyte ratio: a newmarker for predicting steatohepatitis and fibrosis in patients with nonal-coholic fatty liver disease. Liver Int 32: 297–302, 2012.

3. Bahcecioglu IH, Koca SS, Poyrazoglu OK, Yalniz M, Ozercan IH,Ustundag B, Sahin K, Dagli AF, Isik A. Hepatoprotective effect ofinfliximab, an anti-TNF-alpha agent, on carbon tetrachloride-inducedhepatic fibrosis. Inflammation 31: 215–221, 2008.

4. Bahcecioglu IH, Yalniz M, Ataseven H, Ilhan N, Ozercan IH, SeckinD, Sahin K. Levels of serum hyaluronic acid, TNF-alpha and IL-8 inpatients with nonalcoholic steatohepatitis. Hepato-Gastroenterology 52:1549–1553, 2005.

5. Bansal R, Prakash J, Post E, Beljaars L, Schuppan D, Poelstra K.Novel engineered targeted interferon-gamma blocks hepatic fibrogenesisin mice. Hepatology 54: 586–596, 2011.

6. Baroni GS, D’Ambrosio L, Curto P, Casini A, Mancini R, JezequelAM, Benedetti A. Interferon gamma decreases hepatic stellate cell acti-vation and extracellular matrix deposition in rat liver fibrosis. Hepatology23: 1189–1199, 1996.

7. Bertola A, Deveaux V, Bonnafous S, Rousseau D, Anty R, Wakkach A,Dahman M, Tordjman J, Clement K, McQuaid SE, Frayn KN, HuetPM, Gugenheim J, Lotersztajn S, Le Marchand-Brustel Y, Tran A,Gual P. Elevated expression of osteopontin may be related to adiposetissue macrophage accumulation and liver steatosis in morbid obesity.Diabetes 58: 125–133, 2009.

8. Block ET, Cronstein BN. Interferon-gamma inhibits adenosine A2Areceptor function in hepatic stellate cells by STAT1-mediated repressionof adenylyl cyclase. Int J Interferon Cytokine Mediator Res 2010: 113–126, 2010.

9. Border WA, Noble NA. Transforming growth-factor-beta in tissue fibro-sis. N Engl J Med 331: 1286–1292, 1994.

10. Cao Q, Batey R, Pang G, Russell A, Clancy R. IL-6, IFN-gamma andTNF-alpha production by liver-associated T cells and acute liver injury inrats administered concanavalin A. Immunol Cell Biol 76: 542–549, 1998.

11. Chikano S, Sawada K, Shimoyama T, Kashiwamura SI, Sugihara A,Sekikawa K, Terada N, Nakanishi K, Okamura H. IL-18 and IL-12



induce intestinal inflammation and fatty liver in mice in an IFN-gammadependent manner. Gut 47: 779–786, 2000.

12. Day CP. From fat to inflammation. Gastroenterology 130: 207–210, 2006.13. Diehl AM. Nonalcoholic steatosis and steatohepatitis. IV. Nonalcoholic

fatty liver disease abnormalities in macrophage function and cytokines.Am J Physiol Gastrointest Liver Physiol 282: G1–G5, 2002.

14. Dixon LJ, Berk M, Thapaliya S, Papouchado BG, Feldstein AE.Caspase-1-mediated regulation of fibrogenesis in diet-induced steatohepa-titis. Lab Invest 92: 713–723, 2012.

15. Farrell GC, Larter CZ. Nonalcoholic fatty liver disease: from steatosisto cirrhosis. Hepatology 43: S99–S112, 2006.

16. Ferreyra Solari NE, Inzaugarat ME, Baz P, De Matteo E, Lezama C,Galoppo M, Galoppo C, Chernavsky AC. The role of innate cells iscoupled to a Th1-polarized immune response in pediatric nonalcoholicsteatohepatitis. J Clin Immunol 32: 611–621, 2012.

17. Fourcot A, Couchie D, Chobert MN, Zafrani ES, Mavier P, LapercheY, Brouillet A. Gas6 deficiency prevents liver inflammation, steatohepa-titis, and fibrosis in mice. Am J Physiol Gastrointest Liver Physiol 300:G1043–G1053, 2011.

18. Frankova D, Zidek Z. IFN-gamma-induced TNF-alpha is a prerequisitefor in vitro production of nitric oxide generated in murine peritonealmacrophages by IFN-gamma. Eur J Immunol 28: 838–843, 1998.

19. Friedman SL, Flier JS, Epstein F, Glickman R, Scheele G. Seminars inmedicine of the Beth-Israel-Hospital, Boston-the cellular basis of hepatic-fibrosis-mechanisms and treatment strategies. N Engl J Med 328: 1828–1835, 1993.

20. Frucht DM, Fukao T, Bogdan C, Schindler H, O’Shea JJ, Koyasu S.IFN-gamma production by antigen-presenting cells: mechanisms emerge.Trends Immunol 22: 556–560, 2001.

21. Gabele E, Froh M, Arteel GE, Uesugi T, Hellerbrand C, ScholmerichJ, Brenner DA, Thurman RG, Rippe RA. TNF alpha is required forcholestasis-induced liver fibrosis in the mouse. Biochem Biophys ResCommun 378: 348–353, 2009.

22. Gao B, Jeong WI, Tian Z. Liver: An organ with predominant innateimmunity. Hepatology 47: 729–736, 2008.

23. Gonzalez-Reimers E, Castellano-Higuera A, Aleman-Valls R, Alvarez-Arguelles H, de la Vega-Prieto MJ, Abreu-Gonzalez P, Lopez-Prieto J,Santolaria-Fernandez F, Valladares-Parrilla F. Relation between bodyfat and liver fat accumulation and cytokine pattern in non-alcoholicpatients with chronic HCV infection. Ann Nutr Metab 55: 351–357, 2009.

24. Guo H, Cai CQ, Schroeder RA, Kuo PC. Osteopontin is a negativefeedback regulator of nitric oxide synthesis in murine macrophages. JImmunol 166: 1079–1086, 2001.

25. Horras CJ, Lamb CL, Mitchell KA. Regulation of hepatocyte fate byinterferon-gamma. Cytokine Growth Factor Rev 22: 35–43, 2011.

26. Huang W, Metlakunta A, Dedousis N, Zhang PL, Sipula I, Dube JJ,Scott DK, O’Doherty RM. Depletion of liver Kupffer cells prevents thedevelopment of diet-induced hepatic steatosis and insulin resistance.Diabetes 59: 347–357, 2010.

27. Inagaki Y, Nemoto T, Kushida M, Sheng Y, Higashi K, Ikeda K,Kawada N, Shirasaki F, Takehara K, Sugiyama K, Fujii M, YamauchiH, Nakao A, de Crombrugghe B, Watanabe T, Okazaki I. Interferonalfa down-regulates collagen gene transcription and suppresses experi-mental hepatic fibrosis in mice. Hepatology 38: 890–899, 2003.

28. Inoue A, Yanagisawa M, Takuwa Y, Mitsui Y, Kobayashi M, MasakiT. The human preproendothelin-1 gene. Complete nucleotide sequenceand regulation of expression. J Biol Chem 264: 14954–14959, 1989.

29. Ishida Y, Kondo T, Ohshima T, Fujiwara H, Iwakura Y, Mukaida N.A pivotal involvement of IFN-gamma in the pathogenesis of acetamino-phen-induced acute liver injury. FASEB J 16: 1227–1236, 2002.

30. Ishikawa S, Ikejima K, Yamagata H, Aoyama T, Kon K, Arai K,Takeda K, Watanabe S. CD1d-restricted natural killer T cells contributeto hepatic inflammation and fibrogenesis in mice. J Hepatol 54: 1195–1204, 2011.

31. Jaruga B, Hong F, Kim WH, Gao B. IFN-gamma/STAT1 acts as aproinflammatory signal in T cell-mediated hepatitis via induction ofmultiple chemokines and adhesion molecules: a critical role of IRF-1. AmJ Physiol Gastrointest Liver Physiol 287: G1044–G1052, 2004.

32. Kang YH, Seigel B, Bengsch B, Fleming VM, Billerbeck E, SimmonsR, Walker L, Willberg CB, Barnes EJ, Bhagwanani A, Oo YH, BlumHE, Adams DH, Thimme R, Klenerman P. CD161(�)CD4(�) T cellsare enriched in the liver during chronic hepatitis and associated withco-secretion of IL-22 and IFN-gamma (Abstract). Frontiers Immunol 3:346, 2012.

33. Kanzler S, Lohse AW, Keil A, Henninger J, Dienes HP, SchirmacherP, Rose-John S, Zum Buschenfelde KHM, Blessing M. TGF-beta 1 inliver fibrosis: an inducible transgenic mouse model to study liver fibro-genesis. Am J Physiol Gastrointest Liver Physiol 276: G1059–G1068,1999.

34. Kapsenberg ML. Dendritic-cell control of pathogen-driven T-cell polar-ization. Nat Rev Immunol 3: 984–993, 2003.

35. Kawashima R, Mochida S, Matsui A, YouLuTu ZY, Ishikawa K,Toshima K, Yamanobe F, Inao M, Ikeda H, Ohno A, Nagoshi S, UedeT, Fujiwara K. Expression of osteopontin in Kupffer cells and hepaticmacrophages and Stellate cells in rat liver after carbon tetrachlorideintoxication: a possible factor for macrophage migration into hepaticnecrotic areas. Biochem Biophys Res Commun 256: 527–531, 1999.

36. King TE Jr, Albera C, Bradford WZ, Costabel U, Hormel P, Lan-caster L, Noble PW, Sahn SA, Szwarcberg J, Thomeer M, Valeyre D,du Bois RM. Effect of interferon gamma-1b on survival in patients withidiopathic pulmonary fibrosis (INSPIRE): a multicentre, randomised,placebo-controlled trial. Lancet 374: 222–228, 2009.

37. Knight B, Akhurst B, Matthews VB, Ruddell RG, Ramm GA, Abra-ham LJ, Olvnyk JK, Yeoh GC. Attenuated liver progenitor (oval) celland fibrogenic responses to the choline deficient, ethionine supplementeddiet in the BALB/c inbred strain of mice. J Hepatol 46: 134–141, 2007.

38. Knight B, Lim R, Yeoh GC, Olynyk JK. Interferon-gamma exacerbatesliver damage, the hepatic progenitor cell response and fibrosis in a mousemodel of chronic liver injury. J Hepatol 47: 826–833, 2007.

39. Kremer M, Hines IN, Milton RJ, Wheeler MD. Favored T helper 1response in a mouse model of hepatosteatosis is associated with enhancedT cell-mediated hepatitis. Hepatology 44: 216–227, 2006.

40. Kudo H, Takahara T, Yata Y, Kawai K, Zhang W, Sugiyama T.Lipopolysaccharide triggered TNF-alpha-induced hepatocyte apoptosis ina murine non-alcoholic steatohepatitis model. J Hepatol 51: 168–175,2009.

41. Kudo H, Yata Y, Takahara T, Kawai K, Nakayama Y, Kanayama M,Oya T, Morita S, Sasahara M, Mann DA, Sugiyama T. Telmisartanattenuates progression of steatohepatitis in mice: role of hepatic macro-phage infiltration and effects on adipose tissue. Liver Int 29: 988–996,2009.

42. Kumar V. NKT-cell subsets: Promoters and protectors in inflammatoryliver disease. J Hepatol 59: 618–620, 2013.

43. Langenkamp A, Messi M, Lanzavecchia A, Sallusto F. Kinetics ofdendritic cell activation: impact on priming of TH1, TH2 and nonpolarizedT cells. Nat Immunol 1: 311–316, 2000.

44. Larter CZ, Yeh MM. Animal models of NASH: getting both pathologyand metabolic context right. J Gastroenterol Hepatol 23: 1635–1648,2008.

45. Li X, O’Regan AW, Berman JS. IFN-gamma induction of osteopontinexpression in human monocytoid cells. J Interferon Cytokine Rese 23:259–265, 2003.

46. Li Z, Soloski MJ, Diehl AM. Dietary factors alter hepatic innate immunesystem in mice with nonalcoholic fatty liver disease. Hepatology 42:880–885, 2005.

47. Locatelli I, Sutti S, Vacchiano M, Bozzola C, Albano E. NF-kappaB1deficiency stimulates the progression of non-alcoholic steatohepatitis(NASH) in mice by promoting NKT-cell-mediated responses. Clin Sci124: 279–287, 2013.

48. Ludwig J, Viggiano TR, McGill DB, Ott BJ. Non-alcoholic steatohepa-titis - Mayo-Clinic experiences with a hitherto unnamed disease. MayoClinic Proceedings 55: 434–438, 1980.

49. Luo XY, Takahara T, Hou JG, Kawai K, Sugiyama T, Tsukada K,Takemoto M, Takeuchi M, Zhong L, Li XK. Theaflavin attenuatesischemia-reperfusion injury in a mouse fatty liver model. Biochem Bio-phys Res Commun 417: 287–293, 2012.

50. Maher JJ, Leon P, Ryan JC. Beyond insulin resistance: innate immunityin nonalcoholic steatohepatitis. Hepatology 48: 670–678, 2008.

51. Mallat A, Preaux AM, Blazejewski S, Rosenbaum J, Dhumeaux D,Mavier P. Interferon alfa and gamma inhibit proliferation and collagensynthesis of human Ito cells in culture. Hepatology 21: 1003–1010, 1995.

52. Mas E, Danjoux M, Garcia V, Carpentier S, Segui B, Levade T. IL-6deficiency attenuates murine diet-induced non-alcoholic steatohepatitis.Plos One 4: 2009.

53. Olleros ML, Martin ML, Vesin D, Fotio AL, Santiago-Raber ML,Rubbia-Brandt L, Spahr L, Hadengue A, Garcia I. Fat diet andalcohol-induced steatohepatitis after LPS challenge in mice: role of bio-active TNF and Th1 type cytokines. Cytokine 44: 118–125, 2008.



54. Pockros PJ, Jeffers L, Afdhal N, Goodman ZD, Nelson D, Gish RG,Reddy KR, Reindollar R, Rodriguez-Torres M, Sullivan S, Blatt LM,Faris-Young S. Final results of a double-blind, placebo-controlled trial ofthe antifibrotic efficacy of interferon-gamma1b in chronic hepatitis Cpatients with advanced fibrosis or cirrhosis. Hepatology 45: 569–578,2007.

55. Radaeva S, Sun R, Jaruga B, Nguyen VT, Tian Z, Gao B. Natural killercells ameliorate liver fibrosis by killing activated stellate cells in NKG2D-dependent and tumor necrosis factor-related apoptosis-inducing ligand-dependent manners. Gastroenterology 130: 435–452, 2006.

56. Rockey DC, Chung JJ. Interferon gamma inhibits lipocyte activation andextracellular matrix mRNA expression during experimental liver injury:implications for treatment of hepatic fibrosis. J Investiga Med 42: 660–670, 1994.

57. Rockey DC, Maher JJ, Jarnagin WR, Gabbiani G, Friedman SL.Inhibition of rat hepatic lipocyte activation in culture by interferon-gamma. Hepatology 16: 776–784, 1992.

58. Shen H, Zhang M, Minuk GY, Gong Y. Different effects of ratinterferon alpha, beta and gamma on rat hepatic stellate cell proliferationand activation (Abstract). BMC Cell Biol 3: 9, 2002.

59. Shi Z, Wakil AE, Rockey DC. Strain-specific differences in mousehepatic wound healing are mediated by divergent T helper cytokineresponses. Proc Natl Acad Sci USA 94: 10663–10668, 1997.

60. Stienstra R, Saudale F, Duval C, Keshtkar S, Groener JEM, vanRooijen N, Staels B, Kersten S, Muller M. Kupffer cells promote hepaticsteatosis via interleukin-1 beta-dependent suppression of peroxisome pro-liferator-activated receptor alpha activity. Hepatology 51: 511–522, 2010.

61. Susca M, Grassi A, Zauli D, Volta U, Lenzi M, Marchesini G, BianchiFB, Ballardini G. Liver inflammatory cells, apoptosis, regeneration andstellate cell activation in non-alcoholic steatohepatitis. Dig Liver Dis 33:768–777, 2001.

62. Syn WK, Agboola KM, Swiderska M, Michelotti GA, Liaskou E, PangH, Xie G, Philips G, Chan IS, Karaca GF, Pereira Tde A, Chen Y, MiZ, Kuo PC, Choi SS, Guy CD, Abdelmalek MF, Diehl AM. NKT-associated hedgehog and osteopontin drive fibrogenesis in non-alcoholicfatty liver disease. Gut 61: 1323–1329, 2012.

63. Syn WK, Oo YH, Pereira TA, Karaca GF, Jung Y, Omenetti A, WitekRP, Choi SS, Guy CD, Fearing CM, Teaberry V, Pereira FE, AdamsDH, Diehl AM. Accumulation of natural killer T cells in progressivenonalcoholic fatty liver disease. Hepatology 51: 1998–2007, 2010.

64. Tagawa Y, Sekikawa K, Iwakura Y. Suppression of concanavalin A-induced hepatitis in IFN-gamma(�/�) mice, but not in TNF-alpha(�/�)mice: role for IFN-gamma in activating apoptosis of hepatocytes. J Immunol159: 1418–1428, 1997.

65. Tajiri K, Shimizu Y, Tsuneyama K, Sugiyama T. Role of liver-infiltrating CD3�CD56� natural killer T cells in the pathogenesis ofnonalcoholic fatty liver disease. Eur J Gastroenterol Hepatol 21: 673–680,2009.

66. Takahara T, Furui K, Funaki J, Nakayama Y, Itoh H, Miyabayashi C,Sato H, Seiki M, Ooshima A, Watanabe A. Increased expression ofmatrix metalloproteinase-ii in experimental liver fibrosis in rats. Hepatol-ogy 21: 787–795, 1995.

67. Takahashi F, Takahashi K, Maeda K, Tominaga S, Fukuchi Y.Osteopontin is induced by nitric oxide in RAW 264.7 cells. IUBMB Life49: 217–221, 2000.

68. Takahashi K, Takahashi F, Hirama M, Tanabe KK, Fukuchi Y.Restoration of CD44S in non-small cell lung cancer cells enhanced theirsusceptibility to the macrophage cytotoxicity. Lung Cancer 41: 145–153,2003.

69. Tilg H, Moschen AR. Evolution of inflammation in nonalcoholic fattyliver disease: the multiple parallel hits hypothesis. Hepatology 52: 1836–1846, 2010.

70. Tiniakos DG. Liver biopsy in alcoholic and non-alcoholic steatohepatitispatients. Gastroenterol Clinique et Biol 33: 930–939, 2009.

71. Tomita K, Tamiya G, Ando S, Ohsumi K, Chiyo T, Mizutani A,Kitamura N, Toda K, Kaneko T, Horie Y, Han JY, Kato S, ShimodaM, Oike Y, Tomizawa M, Makino S, Ohkura T, Saito H, Kumagai N,Nagata H, Ishii H, Hibi T. Tumour necrosis factor alpha signallingthrough activation of Kupffer cells plays an essential role in liver fibrosisof non-alcoholic steatohepatitis in mice. Gut 55: 415–424, 2006.

72. Tsuji H, Mukaida N, Harada A, Kaneko S, Matsushita E, NakanumaY, Tsutsui H, Okamura H, Nakanishi K, Tagawa Y, Iwakura Y,Kobayashi K, Matsushima K. Alleviation of lipopolysaccharide-inducedacute liver injury in Propionibacterium acnes-primed IFN-gamma-defi-cient mice by a concomitant reduction of TNF-alpha, IL-12, and IL-18production. J Immunol 162: 1049–1055, 1999.

73. Ulloa L, Doody J, Massague J. Inhibition of transforming growthfactor-beta/SMAD signalling by the interferon-gamma/STAT pathway.Nature 397: 710–713, 1999.

74. Vila-del Sol V, Diaz-Munoz MD, Fresno M. Requirement of tumornecrosis factor alpha and nuclear factor-kappaB in the induction byIFN-gamma of inducible nitric oxide synthase in macrophages. J Leuko-cyte Biol 81: 272–283, 2007.

75. Wallace K, Konstantinou DK, Cowie DE, Tucker SJ, Marek CJ,Durward E, Koruth M, Mann DA, Wright MC. PXR activation reduceshepatic TNF-alpha levels: a mechanism for inhibiting liver fibrosis. JHepatol 48: S189–S189, 2008.

76. Wang Y, Mochida S, Kawashima R, Inao M, Matsui A, YouLuTu ZY,Nagoshi S, Uede T, Fujiwara K. Increased expression of osteopontin inactivated Kupffer cells and hepatic macrophages during macrophagemigration in Propionibacterium acnes-treated rat liver. J Gastroenterol 35:696–701, 2000.

77. Weng H, Mertens PR, Gressner AM, Dooley S. IFN-gamma abrogatesprofibrogenic TGF-beta signaling in liver by targeting expression ofinhibitory and receptor Smads. J Hepatol 46: 295–303, 2007.

78. Weng HL, Wang BE, Jia JD, Wu WF, Xian JZ, Mertens PR, Cai WM,Dooley S. Effect of interferon-gamma on hepatic fibrosis in chronichepatitis B virus infection: a randomized controlled study. Clin Gastro-enterol Hepatol 3: 819–828, 2005.

79. Wu YJ, Cai WM, Li Q, Liu Y, Shen H, Mertens PR, Dooley S, WengHL. Long-term antifibrotic action of interferon-gamma treatment in pa-tients with chronic hepatitis B virus infection. Hepatobiliary PancreaticDis Int 10: 151–157, 2011.

80. Yu J, Chu ES, Wang R, Wang S, Wu CW, Wong VW, Chan HL,Farrell GC, Sung JJ. Heme oxygenase-1 protects against steatohepatitisin both cultured hepatocytes and mice. Gastroenterology 138: 694–704,2010.

81. Zhan YT, An W. Roles of liver innate immune cells in nonalcoholic fattyliver disease. World J Gastroenterol 16: 4652–4660, 2010.



Documents

IFN- deficiency attenuates hepatic inflammation and fibrosis in a steatohepatitis model induced by a methionine- and choline-deficient high-fat diet