Am J Physiol Gastrointest Liver Physiol 309: G42–G51, 2015. First published May 21, 2015; doi:10.1152/ajpgi.00443.2014.

The xanthine oxidase inhibitor febuxostat suppresses development of nonalcoholic steatohepatitis in a rodent model Yusuke Nakatsu,1* Yasuyuki Seno,1* Akifumi Kushiyama,2 Hideyuki Sakoda,3 Midori Fujishiro,4 Aya Katasako,1 Keiichi Mori,1 Yasuka Matsunaga,1 Toshiaki Fukushima,1 Ryuhei Kanaoka,1 Takeshi Yamamotoya,1 Hideaki Kamata,1 and Tomoichiro Asano1 1

Department of Medical Science, Graduate School of Medicine, Hiroshima University, Hiroshima, Japan; 2Division of Diabetes and Metabolism, Institute for Adult Disease, Asahi Life Foundation, Tokyo, Japan; 3Division of Neurology, Respirology, Endocrinology and Metabolism, Department of Internal Medicine, Faculty of Medicine, University of Miyazaki, Miyazaki, Japan; and 4Department of Internal Medicine, Graduate School of Medicine, University of Tokyo, Tokyo, Japan

Submitted 15 December 2014; accepted in final form 30 April 2015

Nakatsu Y, Seno Y, Kushiyama A, Sakoda H, Fujishiro M, Katasako A, Mori K, Matsunaga Y, Fukushima T, Kanaoka R, Yamamotoya T, Kamata H, Asano T. The xanthine oxidase inhibitor febuxostat suppresses development of nonalcoholic steatohepatitis in a rodent model. Am J Physiol Gastrointest Liver Physiol 309: G42–G51, 2015. First published May 21, 2015; doi:10.1152/ajpgi.00443.2014.— Xanthine oxidase (XO) is an enzyme involved in the production of uric acid (UA) from purine nucleotides. Numerous recent studies have revealed the likelihood of metabolic syndrome including nonalcoholic fatty liver disease (NAFLD) or steatohepatitis (NASH) to be related to hyperuricemia. However, it remains unclear whether elevated serum UA during the development of NAFLD or NASH is a cause or a consequence of these diseases. In this study, the XO inhibitor febuxostat was administered to two types of NASH model mice. Febuxostat exerted a strong protective effect against NASH development induced by a high-fat diet containing trans fatty acid (HFDT). In contrast, methionine choline-deficient-diet-induced NASH development not accompanied by hyperuricemia showed no UA normalization, suggesting that the ameliorating effect of febuxostat occurs via the normalization of hyperuricemia itself and/or accompanying molecular mechanism(s) such as oxidative stress. In the HFDT-fed mice, hyperuricemia, elevated alanine aminotransferase, and increased Tunnelpositive cells in the liver were normalized by febuxostat administration. In addition, upregulation of fatty acid oxidation-related genes, fibrotic change, and increases in collagen deposition, inflammatory cytokine expressions, and lipid peroxidation in the HFDT-fed mice were also normalized by febuxostat administration. Taken together, these observations indicate that administration of febuxostat has a protective effect against HFDT-induced NASH development, suggesting the importance of XO in its pathogenesis. Thus XO inhibitors are potentially potent therapies for patients with NASH, particularly that associated with hyperuricemia. uric acid; nonalcoholic steatohepatitis; xanthine oxidase inhibitor; hyperuricemia; febuxostat

of uric acid (UA), the end product of purine nucleotide degradation, are reportedly associated with the development of cardiovascular and kidney diseases, hypertension, and obesity (5, 9, 17, 18, 26). Although the causative link between hyperuricemia and disease is clearest for gout, many clinical studies have revealed that the prevalence of metabolic syndrome is closely related to graded

ELEVATED SERUM CONCENTRATIONS

* Y. Nakatsu and Y. Seno contributed equally to this work. Address for reprint requests and other correspondence: T. Asano, Dept. of Medical Science, Graduate School of Medicine, Hiroshima Univ., 1-2-3 Kasumi, Minami-ku, Hiroshima City, Hiroshima, Japan (e-mail: [email protected]). G42

increases in hyperuricemia, strongly suggesting serum UA levels to be a predictor of metabolic syndrome, including impaired glucose and lipid metabolism (2, 8, 11, 23, 24). Although the endogenous production of UA is mainly from the liver, intestines, and other tissues like muscles, kidneys, and the vascular endothelium, UA is also secreted from adipose tissue, which is reportedly increased in obese subjects (24), supporting the view of UA elevation as a consequence of obesity. On the other hand, UA reportedly reduces nitric oxide (NO) bioavailability and thereby impairs endothelial function (10). Furthermore, several reports have demonstrated favorable effects of xanthine oxidase (XO) inhibitors on the disorders characteristic of metabolic syndrome. For example, administration of allopurinol or febuxostat reportedly prevented fructose-induced hyperinsulinemia, hypertension, and weight gain in rats (19), diabetic renal injury in streptozotocin-treated rats (14), and atherosclerosis development in apolipoprotein E knockout mice (12). The relationships of elevated serum UA with nonalcoholic fatty liver disease (NAFLD) and steatohepatitis (NASH) have also been demonstrated in several clinical studies (21, 22), but the details of these relationships remain unclear. Serum UA elevation and NASH development may not have a causal relationship, with both instead being consequences of metabolic syndrome. On the other hand, it is also possible that hyperuricemia itself and/or a molecular mechanism(s) associated with it, such as oxidative stress, plays important roles in promoting and/or exacerbating the development of NAFLD, independently of obesity or insulin resistance. If this is the case, therapy aimed at normalization of the serum UA level could be useful for treating NASH, even if the degree of obesity or metabolic factors other than UA show no improvement. To address this question, we herein examined the effects of an XO inhibitor, febuxostat, on two types of NASH rodent models, with and without elevated serum UA concentrations. We obtained evidence that treatment with this XO inhibitor exerted a strong protective effect against the development of NASH with hyperuricemia induced by a high-fat diet containing trans fatty acid (HFDT), but not on that caused by a methionine choline-deficient (MCD) diet without hyperuricemia. These results suggest the importance of XO in the pathogenesis of NASH development and also raise the possibility of XO inhibitors having therapeutic potential for patients with NASH accompanied by hyperuricemia.

0193-1857/15 Copyright © 2015 the American Physiological Society

http://www.ajpgi.org

G43

XO INHIBITOR FEBUXOSTAT SUPPRESSES NASH DEVELOPMENT MATERIALS AND METHODS

Animals, diets, and febuxostat treatment. C57BL/6 mice were purchased from SLC (Hamamatsu, Japan). They were housed in temperature- and light-controlled rooms with free access to food and water. After a 1-wk acclimatization period, the mice were fed normal chow diet (ND) (Oriental Yeast, Tokyo, Japan) or a diet comprised of high fat (40% kcal), high fructose (22% by weight), and high cholesterol (2% by weight), with trans fats serving as the fat source (Primex partially hydrogenated vegetable oil shortening, cat. no. D09100301; Research Diet, New Brunswick, NJ). The HFDT has been deemed to be superior at inducing NASH. Febuxostat (1 mg/kg per day) (Teijin, Japan) was given to the mice by mixing it into their drinking water. After 8 wk, the mice were killed, and their livers and blood were collected. In the other experiment, the mice were randomly divided into two experimental groups and fed the ND (Oriental Yeast) or the MCD diet (Oriental Yeast), with or without febuxostat administration (1 mg/kg per day), for 8 wk. All animals were handled in accordance with the Guidelines for the Care and Use of Experimental Animals published by Hiroshima University, and all protocols were approved by Hiroshima University. Measurement of serum parameters. Serum was collected after a 10-min centrifugation at 8,000 revolution/min. Alanine transaminase (ALT) and UA were assayed with a Transaminase C-II Test Wako Kit (Wako, Osaka, Japan) and a UA assay kit (Cayman Chemical, Ann Arbor, MI), respectively, according to the manufacturers’ instructions. Homeostasis model of assessment of insulin resistance (HOMA-IR) was calculated as (fasting serum glucose ⫻ fasting serum insulin concentration)/405. Measurement of XO activity. Hepatic XO activity was measured using an XO activity assay kit, according to the manufacturer’s instructions (Sigma, St. Louis, MO). Fluorescence-activated cell sorting analysis. Livers were homogenized immediately after death of the mice. Liver lysates were washed with RPMI medium four times, and debris was removed through a cell strainer. The cells were stained with brilliant violet-CD3, allophycocyanin-CD45, phenylephrine-F4/80, and 7-aminoactinomycin D (7AAD) for 30 min and then analyzed with Fortessa (BD Biosciences, San Jose, CA). Macrophages were gated by side scatter and forward scatter (SSC/FSC), 7-AAD(-), and CD45⫹/F4/80⫹, and T cells were gated by SSC/FSC, 7-AAD(-), and CD45⫹/CD3⫹. Cell populations are presented as the ratios to total recorded cells. Hydroxyproline measurement in mouse livers. Hydroxyproline contents in the livers of mice were measured using a hydroxyproline colorimetric assay kit (BioVision, Mountain View, CA). Briefly, tissue samples were hydrolyzed with HCl at 120°C. Samples were reacted with perchloric acid and dimethylaminobenzaldehyde to produce a chromogen with an absorbance maximum at 560 nm. Western blotting. Livers were homogenized in lysis buffer containing 50 mM Tri·HCl (pH 7.4), 150 mM NaCl, 1 mM EDTA, 1% Triton X-100, 1 mM PMSF, 1 mM orthovanadate, and protease inhibitor cocktail. The lysates were then centrifuged at 15,000 revolution/min for 30 min. After quantitative protein determination, supernatants were mixed with the same volume of 2⫻ sample buffer. Equal amounts of proteins extracted from livers were electrophoresed on SDS-PAGE gel and then transferred to PVDF membranes. After being blocked with 3% bovine serum albumin in phosphate-buffered saline with Tween (PBST) for 1 h, the membranes were reacted with the anti-XO (1:1,000), anti-collagen-1 (1:2,000), or anti-tubulin (1:2,000) antibody for 1 h at room temperature. The membranes were subsequently reacted with horseradish peroxidase-conjugated secondary antibodies for 1 h. The membranes were then washed with PBST three times and reacted with Supersignal West Pico Substrate (Thermo Scientific, Yokohama, Japan) and exposed to X-ray film. Band density was quantified by ChemiDoc XRS⫹ (Bio-Rad, Tokyo, Japan). Quantitative real-time RT-PCR. Total RNA was extracted from mouse livers using Sepasol reagent (Nakalai Tesche, Kyoto, Japan),

according to the manufacturer’s instructions. First-strand cDNAs were synthesized using a verso cDNA synthesis kit with both a random primer and reverse transcriptase enhancer, which degrades doublestranded DNA. Quantitative real-time PCR was performed using SYBR Green PCR master mix (Invitrogen, Tokyo, Japan) on a CFX96 real-time PCR system (Bio-Rad). Relative mRNA gene levels were normalized to the GAPDH mRNA level, and relative expressions were determined by the comparative Ct method. The designed primers are shown in Table 1. Histochemical studies. Livers were fixed with 10% formaldehyde, then dehydrated through an ethanol and xylene series, and finally embedded in paraffin. Paraffin-embedded sections were deparaffinized in xylene and rehydrated. Sections were incubated in 0.1% Triton for 5 min and then heated in citrate buffer with a microwave or treated with protein kinase K (Pro K) for antigen activation. After being washed with PBS three times, sections were incubated with first antibodies [anti-4-hydroxynonenal (4-HNE), 1:100 (Japan Institute for the Control of Aging, Fukuroi, Japan); T anti-smooth muscle actin (SMA), 1:100 (Abcam, Cambridge, MA); anti-active caspase-3, 1:100 (Cell Signaling, Beverly, MA)] overnight at 4°C. For active caspase-3 staining, the slides were reacted with fluorescein isothiocyanateconjugated secondary antibody (Santa Cruz Biotechnology, Santa Cruz, CA) for 1 h. To visualize SMA and 4-HNE, the slides were stained by the diaminobenzidine method. Digital images of lesions were obtained with a multifocal microscope (BZ-9000; KEYENCE, Osaka, Japan). TUNEL staining was performed, using the DeadEnd Fluorometric TUNEL System (Promega, Madison, WI). Briefly, deparaffinized

Table 1. Designed primers Gene Name

ACC1 FASN PPAR␣ CYP4A10 CYP4A14 ACOX Collagen1a1 Collagen1a2 SMA TIMP1 TIMP2 MCP-1 IL-1␤ IL-6 IL-12 GAPDH

Primer Sequence

Gene ID

Amplicon Size

F: gag agg ggt caa gtc ctt cc R: ctg ctg ccg tca taa gac aa F: ttg ctg gca cta cag aat gc R: aac agc ctc aga gcg aca at F: atg cca gta ctg ccg ttt tc R: ggc ctt gac ctt gtt cat gt F: cca gca gtt ccc atc acct c R: tga tcg ccc cag aat cac tt F: ggt tct tcc aat ggg cct tc R: tct gct gga gct cct tgt cc F: ttc aag aca gag ccg tgc aa R: cag agc caa ggg tca cat cc F: gag cg gaga gta ctg gat cg R: gct tct ttt cct tgg ggt tc F: ccg tgc ttc tca gaa cat ca R: gag cag cca tcg act agg ac F: acc aac tgg gac gac atg gaa R: tgt cag cag tgt cgg atg ctc F: att caa ggc tgt ggg aaa tg R: ctc aga gta cgc cag gga ac F: gga tga gtg cct ctg gat gg R: acg ggt cct cga tgt caa ga, F: agg tcc ctg tca tgc ttc tg R: tct gga ccc att cct tct tg F: tgg gcc tca aag gaa aga at R: ctt ggg atc cac act ctc ca F: cca tcc agt tgc ctt ctt gg R: tcc acg att tcc cag aga aca F: tgg agc act ccc cat tcc ta R: tga gct tgc acg cag aca tt F: tga tgg gtg tga acc acg ag R: ggg cca tcc aca gtc ttc tg

107476

200 bp

14104

192 bp

19013

220 bp

13117

190 bp

13119

165 bp

11430

222 bp

12842

158 bp

12843

168 bp

11475

90 bp

21857

183 bp

21858

155 bp

20296

249 bp

16176

86 bp

16193

165 bp

16160

170 bp

14433

178 bp

F, forward; R, reverse. ACC1, acetyl-CoA carboxylase 1; FASN, fatty acid synthase; PPAR, peroxisome proliferator-activated receptor; ACOX, acyl coenzyme A oxidase; SMA, smooth muscle actin; TIMP, tissue inhibitor of metalloproteinase; MCP-1, monocyte chemoattractant protein 1.

AJP-Gastrointest Liver Physiol • doi:10.1152/ajpgi.00443.2014 • www.ajpgi.org

G44

XO INHIBITOR FEBUXOSTAT SUPPRESSES NASH DEVELOPMENT

sections were treated with Pro K. After being washed, the sections were reacted with terminal deoxynucleotidyl transferase and fluorescein-12-dUTP. After being washed, the slides were mounted employing DAPI. Thiobarbituric acid reactive substances assay. Thiobarbituric acid reactive substances (TBARS) measurement was performed using a TBARS assay kit (Cayman Chemical), according to the manufacturer’s instructions. Briefly, malondialdehyde (MDA) in lysates was reacted with thiobarbituric acid at 100°C to produce the MDA-TBA adduct, and the concentration produced was measured at 535 nm. Statistical analysis. Results are expressed as means ⫾ SE. Statistical significance was assessed using ANOVA followed by Tukey’s honestly significant difference test. Variables not normally distributed were log transformed before the analysis. P ⬍ 0.05 was taken to indicate a statistically significant difference.

duced liver weight without affecting whole-body weights. Whereas febuxostat administration did not alter blood glucose concentrations or HOMA-IR scores (Fig. 1B), serum UA concentrations and hepatic XO activity raised by HFDT feeding were completely normalized by febuxostat (Fig. 1, C and D), with no change in the amount of hepatic XO protein (Fig. 1E). Serum ALT levels were markedly elevated by HFDT feeding for 8 wk, but the degree of these increases was significantly suppressed by febuxostat administration (Fig. 1F). Hepatosteatosis induced by HFDT was reversed by febuxostat administration. Hematoxylin-eosin staining showed that the livers of mice fed HFDT developed marked macrovesicular steatosis (Fig. 2A). In contrast, the mice fed HFDT with febuxostat showed much less hepatic fat accumulation than HFDT-fed mice not given febuxostat (Fig. 2, A and B). To reveal the molecular mechanism underlying febuxostat-induced suppression of hepatosteatosis, the mRNA levels of the genes regulating lipid metabolism were investigated. Febuxostat had no significant effect on mRNA levels of acetyl-CoA carboxylase and fatty acid synthase, which are both involved in lipid synthesis (Fig. 2C). Neither the peroxisome proliferator-activated receptor-␣ nor the CYP4A10 mRNA

RESULTS

Febuxostat reduced the elevated serum UA concentration and increased liver weight induced by HFDT feeding. Mice were divided into four groups based on the combinations of HFDT and ND with or without febuxostat in the drinking water. After the 8-wk treatment, body and liver weights were greater in the HFDT-fed mice than in the ND-fed mice (Fig. 1A). Febuxostat administration slightly but significantly re-

15 10 5 0 ND

80 60 40 20

0

*

2000 1500 1000 500 0 HFDT

0

100

50

ND

HFD

HFDT

*

45 40

HFDT

35

n.s.

1.4

***

ND

HFDT

F ND

0 ND

1

ND

1.6 150

Feb

2 20

Cont Feb Cont Feb

200

Cont 4

40

HFDT

***

5

3

0 ND

250 XO activity (µmol/min/mg protein)

2500

60

IB:XO

***

3000

6

80

E

**

3500

Uric acid ( M)

100

Relative XO protein levels

4000

7 100

120

HFDT

8

120

140

ND

*** n.s.

9

n.s.

n.s.

160

D **

140

180

HFDT

C

200

30

1.2

ALT(IU/ml)

20

B

Fed Glucose (mg/dl)

25

Liver Weight (g)

Body Weight (g)

30

*

2 1.8 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0

HOMA-IR

n.s.

Fast Glucose (mg/dl)

A 35

1 0.8 0.6

25 20 15 10

0.4

5

0.2

0

0 ND

HFDT

ND

HFDT

Fig. 1. Febuxostat reduced serum uric acid (UA) elevations and liver weight increases induced by high-fat diet containing trans fatty acid (HFDT) feeding. Mice were divided into 4 groups based on the combination of HFDT and normal chow diet (ND) with or without febuxostat in drinking water (n ⫽ 8 per group). After the 8-wk treatment, the indicated parameters were measured. A: whole-body and liver weights. B: blood glucose concentrations in the fed and fasted states and homeostasis model of assessment of insulin resistance (HOMA-IR). C: serum UA concentration. D: xanthine oxidase (XO) activity in livers. E: immunoblotting (IB) of hepatic XO. Cont, control; Feb, febuxostat. F: serum alanine transaminase (ALT) level. All data are shown as means ⫹ SE. *P ⬍ 0.05, **P ⬍ 0.01, ***P ⬍ 0.001; n.s., not significant. AJP-Gastrointest Liver Physiol • doi:10.1152/ajpgi.00443.2014 • www.ajpgi.org

G45

XO INHIBITOR FEBUXOSTAT SUPPRESSES NASH DEVELOPMENT

HFDT

ND Control water

2

C

*

50 40 Cont

30

Feb

20

2

10

1.5

n.s. 1 0.5

1.2 1 0.8 0.6 0.4

0.2

E

1

0.5

HFDT

ND

ND

HFDT

HFDT

2

**

1.5

**

1 0.5

1.5 Cont

1

Feb 0.5 0

0

0

0

HFDT

2 CYP4A14 mRNA levels

1.4

n.s.

1.5 CYP4A10 mRNA levels

PPARalpha mRNA levels

HFDT

n.s.

ND

0.5 0

ND ND

n.s.

1

0

0

D 1.6

1.5

ACOX mRNA levels

Triglyceride (mg/g tissue)

60

Febuxostat

Control water

ACC mRNA levels

B

Febuxostat

FASN mRNA levels

A

ND

HFDT

ND

HFDT

Fig. 2. Febuxostat normalized HFDT-induced hepatosteatosis. Mice were divided into 4 groups based on the combination of HFDT and ND for 8 wk with or without febuxostat in drinking water. A: paraffin-embedded sections were stained with hematoxylin and eosin. Representative photographs of each group are shown. B: mRNA levels of triglycerides. C: mRNA levels of acetyl-CoA carboxylase (ACC) and fatty acid synthase (FASN). D: mRNA levels of peroxisome proliferator-activated receptor-␣ (PPAR␣) and CYP4A10. E: mRNA levels of CYP4A14 and acyl coenzyme A oxidase (ACOX). All data are shown as means ⫹ SE. *P ⬍ 0.05, **P ⬍ 0.01.

level was affected by the difference in diets or by febuxostat administration (Fig. 2D). On the other hand, although the expressions of fatty acid oxidation-related genes such as acyl coenzyme A oxidase (ACOX) and CYP4A14 were reduced by HFDT feeding, febuxostat normalized these reductions (Fig. 2E). Febuxostat inhibits liver fibrosis attributable to HFDT feeding. To elucidate the impact of febuxostat on liver fibrosis caused by HFDT feeding, collagen deposition in the liver was estimated by Sirius red staining. The mice fed HFDT showed marked liver fibrosis, whereas almost no collagen deposition was observed in the livers of HFDT-fed mice administered febuxostat (Fig. 3, A and B). Hydroxyproline accumulation in the liver was also caused by HFDT feeding, and febuxostat suppressed this effect (Fig. 3C). Consistent with the results of Azan staining, whereas a marked increase was observed in HFDT-fed mice, mRNA expressions of collagen-1a1/1a2 in the febuxostat-treated mice were not increased by HFDT feeding and were significantly lower than those in the untreated mice fed HFDT

(Fig. 3D). In addition, hepatic collagen proteins were upregulated by HFDT, and febuxostat administration reduced the resulting accumulation of collagen (Fig. 3E). Because activation of hepatic stellate cells reportedly plays a critical role in the development of hepatic fibrosis, immunofluorescence staining was carried out using anti-SMA antibody, as a marker of stellate cell activation. Numerous SMApositive cells were observed in the livers of HFDT-fed mice, and this finding was normalized by febuxostat administration (Fig. 4A). These immunochemical staining data were supported by the analysis of mRNA expressions of SMA and tissue inhibitors of metalloproteinases (TIMPs) (Fig. 4B). Febuxostat attenuates oxidative stress and inflammatory cytokine releases. Oxidative stress and inflammatory cytokines are reportedly involved in the activation of hepatic stellate cells. Thus the accumulation of 4-HNE as a marker of oxidative stress was examined (Fig. 5A). Although the amount of 4-HNE was markedly increased in the livers of HFDT-fed mice, febuxostat administration was found to suppress the accumulation of 4-HNE (Fig. 5A). In addition, the amount of

AJP-Gastrointest Liver Physiol • doi:10.1152/ajpgi.00443.2014 • www.ajpgi.org

G46

XO INHIBITOR FEBUXOSTAT SUPPRESSES NASH DEVELOPMENT

A

HFDT

ND

0.6

C

** **

Hydroxy proline (µg/g tissue)

0.5 % of Area

0.4 0.3

Cont

0.2

Feb

0.1 0

*

100 80 60 40 20 0

ND

E

ND

HFDT

HFD

Febuxostat

Control water

D Col 1 mRNA levels

B

Febuxostat

**

3.5 3 2.5 2 1.5 1 0.5 0

n.s. Col 1a2 mRNA levels

Control water

ND

n.s.

1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0

HFDT

**

Cont Feb

ND

HFDT

IB:Collagen1 Cont Feb Cont Feb ND

Relative Protein Levels

1.5

HFDT

***

**

1 Cont 0.5

Feb

0 ND

HFDT

Fig. 3. Febuxostat inhibits liver fibrosis attributable to HFDT feeding. A: sections prepared from the livers of ND- or HFDT-fed mice with or without febuxostat administration were stained with Sirius red. Representative photographs of each group are shown. B: areas staining positive with Sirius red were randomly selected from ⫻60 fields for measurement, and the results are presented as a bar graph. C: hydroxyproline contents of livers were measured. D: mRNA levels of collagen 1 (Col 1) and collagen 2. E: collagen expression levels were measured. All data are shown as means ⫹ SE. *P ⬍ 0.05, **P ⬍ 0.01, ***P ⬍ 0.001.

MDA, a product of lipid peroxidation, was also elevated by HFDT feeding but was completely normalized by febuxostat administration (Fig. 5B). We also investigated the expressions of cytokines such as monocyte chemoattractant protein 1 (MCP-1), IL-1␤, IL-6, and IL-12. In our NASH model, MCP-1 expressions were markedly elevated by HFDT, but the expressions of IL-1␤, IL-6, and IL-12 were unchanged or only slightly elevated. In contrast, febuxostat administration markedly reduced the expressions of MCP-1 and IL-6, while only tending to reduce the expressions of IL-1␤ and IL-12 without statistical significance (Fig. 5C). The populations of immune cells were not affected by either HFDT or febuxostat (Fig. 5D). Febuxostat reduced cell apoptosis induced by HFDT feeding. Finally, the effect of febuxostat on the apoptosis of hepatocytes was investigated. HFDT feeding markedly increased the number of cells positive for TUNEL staining, and this was almost completely reversed by febuxostat administration (Fig. 6, A and B). In addition, active caspase-3-positive cells were significantly increased in the livers of HFDT-fed

mice, a finding that was normalized by the administration of febuxostat (Fig. 6, C and D). Febuxostat had no effect on the NASH development induced by MCD diet feeding. Although mice fed the MCD diet had slightly but not significantly elevated serum UA concentrations compared with those fed the ND, febuxostat administration did not affect the serum UA concentration in the MCD diet-fed mice (Fig. 7A). In addition, febuxostat did not normalize the serum ALT elevation caused by MCD diet feeding (Fig. 7B). Hematoxylin-eosin staining revealed marked increases in fat droplets, increased inflammatory cell infiltration, and balloonlike structures in the livers of the MCD diet-fed mice, and febuxostat intervention did not improve these abnormalities (Fig. 7C). In addition, the results of Sirius red staining indicated that febuxostat had no effect on collagen deposition induced by the MCD diet (Fig. 7C). Similarly, febuxostat administration showed no significant normalizing effects on the elevations of collagen-1a1, collagen-1a2, TIMP1, and TIMP2 mRNA caused by MCD diet feeding (Fig. 7D). Taken together, these results allow us to conclude that febuxostat does

AJP-Gastrointest Liver Physiol • doi:10.1152/ajpgi.00443.2014 • www.ajpgi.org

G47

XO INHIBITOR FEBUXOSTAT SUPPRESSES NASH DEVELOPMENT

A

HFDT

ND Control water

B

Febuxostat

*

3

1.4

6

Cont

1.5

Feb

1 0.5 0

5

TIMP2 mRNA levels

2

TIMP1 mRNA levels

SMA mRNA levels

* 2.5

4 3 2 1

HFDT

*

1.2 1 0.8 0.6 0.4 0.2 0

0 ND

Febuxostat

Control water

ND

HFDT

ND

HFDT

Fig. 4. Febuxostat inhibits HFDT-induced activation of hepatic stellate cells. A: sections prepared from the livers of ND- or HFDT-fed mice with or without febuxostat administration were stained with anti-smooth muscle actin (SMA) antibody, as a marker of stellate cell activation. Representative photographs of each group are shown. B: mRNA levels of SMA, tissue inhibitor of metalloproteinase 1 (TIMP1), and TIMP2. All data are shown as means ⫹ SE. *P ⬍ 0.05.

not improve MCD diet-induced NASH that is not accompanied by hyperuricemia. DISCUSSION

NASH is highly related to metabolic syndrome and obesity, and its incidence has been increasing in many, particularly advanced, nations. However, there are currently no treatment options specifically for NASH, and this disease is usually managed with lifestyle changes such as diet, exercise, and weight reduction. Although several recent lines of evidence have demonstrated a relationship between hyperuricemia and NASH (21, 22), the present study showed clearly that febuxostat exerts a strong protective effect against NASH development induced by HFDT feeding, based on our observations of the inductions of fibrosis genes, steatohepatitis, and inflammatory cytokine expressions. On the other hand, interestingly, febuxostat had no protective effect against the development of MCD diet-induced NASH that is not accompanied by hyperuricemia. Regarding the pathogenesis of NASH, the “two-hit theory” has been proposed as a mechanism underlying the development of NASH (15). The first hit involves simple steatosis, whereas the second hit, factors such as oxidative stress and inflammatory cytokines, exacerbates NASH once it has developed. These two hits are both present in the NASH rodent models created by HFDT and MCD diet feeding, with phenotypic features such as obesity, insulin resistance, glucose intolerance, and hyperuricemia all being present in the HFDT- but not the MCD diet-induced NASH models. Whereas febuxostat did not affect body weight gain, insulin resistance, or glucose intolerance induced by HFDT feeding, serum UA elevation, observed in the HFDT- but not the MCD diet-fed mice, was completely

normalized. With the aforementioned issues in consideration, it may be reasonable to speculate that febuxostat protects against HFDT-induced NASH development via normalization of the elevated UA concentration itself and/or accompanying molecular mechanism(s). Recent reports have shown increased XO activity in HFDT-fed mice and genetically obese mice (24, 25). In good agreement with these reports, XO activity in the livers of HFDT-fed mice was shown to be increased (Fig. 1D). Unexpectedly, XO activity was revealed to also be upregulated in the MCD diet-fed mice (MCD, 197 ⫾ 15 vs. ND, 115 ⫾ 7 ␮mol/mg protein per min), despite there being no serum UA elevation. However, with consideration that MCD-fed mice are malnourished with atrophic adipose tissue, UA production might be suppressed by a deficiency of source materials in the liver and adipocytes, the two major tissues producing UA. On the other hand, HFDT-fed mice with excessive nutrition and hypertrophic adipocytes would presumably release more UA than the ND-fed mice. Although UA is reportedly capable of acting as a prooxidant like other reducing substances such as ascorbic acid, the reaction catalyzed by XO converting xanthine to UA requires oxygen (1). XO is abundantly expressed in the liver and generates hydroxyperoxide when catalyzing substrates. Thus serum UA elevation is considered to accompany increased accumulation of oxidative stress generated by hepatic XO. In fact, 4-HNE or MDA produced by lipid peroxidation was found to be higher in mice given HFDT, but this elevation was normalized by febuxostat administration. In contrast, the expression levels of apoptosis-related genes such as FADD and Bim mRNAs did not significantly differ between the HFDT and ND groups (data not shown). It is considered that oxidative stress affects the membrane potential of mitochondria and

AJP-Gastrointest Liver Physiol • doi:10.1152/ajpgi.00443.2014 • www.ajpgi.org

G48

XO INHIBITOR FEBUXOSTAT SUPPRESSES NASH DEVELOPMENT

A

ND

HFDT

1.2 1 0.8

Cont

0.6

Feb

0.4 0.2

2 1.5 1 0.5

C

IL-1 mRNA Levels

6

3 2 1 0

1.6

1.2

1.4

HFDT

Feb

1

ND

HFDT

**

0.8 0.6 0.4 0.2

n.s. 1.8 1.6

1.2

1

0

ND

Cont 2

HFDT

1.4

IL-6 mRNA levels

7

4

3

n.s.

*

5

4

0 ND

HFDT

n.s.

5

n.s.

0

0 ND

MCP-1 mRNA levels

2.5

Febuxostat

IL-12 mRNA levels

MDA (nmol/mg protein)

D

*

1.4

Control Water

CD45+CD3+ % of recorded total cells

B

Febuxostat

CD45+ F4/80+ % of recorded total cells

Control Water

1 0.8 0.6 0.4 0.2

HFDT

1.2 1 Cont

0.8

Feb

0.6 0.4 0.2

0 ND

1.4

0 ND

HFDT

ND

HFDT

Fig. 5. Febuxostat attenuates oxidative stress and releases of inflammatory cytokines induced by HFDT feeding. A: sections prepared from the livers of ND- or HFDT-fed mice with or without febuxostat administration were stained with anti-4-hydroxynonenal (4-HNE) antibody. Representative photographs of each group are shown. B: amounts of malondialdehyde (MDA) in the liver tissues were measured by the thiobarbituric acid reactive substances method. C: mRNA levels of monocyte chemoattractant protein 1 (MCP-1), IL-1b, IL-6, and IL-12. D: fluorescence-activated cell sorting analysis was performed to investigate the populations of immune cells. Left: T-cells were gated by forward scatter/side scatter (FSC/SSC), 7-aminoactinomycin D (7-AAD)(-), CD45⫹/CD3⫹. Right: macrophages were gated by FSC/SSC, 7-AAD(-), CD45⫹/F4/80⫹. All data are shown as means ⫹ SE. *P ⬍ 0.05, **P ⬍ 0.01.

triggers the release of cytochrome c into the cytosol, inducing apoptosis (6). Therefore, we speculate the involvement of oxidative stress on the induction of hepatocyte apoptosis probably via the impaired mitochondrial function. Furthermore, accumulated oxidative stress reportedly leads to increased expressions of inflammatory cytokines, such as MCP-1 and IL-6, findings also observed in mice given HFDT and normalized by febuxostat. These findings may represent phenomena similar to those reported in the previous study showing the suppressive effect of febuxostat on lipopolysaccharide (LPS)induced MCP-1 release from macrophages (20), but further study is necessary to clarify this issue. The accumulation of oxidative stress and inflammatory cytokine expressions are also likely to be involved in, not only the inflammation and fibrosis occurring as a result of the “second hit,” but also in lipid accumulation because continuous injection of LPS into rats reportedly results in fatty liver development (7). This study revealed that expressions of fatty acid oxidation-related genes such as ACOX and CYP4A14 are

reduced by HFDT feeding and that these abnormalities were normalized by febuxostat administration. Because MCD diet feeding decreases the expressions of catalase and superoxide dismutase, oxidative stress also accumulates in the livers of mice given the MCD diet (4, 13, 16) although its generation mechanism is not related to either the XO activity or the UA concentration. Thus it is reasonable to speculate that this difference in the process of generating oxidative stress explains why febuxostat exerts no effect on MCD diet-induced NASH development. Finally, it should be noted that, besides the mechanism involving increased oxidative stress, an elevated UA concentration may itself also contribute to hepatic lipid accumulation because a previous report showed UA to promote triglyceride accumulation in hepatocytes by inhibiting AMP kinase (3). In conclusion, to our knowledge, this is the first study clearly demonstrating that an XO inhibitor exerts a strong protective effect against NASH development, independently of the normalization of impaired glucose metabolism or

AJP-Gastrointest Liver Physiol • doi:10.1152/ajpgi.00443.2014 • www.ajpgi.org

G49

XO INHIBITOR FEBUXOSTAT SUPPRESSES NASH DEVELOPMENT

Control

A

Febuxostat

B 2.5

Cont

Number of TUNEL positive cells/area

Feb

ND

HFDT

**

2

1.5

1

0.5

0 ND

Control

Febuxostat

D 1.8 Number of active Caspase -3 positive cells /area

C

ND

HFDT

HFDT

1.6

Cont

**

Feb

1.4 1.2 1

0.8 0.6 0.4 0.2 0 ND

HFDT

Fig. 6. Febuxostat reduced cellular apoptosis induced by HFDT feeding. Sections prepared from the livers of ND- or HFDT-fed mice with or without febuxostat administration were subjected to TUNEL staining (A and B) or active caspase-3 staining (C and D). TUNEL-positive cells or active caspase-3 cells were counted in 3 randomly selected areas for each mouse (magnification fields ⫻600). **P ⬍ 0.01.

obesity. Because the results obtained for the HFDT-fed NASH model are more clinically relevant than those of the MCD diet-fed model, our results raise the possibility of an XO inhibitor such as febuxostat as effective for the treatment of at least some human patients with NASH, perhaps those with borderline or high serum UA concentrations. Clinical trials will be required to test the therapeutic potential of XO inhibitors. DISCLOSURES No conflicts of interest, financial or otherwise, are declared by the authors. AUTHOR CONTRIBUTIONS Author contributions: Y.N., A. Kushiyama, T.F., H.K., and T.A. conception and design of research; Y.N., Y.S., A. Katasako, K.M., and Y.M. performed

experiments; Y.N., Y.S., T.F., R.K., and T.A. analyzed data; Y.N., A. Kushiyama, H.S., M.F., Y.M., T.F., R.K., T.Y., H.K., and T.A. interpreted results of experiments; Y.N. and Y.S. prepared figures; Y.N. and T.A. drafted manuscript. REFERENCES 1. Batteli MG, Boloqnesi A, Polito L. Pathophysiology of circulating xanthine oxidoreductase: new emerging roles for a multi-tasking enzyme. Biochem Biophys Acta 1842: 1502–1517, 2014. 2. Bokyo EJ, de Courten M, Zimmet PZ, Chitson P, Tuomilehto J, Alberti KG. Features of the metabolic syndrome predict higher risk of diabetes and impaired glucose tolerance: a prospective study in Mauritius. Diabetes Care 23: 1242–1248, 2000. 3. Cicerchi C, Li N, Kratzer J, Garcia G, Roncal-Jimenez CA, Tanabe K, Hunter B, Rivard CJ, Sautin YY, Gaucher EA, Johnson RJ, Lanaspa MA. Uric acid-dependent inhibition of AMP kinase induces hepatic glucose production in diabetes and starvation: evolutionary implications of the uricase loss in hominids. FASEB J 28: 34339 –34350, 2014.

AJP-Gastrointest Liver Physiol • doi:10.1152/ajpgi.00443.2014 • www.ajpgi.org

G50

XO INHIBITOR FEBUXOSTAT SUPPRESSES NASH DEVELOPMENT

A

B

C

HE MCD+Cont

S ir iu s r e d MCD+Cont

MCD+Feb

MCD+Feb

n.s. 2500

60

n.s.

*

Cont 50

Feb

40

1500

ALT (IU/ml)

1000

500

30 20 10 0

ND

D

MCD

ND

MCD

n.s.

n.s.

18 14

Col1a2 mRNA levels

Col1a1 mRNA lvels

16 12 10 8 6 4 2 0 ND

MCD

9 8 7 6 5 4 3 2 1 0

n.s.

n.s.

30

4 3.5

25

TIMP2 mRNA levels

0

TIMP1 mRNA levels

Uric acid (µM)

2000

20 15 10 5 0

ND

ND

MCD

MCD

3 2.5

Cont

2

Feb

1.5 1 0.5 0 ND

MCD

Fig. 7. Febuxostat had no effect on the nonalcoholic steatohepatitis development induced by methionine choline-deficient (MCD) diet feeding. A: serum UA concentration. B: serum ALT level. C: paraffin-embedded sections were stained with hematoxylin and eosin (HE) or with Sirius red. Representative photographs of each group are shown. D: mRNA levels of collagen 1a1, collagen 1a2, TIMP1, and TIMP2. All data are shown as means ⫹ SE. *P ⬍ 0.05. 4. Dela Pena A, Leclercq IA, Williams J, Farrell GC. NADPH oxidase is not an essential mediator of oxidative stress or liver injury in murine MCD diet-induced steatohepatitis. J Hepatol 46: 304 –313, 2007. 5. Feig DI, Kang Johnson RJ DH. Uric acid and cardiovascular risk. N Engl J Med 359: 1811–1821, 2008. 6. Fleury C, Mignotte B, Vayssière JL. Mitochondrial reactive oxygen species in cell death signaling. Biochimie 84: 131–141, 2002. 7. Hsieh PS, Chan YJ, Shyu JF, Chen YT, Loh CH. Mild portal endotoxaemia induces subacute hepatic inflammation and pancreatic beta-cell dysfunction in rats. Eur J Clin Invest 38: 640 –648, 2008. 8. Ishizaka N, Ishizaka Y, Toda E, Nagai R, Yamakado M. Association between serum uric acid, metabolic syndrome, and carotid atherosclerosis in Japanese individuals. Arterioscler Thromb Vasc Biol 25: 1038 –1044, 2005. 9. Johnson RJ, Segal MS, Srinivas T, Ejaz A, Mu W, Roncal C, SánchezLozada LG, Gersch M, Rodriguez-Iturbe B, Kang DH, Acosta JH. Essential hypertension, progressive renal disease, and uric acid: a pathogenetic link? J Am Soc Nephrol 16: 1909 –1919, 2005. 10. Khosla UM, Zharikov S, Finch JL, Nakagawa T, Roncal C, Mu W, Krotova K, Block ER, Prabhakar S, Johnson RJ. Hyperuricemia induces endothelial dysfunction. Kidney Int 67: 1739 –1742, 2005. 11. Klein BE, Klein R, Lee KE. Components of the metabolic syndrome and risk of cardiovascular disease and diabetes in Beaver Dam. Diabetes Care 25: 1790 –1794, 2002. 12. Kushiyama A, Okubo H, Sakoda H, Kikuchi T, Fujishiro M, Sato H, Kushiyama S, Iwashita M, Nishimura F, Fukushima T, Nakatsu Y,

13. 14. 15. 16. 17. 18.

19.

Kamata H, Kawazu S, Higashi Y, Kurihara H, Asano T. Xanthine oxidoreductase is involved in macrophage foam cell formation and atherosclerosis development. Arterioscler Thromb Vasc Biol 32, 291–298, 2012. Leclercq IA, Farrell GC, Field J, Bell DR, Gonzalez FJ, Robertson GR. CYP2E1 and CYP4A as microsomal catalysts of lipid peroxides in murine nonalcoholic steatohepatitis. J Clin Invest 105: 1067–1075, 2000. Lee HJ, Jeong KH, Kim YG, Moon JY, Lee SH, Ihm CG, Sung JY, Lee TW. Febuxostat ameliorates diabetic renal injury in a streptozotocininduced diabetic rat model. Am J Nephrol 40: 56 –63, 2014. Lim JS, Mietus-Snyder M, Valente A, Schwarz JM, Lustig RH. The role of fructose in the pathogenesis of NAFLD and the metabolic syndrome. Nat Rev Gastroenterol Hepatol 7: 251–264, 2010. Lin CC, Yin MC, Liu WH. Alleviative effects of s-allyl cysteine and s-ethyl cysteine on MCD diet-induced hepatotoxicity in mice. Food Chem Toxicol 46: 3401–3406, 2008. Masuo K, Kawaguchi H, Mikami H, Ogihara T, Tuick ML. Serum uric acid and plasma norepinephrine concentrations predict subsequent weight gain and blood pressure elevation. Hypertension 42: 474 –480, 2003. Matsuura F, Yamashita S, Nakamura T, Nishida M, Nozaki S, Funahashi T, Matsuzawa Y. Effect of visceral fat accumulation on uric acid metabolism in male obese subjects: visceral fat obesity is linked more closely to overproduction of uric acid than subcutaneous fat obesity. Metabolism 47: 929 –933, 1998. Nakagawa T, Hu H, Zharikov S, Tuttle KR, short RA, Glushakova O, Ouyang X, Feig DI, Block ER, Herrera-Acosta J, Patel JM, Johnson

AJP-Gastrointest Liver Physiol • doi:10.1152/ajpgi.00443.2014 • www.ajpgi.org

XO INHIBITOR FEBUXOSTAT SUPPRESSES NASH DEVELOPMENT RJ. A causal role for uric acid in fructose-induced metabolic syndrome. Am J Physiol Renal Physiol 290: F625–F631, 2006. 20. Nomura J, Busso N, Ives A, Tsujimoto S, Tamura M, So A, Yamanaka Y. Febuxostat, an inhibitor of xanthine oxidase, suppresses lipopolysaccharide-induced MCP-1 production via MAPK phosphatase-1-mediated inactivation of JNK. PLoS One 8: e75527, 2013. 21. Sertoglu E, Ercin CN, Celebi G, Gurel H, Kayadibi H, Genc H, Kara M, Dogru T. The relationship of serum uric acid with non-alcoholic fatty liver disease. Clin Biochem 47: 383–388, 2014. 22. Sesti G, Fiorentino TV, Arturi F, Perticone M, Sciacqua A, Perticone F. Association between noninvasive fibrosis markers and chronic kidney disease among adults with nonalcoholic fatty liver disease. PLoS One 9: e88569, 2014.

G51

23. Tsouli SG, Liberopoulos EN, Mikhailidis DP, Athyros VG, Elisaf MS. Elevated serum uric acid levels in metabolic syndrome: an active component or an innocent bystander? Metabolism 55: 1293–1301, 2006. 24. Tsushima Y, Nishizawa H, Tochino Y, Nakatsuji H, Sekimoto R, Nagao H, Shirakura T, Kato K, Imaizumi K, Takahashi H, Tamura M, Maeda N, Funahashi T, Shimomura I. Uric acid secretion from adipose tissue and its increase in obesity. J Biol Chem 288: 27138 –27149, 2013. 25. Xu C, Wan X, Xu L, Weng H, Yan M, Miao M, Sun Y, Xu G, Dooley S, Li Y, Yu C. Xanthine oxidase in nonalcoholic fatty liver disease and hyperuricemia: One stone hits two birds. J Hepatol 62: 1412–1419, 2015. 26. Zoccali C, Mallamaci F. Uric acid, hypertension, and cardiovascular and renal complications. Curr Hypertens Rep 15: 531–537, 2013.

AJP-Gastrointest Liver Physiol • doi:10.1152/ajpgi.00443.2014 • www.ajpgi.org

The xanthine oxidase inhibitor febuxostat suppresses development of nonalcoholic steatohepatitis in a rodent model.

Xanthine oxidase (XO) is an enzyme involved in the production of uric acid (UA) from purine nucleotides. Numerous recent studies have revealed the lik...
1MB Sizes 2 Downloads 11 Views