The hepatitis C virus core protein indirectly induces alpha-smooth muscle actin expression in hepatic stellate cells via interleukin-8

Research Article

The hepatitis C virus core protein indirectly induces alpha-smooth muscle actin expression in hepatic stellate cells via interleukin-8 Sophie Clément1,*, Stéphanie Pascarella1, Stéphanie Conzelmann1, Carmen Gonelle-Gispert3, Kévin Guilloux1, Francesco Negro1,2 1

Division of Clinical Pathology, Geneva University Hospital, 1 rue Michel-Servet, 1211 Geneva 4, Switzerland; 2Division of Gastroenterology and Hepatology, Geneva University Hospital, 4 rue Gabrielle-Perret-Gentil, 1211 Geneva 14, Switzerland; 3Surgical Research Unit, Department of Surgery, Geneva University Hospital, 1 rue Michel-Servet, 1211 Geneva 4, Switzerland

Background & Aims: Progressive deposition of liver fibrosis is a common feature of chronic hepatitis associated with hepatitis C virus (HCV) infection, and it may eventually lead to cirrhosis and liver failure. Although this fibrogenic process appears to be linked to HCV protein expression and replication via indirect mechanisms, i.e., to be mediated by virally-driven inflammation, a direct role of HCV in inducing fibrosis deposition has never been entirely excluded. Methods: We established an in vitro system in which the human hepatic stellate cell line LX-2 was cultured in the presence of conditioned medium from human hepatoma Huh-7 cells transduced with a lentiviral vector expressing HCV core proteins of different genotypes. Results: Treatment of LX-2 cells, with conditioned medium from Huh-7 cells expressing HCV core protein, led to the activation of a-smooth muscle actin expression. Among the chemokines secreted by cells transduced with HCV core, interleukin-8 was identified as the strongest inducer of a-smooth muscle actin expression in LX-2 and primary hepatic stellate cells. This effect was accompanied by a decrease in cell migration and increased focal contact organisation. Conclusions: The expression of the HCV core in hepatocytes may contribute to the establishment of a profibrogenic microenvironment. Ó 2010 European Association for the Study of the Liver. Published by Elsevier B.V. All rights reserved.

Keywords: Liver fibrosis; Cytokines; Hepatitis C. Received 22 June 2009; received in revised form 8 October 2009; accepted 11 October 2009; available online 5 March 2010 * Corresponding author. Tel.: +41 22 3795060; fax: +41 22 3724920. E-mail address: [email protected] (S. Clément). Abbreviations: HCV, hepatitis C virus; HSC, hepatic stellate cells; IL-8, interleukin8; PCR, polymerase chain reaction; EF1, elongation factor 1; EGFP, empty and green fluorescent protein; SMA, smooth muscle actin; FCS, foetal calf serum; CM, conditioned medium; SD, standard deviation; MMP2, matrix metalloproteinase 2; TIMP1, tissue inhibitor of metalloproteinase 1; CTGF, connective tissue growth factor; EMT, epithelial to mesenchymal transition.

Introduction Hepatitis C virus (HCV) infects approximately 170 million people worldwide, and a proportion of them develop cirrhosis and hepatocellular carcinoma. Chronic hepatitis C is characterised by the relentless deposition of liver fibrosis, a process that may take decades before the cirrhotic stage is reached. Hepatic fibrosis occurring in chronic hepatitis C is a woundhealing response to persistent liver injury [1]. Damaged hepatocytes infected by HCV release various mediators that recruit inflammatory cells [2]. These, in turn, activate hepatic stellate cells (HSC) to secrete collagen [3]. Since the latter are a major source of proinflammatory chemokines, a vicious circle ensues, with inflammation and fibrogenesis amplifying each other [1]. Thus, in chronic hepatitis C, the fibrogenic process seems to be mostly mediated by virally-driven inflammation. The liver fibrosis progression rate, however, varies considerably among chronically HCV-infected patients. Some patients present with no fibrosis even after decades of infection while others rapidly develop cirrhosis in a few years. Host cofactors like age at infection, insulin resistance, gender, alcohol consumption, coinfections, and genetic polymorphisms influence the progression of hepatitis C [4,5]. Conversely, the impact of viral factors is debated, although recent reports suggest a direct fibrogenic effect of HCV [4,6]. Several viral proteins may induce oxidative stress, steatosis, and apoptosis, leading to direct HSC activation without the participation of the inflammatory response. Thus, some patients may present with significant liver fibrosis despite minimal inflammation. In the present work, we show that HCV core protein-expressing hepatocytes secrete interleukin-8 (IL-8), and that this cytokine participates in the activation of HSCs, suggesting a fibrogenic effect of HCV via paracrine mechanisms.

Materials and methods Antibodies and chemicals Anti-a-SMA (asm-1, [7]) and anti-b-cytoplasmic actin antibodies were obtained from C. Chaponnier (Geneva). Monoclonal anti-core (C7–50) antibody was from Axxora Europe (Lausen, Switzerland). Monoclonal anti-vinculin antibody (clone

Journal of Hepatology 2010 vol. 52 j 635–643

Research Article vector. The lentivector particles were produced by transient transfection in HEK 293T cells, collected after 72 h and concentrated 120-fold by ultracentrifugation. Viral titre was estimated by quantitative RT-PCR [9] and immunofluorescence (see below). Empty and green fluorescent protein (EGFP)-encoding lentivectors were used as controls.

hVin-1) was from Sigma (St Louis, MO). Blocking anti-IL-8 and human recombinant IL-8 were from R&D Systems Europe (Abingdon, UK). The chemokine (C– X–C motif) receptor 2 (CXCR2) antagonist (SB 265610) and TGF-b1 were from Tocris (Ellisville, MO) and ABD Serotech (Düsseldorf, Germany), respectively. Cloning of HCV core in a lentiviral vector

Cell culture, transduction, and treatment Core-encoding sequences were amplified by PCR from pIRES2-EGFP plasmids [8] and inserted into pENTR4 (Gateway lentiviral system, Invitrogen) after digestion with XhoI and BamHI. With respect to the originally reported sequence, the genotype 1b core variant used in the present study has a threonine replacing an alanine at position 131. This plasmid was recombined with pENTR-EF1-a containing the ubiquitous Elongation Factor 1 (EF1) promoter into the destination

Human hepatoma Huh-7 cells were cultured in DMEM (Invitrogen Life Technologies, Basel, Switzerland, Ref. 31885–023) supplemented with 10% foetal calf serum (FCS), 100 U/ml penicillin and 100 lg/ml streptomycin (Invitrogen Life Technologies). Twenty-four hours before transduction with lentiviral vectors, the growth medium was replaced with serum-free medium consisting of

A

B

55

35 27

1

2

3

4

5

1

2

3

4

5

14

*

12 α- SMA protein level (ratio to CM Control)

kDa 130 100 70

C

α-SMA

Ponceau red

10 8 6

*

4 2 0 CM 1b

CM 2a

CM 3a

CM 4h CM Huh-7

Fig. 1. Effect of CM from Huh-7 cells expressing HCV core protein on LX-2 activation. CM obtained from Huh-7 cells expressing HCV core 1b, 2a, 3a, and 4h was applied to LX-2 cells for 72 h. Activation of LX-2 cells was estimated by a-SMA expression levels. CM control (CM cont) corresponds to the mean of two controls used in these experiments: cells untransduced or transduced with an EGFP-encoding lentivector. (A) a-SMA was stained by immunofluorescence and nuclei by DAPI. (a: CM Huh-7, b: CM GFP, c: CM 1b, d: CM 2a, e: CM 3a, f: CM 4h) Scale bar: 20 lm. (B and C) a-SMA expression was also estimated by immunoblot. (B) Total protein extracts from LX-2 cells treated with the different CM were subjected to SDS–PAGE, transferred to nitrocellulose (first 5 lanes, Ponceau red staining) and blotted with anti-a-SMA. (1: CM 1b, 2: CM 2a, 3: CM 3a, 4: CM 4h, 5: cont CM) (C) a-SMA signals were quantified by densitometry and normalised using ponceau red staining as a loading control. Bars represent the SD of three independent experiments (*p 60.05 compared to control).

636

Journal of Hepatology 2010 vol. 52 j 635–643

JOURNAL OF HEPATOLOGY DMEM/Ham’s nutrient F12 (1:1) supplemented with insulin–transferrin–selenium (Invitrogen) and vitamin C (0.2 mM). Conditioned media (CM) were collected 72 h later. Core protein in transduced cells was quantified using the Ortho HCV core Ag ELISA (Wako Chemicals, Osaka, Japan). LX-2 cells [10] and human primary HSCs, isolated using Nycodenz (Sigma) density gradient centrifugation as described [11] with minor modifications, were cultured in IMDM medium (Invitrogen) containing 10% FCS, penicillin, and streptomycin (Invitrogen). HSCs were treated with CM of core-expressing Huh-7 cells or recombinant IL-8. For IL-8 (1–100 ng/ml for 48–120 h) treatment, the growth medium was replaced with serum-free medium (see above). For migration assays, confluent LX-2 cell monolayers were scratched with a 20-ll pipette tip, washed with serum-free medium and incubated for 8–20 h with 10 ng/ml IL-8, with or without mitomycin C (2.5 lM, pre-treatment for 2 h before wounding). Images were acquired using an Axiophot photomicroscope (Zeiss, Oberkochen, Germany) coupled to OpenLab™ Software (Improvision, Coventry, England). Quantitative analyses were performed using Metamorph software (Molecular Devices Corporation, Sunnyvale, CA).

Results Lentiviral expression of HCV core proteins of genotypes 1b, 2a, 3a, and 4h All genotypes were subcloned in the Gateway lentiviral system. Titrated lentiviral constructs were used to transduce Huh-7 cells,

-

+

A

Measurement of cytokines Secretion profiles of transduced cells were determined using human cytokine antibody arrays (C series 1000; Raybiotech, Inc., Norcross, GA). IL-8 levels in CM were measured by ELISA (R&D Systems).

Indirect immunofluorescence and confocal microscopy

B

RNA isolation, reverse transcription, and real-time PCR Total RNA was extracted using the RNeasy™ Mini Kit (Qiagen, Hombrechtikon, Switzerland). cDNA was synthesised from 100 ng total RNA with SuperScript™ II RNase H() reverse transcriptase (Invitrogen Life Technologies) and random hexanucleotides. For real-time PCR, the following SYBR Green human QuantiTect Primers (from Qiagen) were used: IL-8 (QT00000322), collagen I (QT00037793), MMP2 (QT00088396), TIMP1 (QT00084168), CXCR2 (QT00000518), and CTGF (QT00052899). Relative quantification was performed by real-time PCR as described [12]. For qualitative RT-PCR, specific primers (Invitrogen) were used for human CXCR1 (forward GCAGCTCCTACTGTTGGACA and reverse GGCATGCCAGTGAAATTTAGA) and CXCR2 (forward GCTCTGACTACCACCCAACC and reverse GCTGGGCTTTTCACCTGTAG).

Electrophoretic and immunoblot analysis Proteins from lysed cells were separated on polyacrylamide gels and transferred to nitrocellulose membranes (Milian, Geneva, Switzerland). Membranes were blocked with 5% skim milk in wash buffer (20 mM Tris–HCl, pH 7.6, 140 mM NaCl, 0.1% Tween 20) and incubated with anti-a-SMA or anti-b-cytoplasmic actin antibodies diluted in blocking solution. Following three washes, membranes were incubated with peroxidase-conjugated goat anti-mouse antibodies (BioRad, Hercules, CA) diluted 1:10,000 in wash buffer. Proteins were revealed by chemiluminescence (ECL, Amersham Biosciences AB, Uppsla, Sweden). Film was scanned and spots were quantified using the NIH-Image analysis program Scion IMAGE (Scion Corp., Frederick, MD). Statistical analysis Results were expressed as means ± standard deviation (SD) of three independent experiments. Results were analysed by Student’s t-test. Values of ***p <0.001, **p < 0.01, and *p < 0.05 were considered statistically significant.

8 7

***

6 5 4

**

3 2 1 0 CM 1b

C

IL-8 mRNA level (ratio to control)

Cells were fixed in 3% paraformaldehyde for 10 min at RT and permeabilised with 0.3% Triton X-100 for 10 min. Cells were first incubated with anti-core, anti-aSMA (IgG2a) and mouse anti-vinculin (IgG1) diluted in PBS–Tween 0.1% for 30 min at RT, and then with FITC-conjugated anti-mouse antibodies (IgG2a specific; Southern Biotechnology, Birmingham, AL); rhodamine-conjugated antimouse antibodies (IgG1 specific; Southern Biotechnology) and DAPI for nuclear staining for 30 min at RT. After washing in PBS, cells were mounted in Moviol. Images were acquired using either an Axiophot microscope (Carl Zeiss) equipped with an Axiocam camera (Carl Zeiss) or a confocal microscope (LSM510 Meta, Zeiss). Percentages of a-SMA positive/negative cells (or core positive/negative cells) were calculated by counting P200 cells.

IL-8 amount (ratio to control)

- +

CM 2a

CM 3a

CM 4h CM control

7 6 5 4 3

***

2

*

1 0 1b

2a

3a

4h

Control

Fig. 2. Huh-7 cells expressing HCV core 1b produce high levels of IL-8. (A) A cytokine antibody array was used for the simultaneous detection of 40 inflammatory factors. The antibody-coated membrane was incubated with CM from Huh-7 cells expressing HCV core 1b. +, positive control; , negative control; (1) IL-8, (2) MIP-1b, (3) MIP-1d, and (4) RANTES. (B) The level of IL-8 in the CM of Huh-7 cells expressing core protein of different genotypes was measured by ELISA. *p <0.05, ***p 60.001 versus cont CM. (C) Quantification of IL-8 mRNA level was assessed in Huh-7 expressing the core of different genotypes by real-time PCR. **p 60.01, ***p 60.001 versus control. Bars represent SD of three independent experiments.

Journal of Hepatology 2010 vol. 52 j 635–643

637

Research Article and the expression level and correct localisation of HCV core proteins were investigated by immunofluorescence. Approximately 60–80% of cells expressed core protein 48 h after transduction (Supplementary Fig. 1A and B). The core protein content (by ELISA) was comparable across genotypes (Supplementary Fig. 1C).

detected at high levels, the expression of IL-8 being the strongest (see Fig. 2A and Supplementary Fig. 2 for the relative expression levels of all 40 inflammatory factors). Real-time PCR amplification of mRNA extracted from Huh-7 cells expressing HCV cores of different genotypes (Fig. 2C), and ELISA quantification of IL-8 secreted into the CM (Fig. 2B) showed that the production of IL-8 by Huh-7 cells was influenced by the core protein genotype. When genomic-length HCV RNAs were used to transfect Huh-7.5 cells (see Supplementary Materials and methods), both genotypes 1b and 4a strongly induced IL-8 expression. Genotype 2a also activated IL-8, whereas the activation induced by genotype 3a was lower (Supplementary Fig. 3).

CM of Huh-7 cells expressing HCV core protein induces a-SMA expression in HSCs To determine whether HCV core protein expression may indirectly activate HSCs, Huh-7 cells were transduced with lentivectors expressing the core of different HCV genotypes, and five days after transduction, CM was collected and used to culture LX-2 cells for another 72 h. The CM of Huh-7 cells expressing core 1b or 4 h induced an increase in a-SMA expression, a marker of myofibroblast differentiation (Fig. 1) [13]. Quantitative analysis of immunoblots indicated that CM of Huh-7 cells expressing core 1b had the strongest effect on a-SMA expression.

IL-8 secreted by Huh-7 cells expressing the HCV 1b core protein induces a-SMA expression Blocking antibodies against IL-8, but not control antibodies, blocked a-SMA expression in LX-2 cells induced by the CM of Huh-7 cells expressing the 1b core protein (Fig. 3). In a parallel experiment, recombinant IL-8 induced a-SMA expression and organisation of stress fibres in both LX-2 cells and primary HSCs (Fig. 4). The IL-8-mediated increase of a-SMA expression was dose-dependent (2.5 ± 0.6-fold increase with 1 ng/ml IL-8; 3.7 ± 0.4 with 10 ng/ml IL-8). A higher concentration of IL-8 (100 ng/ml) had a lower effect on a-SMA expression

Huh-7 cells expressing the HCV core protein secrete IL-8 To identify the factor responsible for a-SMA activation, we characterised the cytokine profile of CM obtained from Huh-7 cells expressing core 1b. MIP-1d, MIP-1b, RANTES, and IL-8 were

A

30 α-SMA positive cells /nuclei

D

B

C

E

**

25

α-SMA

β-cyto actin

20 15

1

2

3

1

2

3

10 5 0 CM 1b

CM 1b + anti - IL8

CM 1b + IgG cont

Fig. 3. Blocking antibodies against IL-8 inhibit CM-induced a-SMA expression in LX-2. a-SMA expression was assessed by immunostaining after 72 h in the presence of CM core 1b (A), CM core 1b incubated with blocking anti-IL-8 antibodies (B) 2.5 lg/ml, or CM core 1b incubated with a non-specific control antibody (C). Scale bar: 20 lm. (D) The percentage of a-SMA-positive cells over total (calculated as number of DAPI-stained nuclei) was estimated by counting P200 cells. Bars represent the SD. **p 60.01. (E) Protein extracts from LX-2 cells treated with CM core 1b (lane 1), CM core 1b incubated with blocking anti-IL-8 antibodies (lane 2) or CM core 1b incubated with a nonspecific control antibody (lane 3) were immunoblotted with anti-a-SMA or anti-b-cytoplasmic actin antibodies.

638

Journal of Hepatology 2010 vol. 52 j 635–643

JOURNAL OF HEPATOLOGY A

α-SMA

Coomassie Blue 1

2

3

4

1

2

3

β-cyto actin 4

1

2

3

4

kDa 100 70 55 35 27 15

C α- SMA protein level (ratio to control)

B

4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0

*** *

1 Control

10

100

ng/ml

IL-8

Fig. 4. Induction of a-smooth muscle actin expression in HSCs by IL-8. (A) Representative pictures of LX-2 cells (a–d) or primary human HSCs (e–h) incubated for 72 h with control medium (untreated, a and e) or IL-8 at increasing concentrations (1 ng/ml, b and f; 10 ng/ml, c and g; 100 ng/ml, d and h) and immunostained with anti-a-SMA. Scale bar: 50 lm. (B) Total protein extracts from LX-2 cells either untreated (lane 1) or treated with IL-8 (lanes 2–4) were subjected to SDS–PAGE (first 4 lanes, Coomassie blue staining), transferred to nitrocellulose and immunoblotted with anti-a-SMA or anti-b-cytoplasmic actin antibodies. (C) The a-SMA signal was quantified by densitometry and normalised using Coomassie blue as a loading control. Bars represent the SD of three independent experiments. *p <0.05, ***p 60.001 compared to controls.

(2.1 ± 0.8-fold increase), suggesting that the receptor may undergo desensitisation [14]. We investigated the expression of the IL-8 receptors CXCR1 and CXCR2 in LX-2 cells by RT-PCR. Fig. 5A shows that LX-2 cells express CXCR2 but not CXCR1. Its expression was twice as high in quiescent conditions (1% FCS) compared to 10% FCS conditions (Fig. 5B). Moreover, treatment with SB265610, a CXCR2 antagonist, abolished IL-8-induced a-SMA expression (Fig. 5C). IL-8 does not induce collagen 1, MMP2, TIMP1 or CTGF mRNA expression in LX-2 cells Activated HSCs display increased production of extracellular matrix and profibrotic factors such as collagen 1, matrix metalloproteinase 2 (MMP2), tissue inhibitor of metalloproteinase 1 (TIMP1), and connective tissue growth factor (CTGF). Therefore, we tested the effect of IL-8 on the mRNA levels of these factors.

Although incubation with TGF-b1 (used as positive control) induced a significant increase in all four mRNA levels in LX-2 cells, no effect was observed upon treatment with IL-8 for 48 h (Fig. 6). A time-course analysis revealed that IL-8 did not induce mRNA transcription after 8 or 24 h (not shown). IL-8 decreases LX-2 migration To investigate the effect of IL-8 on HSC migration, a key event in fibrogenesis, we performed an in vitro scratch wound-healing assay. IL-8 reduced HSC migration by approximately 2-fold (Fig. 7). To determine the contribution of cell proliferation in the wound-healing assay, cells were pre-treated with mitomycin C for 2 h before scratching. Although mitomycin C treatment blocked cell proliferation as evaluated by cell counting (not shown), the difference in cell migration in the presence or absence of IL-8 was maintained (Fig. 7B).

Journal of Hepatology 2010 vol. 52 j 635–643

639

Research Article 12

PBMC

mRNA level (ratio to control)

CXCR2 LX 2 (10%)

PBMC

LX 2 (10%)

bp 300 200

MW

LX 2 (1%)

CXCR1

LX 2 (1%)

A

(arbitrary units)

B

CXCR2 mRNA expression

100

0.18 0.16 0.14 0.12 0.10 0.08 0.06 0.04 0.02 0

***

10 8

Control TGF-beta (50 ng/ml) IL-8 (1 ng/ml) IL-8 (10 ng/ml) IL-8 (100 ng/ml)

6 4

***

*** ***

2 0

Collagen I

MMP-2

TIMP-1

CTGF

Fig. 6. Effect of IL-8 on the expression of collagen I, MMP2, TIMP1, and CTGF evaluated by real-time PCR. mRNA levels of collagen I, MMP-2, TIMP-1 and CTGF were quantified by real-time PCR in cells treated for 48 h with increasing concentrations of IL-8 (1–100 ng/ml) in serum-free conditions. Untreated cells and TGF-b1-treated cells were used as negative and positive controls, respectively. Bars represent the SD of three independent experiments. ***p 60.001 compared to controls.

A

LX2- 1% FCS LX2- 10% FCS

C

α-SMA

B 1

2

3

Fig. 5. LX-2 hepatic stellate cells express CXCR2 but not CXCR1. (A) CXCR1 (lanes 1–3) and CXCR2 (lanes 4–6) mRNA levels were assessed in LX-2 HSCs cultured either in 1% FCS (lanes 1 and 4) or 10% FCS (lanes 2 and 5) by RT-PCR, using PBMC (lanes 3 and 6) as a positive control. (B) Comparison of expression levels between 1% and 10% FCS conditions was assessed by real-time PCR. (C) aSMA expression in LX-2 cells was evaluated in the presence of the CXCR2 antagonist SB265610 by immunofluorescence (untreated (a) or treated with 10 ng/ml IL-8 in the absence (b) or presence (c) of SB265610) and immunoblotting (d) with anti-a-SMA or anti-b-cytoplasmic actin antibodies (untreated (lane 1) or treated with IL-8 in the absence (lane 2) or presence (lane 3) of SB265610). Scale bar: 50 lm.

Cell migration (ratio to control)

β-cyto actin

Control IL-8 (10 ng/ml)

1.2 1.0 0.8

**

**

0.6 0.4 0.2 0 No Mitomycin

Mitomycin (2.5 μM)

As focal adhesion assembly and disassembly are determinant events in HSC migration ability, we investigated the organisation of focal contacts. As expected, the decrease in cell migration correlated with an increased organisation of both a-SMA-positive stress fibres and focal contacts, as visualised by staining of one of their constituent vinculin (Fig. 8, red staining). 640

Fig. 7. Analysis of HSC migration. LX-2 cells were either untreated (a and b) or treated with 10 ng/ml IL-8 (c and d), with or without 2.5 lM mitomycin C. (A) Images were acquired at t = 0 (a and c) and t = 20 h (b and d) in in vitro scratch assays using a 10 objective. Dotted black lines delineate acellular areas at t = 0. (B) Acellular areas before and after IL-8 treatment were measured, and the difference, indicating cell migration, was plotted. Bars represent the SD of three independent experiments. **p 60.01 compared to controls.

Journal of Hepatology 2010 vol. 52 j 635–643

JOURNAL OF HEPATOLOGY

Fig. 8. a-Smooth muscle actin and vinculin distribution in primary HSCs. Confocal micrographs of HSCs untreated (a–c) or treated with 10 ng/ml IL-8 (d–f) were stained with anti-a-SMA (a and d) and vinculin (b and e). Image overlays are shown in c and f. Scale bars: 50 lm.

Discussion There is some evidence that HCV may influence the rate of liver disease progression [4,6,15–17]. Some studies have suggested that specific HCV genotypes are associated with accelerated evolution of chronic hepatitis C. Genotype-specific disease expression is a strong indicator of a virally-induced effect on pathogenesis. However, the association between HCV genotype and disease progression is still debated. Although there is good evidence that genotype 1b may be a risk factor associated with hepatocellular carcinoma [18], when it comes to fibrogenesis, a single recent study has suggested that genotype 3 may lead to accelerated fibrosis progression [6]. In this work, we present evidence supporting the ability of HCV to induce paracrine HSC activation via IL-8. The effect was influenced by the viral genotype, a phenomenon observed using either lentivectors expressing the isolated core protein or genomic-length HCV RNA. In the former model, IL-8 activation was the strongest with genotype 1b and less for genotype 4a, while using the full-length HCV, we noticed significant IL-8 activation with genotypes 1b, 2a, and 4a. To reconcile these data, we suggest that IL-8 activation is due to core-encoding sequence polymorphisms that are only partially genotype-specific. Alternatively, proteins other than the core may modulate this effect, similarly to what was suggested for HCV-induced steatosis [19]. Whatever

the explanation, our results support the hypothesis that HCV may actively contribute to the fibrogenic process via the paracrine effect of IL-8 secreted by infected hepatocytes. Liver fibrosis is a wound-healing response characterised by the deposition of extracellular matrix. Many potentially fibrogenic cell populations in the liver have been described, such as portal fibroblasts, mesenchymal cells derived from the bone marrow, and even hepatocytes and biliary epithelial cells, which may undergo an epithelial to mesenchymal transition (EMT) [20]. However, HSCs are recognised as the major source of liver fibrosis. We report that HSCs incubated in the presence of CM from HCV core-expressing hepatocytes increased their expression of a-SMA, a key marker of myofibroblastic differentiation [13]. Thus, although HCV may not be able to induce the EMT in infected hepatocytes (no expression of a-SMA was observed in HCV core protein-expressing hepatocytes, data not shown), it appears to induce the secretion of mediators that activate the expression of a-SMA in HSCs. Interestingly, similar results were reported for hepatocytes expressing the hepatitis B virus X protein, which induced the paracrine activation of HSCs in a TGFb-dependent manner [21]. Here, we show that the HCV core-induced paracrine activation of HSCs is partly due to IL-8. IL-8 belongs to the C–X–C superfamily of chemokines that recruit and activate leukocytes during inflammation. Several cell types express the IL-8 receptors CXCR1

Journal of Hepatology 2010 vol. 52 j 635–643

641

Research Article and CXCR2 [22,23]. We show that the HSC response to IL-8 is probably mediated via CXCR2. IL-8 is implicated in several fibrotic disorders, such as chronic pancreatitis and pulmonary fibrosis [24,25]. However, only few reports have indicated that IL-8 directly acts as profibrotic chemokine: in human prostate fibroblast cultures [26] and in wound healing [27], where it was shown to induce a-SMA expression and stress fibre formation. In HCV infection, IL-8 has been implicated in HCV–host interactions as well as in HCV-associated pathogenesis. IL-8 is up-regulated in hepatitis C and seems to correlate with clinical pathologic manifestations and disease progression [28–30]. Several HCV proteins have been shown to increase IL-8 expression [31–33] in hepatocytes. Regarding the ability of IL-8 to induce HSC activation, it is noteworthy that among the different markers of myofibroblastic differentiation analysed, only a-SMA expression was markedly increased upon IL-8 treatment. Other effectors have been shown to induce the expression of only some HSC activation markers (e.g., anaphylatoxin C5a only induced fibronectin [34]). The fact that IL-8 does not induce MMP-2 is in agreement with our results showing that IL-8 stops the migratory activity of LX-2 cells, as it has been reported that growth factors, that induce MMP-2 and potentially facilitate degradation of the normal space of Disse microenvironment, induce migration [35]. The observation that IL-8 blocked LX-2 cell migration in our experimental settings is of great interest. This could indicate that IL-8 can, under certain conditions, provide a stop signal to HSCs once they have reached the site of injury. Similarly, Toll-Like Receptor 9 activation inhibits PDGF-mediated migration and also induces fibrogenesis [36]. Although many molecules serve as chemoattractant stimuli, little data are available on the mechanisms leading to cell migration arrest. Receptor desensitisation at high concentrations of chemokines has been proposed. Concerning IL-8, Feniger-Barish et al. [37] have demonstrated that depending on its concentration, IL-8 can either activate cell migration or induce a migratory shut-off. The dilution range used in their report is between 10 (low) and 1000 (high) ng/ml. In our conditions, the concentration of 10 ng/ml already induced a blockade of LX-2 migration. This discrepancy could be explained by the fact that Feniger-Barish et al. [37] used IL-8 receptor-overexpressing cells, which probably need a higher concentration to reach the status of non-migratory cells. We also found that this inhibition of migration induced by IL-8 was accompanied by increased organisation of vinculin within focal contacts. This observation is in complete agreement with previous data obtained from cells depleted of vinculin displaying reduced adhesion to a variety of ECM proteins and increased migration rates [38]. Thus, IL-8 may act in concert with other chemokines as a factor leading to stress fibre formation via (i) formation of fibronexus, and (ii) formation of stress fibres and their activation via focal contacts. In conclusion, we propose that HCV core-expressing hepatocytes produce IL-8 and that IL-8 acts as a mediator of HSC differentiation by (i) providing a stop signal to migrating HSCs when they have reached a fibrosis site, and (ii) inducing a contractile phenotype associated with focal adhesion and up-regulation of a-SMA production. This may provide an additional mechanism of induction of liver fibrosis in chronic hepatitis C. IL-8 secreted by HCV-infected hepatocytes may not only be responsible for local inflammatory cell recruitment, but also for the direct activation of HSCs.

642

Acknowledgments The authors who have taken part in this study declared that they do not have anything to declare regarding funding from industry or conflict of interest with respect to this manuscript. This work was supported by research Grants 320000–116544 and 3100A0– 109888 from the Swiss National Science Foundation. We thank Scott Friedman (Mount Sinai School of Medicine, New York, NY) for providing the LX-2 cells, Charles Rice (Rockfeller University, New York, NY) for the Huh-7.5 cells; Ralf Bartenschlager (Heidelberg, Germany) for providing full length HCV RNA of genotypes 1b and 2a; Jens Bukh (Copenhagen, Denmark) for providing full length HCV RNA of genotypes 3a and 4a; and the Genomics Platform of the NCCR program ‘‘Frontiers in Genetics” for performing and analysing the real-time PCR experiments.

Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.jhep.2009.10.035. References [1] Bataller R, Brenner DA. Liver fibrosis. J Clin Invest 2005;115:209–218. [2] Heydtmann M, Adams DH. Chemokines in the immunopathogenesis of hepatitis C infection. Hepatology 2009;49:676–688. [3] Casini A, Ceni E, Salzano R, Biondi P, Parola M, Galli A, et al. Neutrophilderived superoxide anion induces lipid peroxidation and stimulates collagen synthesis in human hepatic stellate cells: role of nitric oxide. Hepatology 1997;25:361–367. [4] Missiha SB, Ostrowski M, Heathcote EJ. Disease progression in chronic hepatitis C: modifiable and nonmodifiable factors. Gastroenterology 2008;134:1699–1714. [5] Jonsson JR, Purdie DM, Clouston AD, Powell EE. Recognition of genetic factors influencing the progression of hepatitis C: potential for personalized therapy. Mol Diagn Ther 2008;12:209–218. [6] Bochud PY, Cai T, Overbeck K, Bochud M, Dufour JF, Müllhaupt B, et al. Genotype 3 is associated with accelerated fibrosis progression in chronic hepatitis C. J Hepatol 2009;51:655–666. [7] Skalli O, Ropraz P, Trzeciak A, Benzonana G, Gillessen D, Gabbiani G. A monoclonal antibody against alpha-smooth muscle actin: a new probe for smooth muscle differentiation. J Cell Biol 1986;103:2787–2796. [8] Abid K, Pazienza V, de Gottardi A, Rubbia-Brandt L, Conne B, Pugnale P, et al. An in vitro model of hepatitis C virus genotype 3a-associated triglycerides accumulation. J Hepatol 2005;42:744–751. [9] Sastry L, Johnson T, Hobson MJ, Smucker B, Cornetta K. Titering lentiviral vectors: comparison of DNA, RNA and marker expression methods. Gene Ther 2002;9:1155–1162. [10] Xu L, Hui AY, Albanis E, Arthur MJ, O’Byrne SM, Blaner WS, et al. Human hepatic stellate cell lines, LX-1 and LX-2: new tools for analysis of hepatic fibrosis. Gut 2005;54:142–151. [11] Schafer S, Zerbe O, Gressner AM. The synthesis of proteoglycans in fatstoring cells of rat liver. Hepatology 1987;7:680–687. [12] Clement S, Juge-Aubry C, Sgroi A, Conzelmann S, Pazienza V, Pittet-Cuenod B, et al. Monocyte chemoattractant protein-1 secreted by adipose tissue induces direct lipid accumulation in hepatocytes. Hepatology 2008;48:799–807. [13] Skalli O, Vandekerckhove J, Gabbiani G. Actin-isoform pattern as a marker of normal or pathological smooth-muscle and fibroblastic tissues. Differentiation 1987;33:232–238. [14] Govindaraju V, Michoud MC, Al-Chalabi M, Ferraro P, Powell WS, Martin JG. Interleukin-8: novel roles in human airway smooth muscle cell contraction and migration. Am J Physiol Cell Physiol 2006;291:C957–C965. [15] Bataller R, Paik YH, Lindquist JN, Lemasters JJ, Brenner DA. Hepatitis C virus core and nonstructural proteins induce fibrogenic effects in hepatic stellate cells. Gastroenterology 2004;126:529–540. [16] Shin JY, Hur W, Wang JS, Jang JW, Kim CW, Bae SH, et al. HCV core protein promotes liver fibrogenesis via up-regulation of CTGF with TGF-beta1. Exp Mol Med 2005;37:138–145.

Journal of Hepatology 2010 vol. 52 j 635–643

JOURNAL OF HEPATOLOGY [17] Schulze-Krebs A, Preimel D, Popov Y, Bartenschlager R, Lohmann V, Pinzani M, et al. Hepatitis C virus-replicating hepatocytes induce fibrogenic activation of hepatic stellate cells. Gastroenterology 2005;129:246–258. [18] Bruno S, Crosignani A, Maisonneuve P, Rossi S, Silini E, Mondelli MU. Hepatitis C virus genotype 1b as a major risk factor associated with hepatocellular carcinoma in patients with cirrhosis: a seventeen-year prospective cohort study. Hepatology 2007;46:1350–1356. [19] Piodi A, Chouteau P, Lerat H, Hezode C, Pawlotsky JM. Morphological changes in intracellular lipid droplets induced by different hepatitis C virus genotype core sequences and relationship with steatosis. Hepatology 2008;48:16–27. [20] Kalluri R, Neilson EG. Epithelial–mesenchymal transition and its implications for fibrosis. J Clin Invest 2003;112:1776–1784. [21] Mart?n-Vilchez S, Sanz-Cameno P, Rodr?guez-Muñoz Y, Majano PL, Mol?naJiménez F, López-Cabrera M, et al. The hepatitis B virus X protein induces paracrine activation of human hepatic stellate cells. Hepatology 2008;47:1872–1883. [22] Horuk R, Martin AW, Wang Z, Schweitzer L, Gerassimides A, Guo H, et al. Expression of chemokine receptors by subsets of neurons in the central nervous system. J Immunol 1997;158:2882–2890. [23] Murdoch C, Monk PN, Finn A. Cxc chemokine receptor expression on human endothelial cells. Cytokine 1999;11:704–712. [24] Carre PC, Mortenson RL, King Jr TE, Noble PW, Sable CL, Riches DW. Increased expression of the interleukin-8 gene by alveolar macrophages in idiopathic pulmonary fibrosis. A potential mechanism for the recruitment and activation of neutrophils in lung fibrosis. J Clin Invest 1991;88:1802–1810. [25] Motoo Y, Xie MJ, Mouri H, Sawabu N. Expression of interleukin-8 in human obstructive pancreatitis. JOP 2004;5:138–144. [26] Schauer IG, Ressler SJ, Tuxhorn JA, Dang TD, Rowley DR. Elevated epithelial expression of interleukin-8 correlates with myofibroblast reactive stroma in benign prostatic hyperplasia. Urology 2008;72:205–213. [27] Feugate JE, Wong L, Li QJ, Martins-Green M. The CXC chemokine cCAF stimulates precocious deposition of ECM molecules by wound fibroblasts, accelerating development of granulation tissue. BMC Cell Biol 2002;3:13. [28] Mahmood S, Sho M, Yasuhara Y, Kawanaka M, Niiyama G, Togawa K, et al. Clinical significance of intrahepatic interleukin-8 in chronic hepatitis C patients. Hepatol Res 2002;24:413–419.

[29] Polyak SJ, Khabar KS, Rezeiq M, Gretch DR. Elevated levels of interleukin-8 in serum are associated with hepatitis C virus infection and resistance to interferon therapy. J Virol 2001;75:6209–6211. [30] Tachibana Y, Nakamoto Y, Mukaida N, Kaneko S. Intrahepatic interleukin-8 production during disease progression of chronic hepatitis C. Cancer Lett 2007;251:36–42. [31] Balasubramanian A, Ganju RK, Groopman JE. Hepatitis C virus and HIV envelope proteins collaboratively mediate interleukin-8 secretion through activation of p38 MAP kinase and SHP2 in hepatocytes. J Biol Chem 2003;278:35755–35766. [32] Polyak SJ, Khabar KS, Paschal DM, Ezelle HJ, Duverlie G, Barber GN, et al. Hepatitis C virus nonstructural 5A protein induces interleukin-8, leading to partial inhibition of the interferon-induced antiviral response. J Virol 2001;75:6095–6106. [33] Oem JK, Jackel-Cram C, Li YP, Kang HN, Zhou Y, Babiuk LA, et al. Hepatitis C virus non-structural protein-2 activates CXCL-8 transcription through NFkappaB. Arch Virol 2008;153:293–301. [34] Schlaf G, Schmitz M, Heine I, Demberg T, Schieferdecker HL, Götze O. Upregulation of fibronectin but not of entactin, collagen IV and smooth muscle actin by anaphylatoxin C5a in rat hepatic stellate cells. Histol Histopathol 2004;19:1165–1174. [35] Yang C, Zeisberg M, Mosterman B, Sudhakar A, Yerramalla U, Holthaus K, et al. Liver fibrosis: insights into migration of hepatic stellate cells in response to extracellular matrix and growth factors. Gastroenterology 2003;124:147–159. [36] Watanabe A, Hashmi A, Gomes DA, Town T, Badou A, Flavell RA, et al. Apoptotic hepatocyte DNA inhibits hepatic stellate cell chemotaxis via tolllike receptor 9. Hepatology 2007;46:1509–1518. [37] Feniger-Barish R, Yron I, Meshel T, Matityahu E, Ben-Baruch A. IL-8-induced migratory responses through CXCR1 and CXCR2: association with phosphorylation and cellular redistribution of focal adhesion kinase. Biochemistry 2003;42:2874–2886. [38] Saunders RM, Holt MR, Jennings L, Sutton DH, Barsukov IL, Bobkov A, et al. Role of vinculin in regulating focal adhesion turnover. Eur J Cell Biol 2006;85:487–500.

Journal of Hepatology 2010 vol. 52 j 635–643

643