Mechanistic link between the anti-HCV effect of interferon gamma and control of viral replication by a ras-MAPK signaling cascade†
Article first published online: 22 DEC 2005
Copyright © 2005 American Association for the Study of Liver Diseases
Volume 43, Issue 1, pages 81–90, January 2006
How to Cite
Huang, Y., Chen, X. C., Konduri, M., Fomina, N., Lu, J., Jin, L., Kolykhalov, A. and Tan, S.-L. (2006), Mechanistic link between the anti-HCV effect of interferon gamma and control of viral replication by a ras-MAPK signaling cascade. Hepatology, 43: 81–90. doi: 10.1002/hep.21011
Potential conflict of interest: Nothing to report.
- Issue published online: 22 DEC 2005
- Article first published online: 22 DEC 2005
- Manuscript Accepted: 24 OCT 2005
- Manuscript Received: 21 JUL 2005
Interferon-gamma (IFN-γ) exerts potent antiviral activity in the hepatitis C virus (HCV) replicon systems. However, the mechanisms underlying the direct antiviral effect have not been determined. We found that the type II transcriptional response to IFN-γ could be suppressed by inhibition of MEK1/2 kinase activity by MEK1/2 inhibitor U0126 in the hepatoma cell line Huh-7. Using a bicistronic HCV replicon system expressing a luciferase reporter gene in Huh-7 cells (RLuc-replicon), we showed that inhibition of MEK1/2 kinase activity is sufficient to counteract the antiviral activity of IFN-γ. Expression of a constitutive active form of Ras inhibited the luciferase activity of RLuc-replicon, whereas a dominant-negative mutant of Ras enhanced the reporter activity, indicating that the Ras-MAPK pathway has a role in limiting replication of the viral RNA. Consistent with the involvement of the Ras-MAPK pathway, treatment with epidermal growth factor suppressed HCV protein expression in the RLuc-replicon cells, an effect that could be abolished by U0126. Inhibition of MEK1/2 kinase activity correlated with reduced phosphorylation of the HCV NS5A protein and enhanced RLuc-replicon luciferase reporter activity, in line with recent reports that phosphorylation of NS5A negatively modulates HCV RNA replication. Finally, genetic deletion analysis in yeast supported the role of a MEK-like kinase(s) in the regulation of NS5A phosphorylation. In conclusion, the direct anti-HCV effect of IFN-γ in cell culture is, at least in part, mediated through the Ras-MAPK signaling pathway, which possibly involves a direct or indirect modulation of NS5A protein phosphorylation. (HEPATOLOGY 2006;43:81–90.)
Chronic infection with hepatitis C virus (HCV) is associated with cirrhosis that often leads to hepatic failure and hepatocellular carcinoma.1 With over 170 million people chronically infected with HCV and 3 to 4 million new infections occurring each year, HCV infection represents a serious global medical, social, and economic burden.2 The situation is worsened by the fact that currently approved anti-HCV therapies, including pegylated interferon-alpha (IFN-α) alone or in combination with ribavirin, are not effective in a substantial fraction of patients and plagued with many undesirable side effects.3 Thus, a pressing need exists for improvement of current anti-HCV therapies or a novel therapeutic approach in the form of new antivirals or immunomodulation agents.
Despite significant advances in understanding HCV at the biochemical and molecular level in recent years,4 the development of effective anti-HCV therapeutics has been stymied owing in part to the lack of a readily available small animal model and a cell culture system for propagating infectious HCV. Recently, human hepatoma-derived Huh-7 cell lines supporting replication of an engineered copy of subgenomic HCV RNA containing the neomycin phosphotransferase gene (NEO) in place of the viral structural genes were established.5 Subsequent variants of the so-called HCV replicon system6–9 have afforded the study of HCV RNA translation and replication, as well as evaluation of candidate anti-HCV agents.10 Initial validation of HCV replicon systems for anti-HCV screening relied on the use of recombinant human IFN-α,5–7, 11 IFN-β, and IFN-γ.11, 12 The HCV replicon system also was used to show that IFN-α and IFN-γ use alternate antiviral mechanisms to inhibit HCV.11, 12 Consistent with this notion, type I/II IFN combinations resulted in a synergistic antiviral effect in the HCV replicon system, whereas type I/I combinations produced an additive profile.13
The therapeutic efficacy of IFN against HCV in vivo is attributable to its effects on infected cells both directly, by inducing an antiviral state via downstream mediators, and indirectly, by stimulating immune cells. In the HCV replicon system, various mechanisms have been proposed for the direct antiviral effects of IFN-α, including activation of STAT proteins14–16 and the p38-MK2 signaling pathway,17 a preferential inhibition of viral translation through PKR and p56,18 and double-stranded RNA-specific editing of adenosine residues by ADAR1.19 However, little is known of the molecular mechanisms by which IFN-γ exerts its direct antiviral function. In this study, we provide evidence suggesting that the direct antiviral effect of IFN-γ in the HCV replicon system is at least in part mediated through a Ras-MAPK signaling pathway, which possibly affects phosphorylation of the viral NS5A protein.
Materials and Methods
Huh-7 cells were cultured at 37°C in Dulbecco's modified Eagle medium containing 10% inactivated fetal bovine serum, 2 mmol/L glutamine, 100 U/mL penicillin, and 100 μg/mL streptomycin under 5% CO2 conditions. Subgenomic HCV replicon cells carrying a Renilla luciferase reporter (RLuc) were maintained in the same medium with 0.25 mg/mL G418. A stable cell line with the Renilla luciferase gene integrated into the genome of Huh-7 cells, termed genome Renilla control cell line (Gen-RLuc), was maintained in Dulbecco's modified Eagle medium. Cell culture medium, supplements, and antibiotics were all purchased from Invitrogen (Carlsbad, CA).
Antibodies and Reagents.
Anti-HCV NS5A and -NS3 antibodies were from ViroStat (Portland, ME). Anti-HCV NS5B antibody was from Santa Cruz Biotechnology (Santa Cruz, CA). Antibodies to Ras (L2 region), MEK1, Raf, phosphorylated (S218/222) MEK1/(S222/226) MEK2, phosphorylated (S383) Rac, and phosphorylated (Y705) ERK1/2 were from Upstate Biotechnology (Lake Placid, NY). Anti-ERK antibody was from Stressgen Biotechnologies (Victoria, BC Canada). Anti-β-actin antibody was from Abcam Inc. (Cambridge, MA). Phospho-specific NS5A antiserum (3pNS5A) was raised against PPpSLASSpSApSQ, where pS denotes a phospho-serine. Phospho-peptide synthesized by Dr. Michael Berne and colleagues (Tufts University peptide core facility) was coupled to keyhole limpet hemocyanin and injected into rabbits at Zymed Laboratories (South San Francisco, CA).
JAK inhibitor (Cat. No. 420099-500UG), SP600125 (Cat. No. 420119), SB203580 (Cat. No. 559389), wortmannin (Cat. No. 681675), rapamycin (Cat. No. 553210), U0126 (Cat. No. 662005), U0124 (Cat. No. 662006), SL327 (Cat. No. 444939), Gö6976 (Cat. No. 365250), rottlerin (Cat. No. 557370), daidzein (Cat. No. 251600), PP2 (Cat. No. 529573), H89 (Cat. No. 173962), and PKR inhibitor (Cat. No. 527450) were purchased from Calbiochem (San Diego, CA). All compounds were diluted in dimethylsulfoxide (DMSO). Recombinant IFNs were purchased from PBL Biomedical Lab (Piscataway, NJ). Epidermal growth factor (EGF) was obtained from BD Biosciences.
Construction of Luciferase Reporter Cell Lines.
A subgenomic HCV replicon was synthesized according to Lohmann et al.5 Cell culture adaptive mutations5, 6 were introduced into the original replicon sequence, and the resulting replicon sequences were tested in a colony formation assay in Huh-7 cells. A variant, which contained 2 mutations (E1202G and S2204I), generated approximately 10 times more colonies than the next most potent variant (S2204I alone) and over 1,000 times more colonies than any other variants tested. This variant was used to generate a bicistronic HCV subgenomic replicon with a RLuc gene. For this purpose, the entire RLuc open reading frame (ORF) was amplified from the phRL-TK plasmid (Promega, Madison, WI) with primers 5′ CACCAACGGGCGCGCCATGGCTTCCAAGGTGTACGAC and 5′CATGGACGCGTTGCTCGTTCTTCAGCACGCGCTC. The resulting DNA fragment was digested with AscI and MluI, and cloned into the AscI site of the E1202G+S2204 replicon plasmid such that the translation of the RLuc ORF is controlled by the 5′NTR (non-translated region) of HCV. The resulting plasmid contains an in-frame fusion of the RLuc ORF with the coding sequence of Neo (Fig. 1). In vitro RNA transcripts from this plasmid were electroporated into Huh-7 cells, and G418-resistant colonies were selected. Established clones (RLuc-replicon) were expanded, and the RLuc-replicon-containing cells were frozen at passage 10. Two independent clones (# 10 and 42) were used in this study. TaqMan real-time quantitative reverse transcription polymerase chain reaction and data analysis, performed as described previously,13 showed 100 to 1,000 copies of HCV RNA per cell. The numbers fluctuated from passage to passage but did not decrease up to passage 50. The sequence of the replicon RNA from cells up to passage 30 showed no additional mutations in the entire HCV part but had mutations in EMCV IRES region (at positions 2250 and 2251 deleted from CTATATG to CTATG; at position 2519 change from GGCGG to GGTGG; and at position 2645 change from TCTGG to TCAGG). All experiments were performed using RLuc-replicon cells from passages below 20. Correlation between the amount of the replicon RNA and luciferase activity or RLuc RNA was established using various recombinant IFNs and TaqMan real-time quantitative reverse transcription polymerase chain reaction.13
A genome Renilla control cell line carrying an RLuc reporter integrated into the genome was generated by co-transfecting linearized pRL-SV40 (Promega) and pPur (Clontech, Mountain View, CA) plasmids into Huh-7 cells and selection of puromycin-resistant and RLuc-positive clones. Established clones (Gen-RLuc) were expanded, and the Gen-RLuc–containing cells were frozen at passage 10. Gen-RLuc clone #10 was used in this study.
Compound Testing in the RLuc-Replicon Cell Line.
RLuc-replicon cells were seeded at 5 × 104 cells per well in 96-well plates in medium without G418. After overnight incubation, DMSO-dissolved compounds were added to the medium at 0.001 to 20 μmol/L at 1:3 serial dilutions in duplicate samples. After incubation for 48 hours, luciferase activity was determined by using Promega Renilla Luciferase Assay System according to the manufacturer's instructions in a Perkin-Elmer Victor2 Multi-label Reader (Boston, MA).
Transient Transfection and Luciferase Assays.
Transient transfections were performed using Roche Fugene 6 Transfection Kit (Indianapolis, IN) according to the manufacturer's instructions. To evaluate the effects of compounds on IFN response, Huh-7 cells were transfected with pGAS-TA-Luc or pISRE-TA-Luc (Promega) and were seeded onto 96-well plates 24 hours post-transfection. Transfected cells were incubated with indicated concentrations of IFN for 16 hours, either with or without pretreatment with the indicated compound for 4 hours. The effect of the IFN treatment was measured with the Bright-Glo luciferase assay (Promega) according to the manufacturer's instructions.
To evaluate the effects of MEK1/2 activation or inhibition, RLuc-replicon cells were co-transfected with the constitutively active (Q61L) or dominant-negative (S17N) form of H-Ras plasmid (Upstate Biotechnology), respectively, and pGL2-Luc (Promega). pGL2-Luc containing a firefly luciferase (FLuc) gene driven by the HSV promoter was used to normalize for the transfection efficiency. Forty-eight hours post-transfection, cells were harvested for FLuc and RLuc activity measurement using Promega Dual-Luciferase Reporter Assay System according to the manufacturer's recommendations in a Victor2 Multi-label Reader (Perkin-Elmer). The RLuc activity in each sample was normalized to FLuc activity.
NS5A Expression in Yeast.
Yeast PBS2, STE7, HOG1, PAK1, and IRE1 deletion mutants and parental strain BY4742 (MATα his3Δ1 leu2Δ0 lys2Δ0 ura3Δ0) were obtained from the former Research Genetics (now Invitrogen) and maintained according to the company's instructions. Yeast strains were transfected with p426-NS5A (provided by Kirk Staschke, Eli Lilly and Company, Indianapolis, IN), using the Bio 101 EZ-Yeast Transformation Kit (now Qbiogene, Morgan Irvine, CA) according to the manufacturer's instructions. Protein lysates were prepared using Bio 101 FastPROTEIN RED and FastPrep Instrument.
Western Blot Analysis.
Cells were lysed using a buffer (20 mmol/L Tris-HCl, pH 7.4, 5 mmol/L EDTA, 50 mmol/L NaCl, 10 mmol/L Na-Pyrophosphate, 50 mmol/L NaF, 1% NP-40, 1 mmol/L Na3VO4) containing the Roche Protease Inhibitor Cocktail Tablets (Indianapolis, IN). After centrifugation (10,000 rpm) at 4°C for 5 minutes, supernatants were collected and protein concentrations determined by the Pierce BCA Protein Assay Kit (Rockford, IL). Cell lysates were subjected to 4%–12% Bis-Tris gel (Invitrogen). After electrophoresis, proteins were transferred onto polyvinylidene fluoride membranes (Invitrogen) and incubated with the indicated antibodies. Signals were detected using the Pierce Femto Supersensitive Western Kit.
Comparison of luciferase activities between the various treatment groups was analyzed by unpaired Student t test. Values of P < .05 were considered significant.
IFN-γ has a potent inhibitory effect on the RNA replication of HCV subgenomic replicon in Huh-7 cells.11–13 Although pharmacological inhibition of the JAK-STAT pathway inhibited transcriptional response to IFN-γ (Table 1),55–59 it did not relieve the inhibitory effects mediated by IFN-γ in the HCV replicon system (data not shown), suggesting that the anti-HCV effect of IFN-γ is mediated through another pathway. To determine the mechanisms mediating the IFN-γ antiviral effect in Huh7 cells, we first examined the effect of pharmacological inhibition of other signaling pathways that have been reported to be activated by IFNs (IFN-α or IFN-γ). Huh-7 cells transiently transfected with a GAS-dependent luciferase reporter construct were pre-incubated in the presence or absence of kinase-specific inhibitor before they were treated with IFN-γ. Total cell lysates were subsequently analyzed for luciferase activity to assess IFN response. We found that treatment of Huh-7 cells with a mitogen-activated/extracellular response protein kinase (MEK)1/2-specific inhibitor U012620 significantly reduced IFN-γ-dependent transcriptional activation (Table 1). In contrast, inhibition of the kinase activity of PKR, PI-3K, mTOR, P38, JNK, classical PKC, or PKCδ did not significantly affect IFN-γ response in this system. Interestingly, U0126 had no effect on type I IFN response as judged by insensitivity of transcriptional response from an ISRE-driven luciferase reporter construct to IFN-α treatment (data not shown). These results suggest that MEK1/2- or a MEK1/2-dependent pathway is selectively involved in mediating IFN-γ-induced response in Huh-7 cells.
|Inhibitor (10 μmol/L)||Target||Inhibition of IFN-γ Response||References|
|JAK inhibitor||JAK||+||Briscoe et al.55|
|PKR inhibitor||PKR||−||Chu et al.32; Deb et al.33|
|Wortmannin||PI3K||−||Nguyen et al.34|
|Rapamycin||mTOR||−||Kristof et al.35; Parmar et al.36|
|SB203580||P38||−||Uddin et al.37; Li et al.38|
|SP600125||JNK||−||Chu et al.32; Takada et al.56|
|U0126||MEK1/2||+||Xia et al.57; Sakatsume et al.58|
|Rottlerin||PKCδ||−||Deb et al.59|
|Gö6976||cPKC||−||Fimia et al.15|
We next examined whether inhibition of MEK1/2 activity would be sufficient to counteract the antiviral effect of IFN-γ in an HCV subgenomic luciferase-replicon reporter system (Fig. 1). In this bicistronic system, the RLuc reporter gene is expressed by the HCV 5′NTR, whereas the viral proteins are translated by means of the encephalomyocarditis virus (EMCV) internal ribosomal entry site (IRES). The RLuc-replicon system was validated using various recombinant IFNs; the RLuc-replicon luciferase activities correlated well with the viral RNA levels (Fig. 2A), as reported for other similar reporter-based replicon systems.8, 21, 22 Consistent with the reported concentration that inhibits 50% for IFN-γ in reducing the HCV subgenomic replicon RNA levels,13 treatment with 1 U/mL IFN-γ resulted in 50% inhibition of the RLuc activity in Huh-7 cells (Fig. 2B). Importantly, the inhibitory effect was reversible by pretreatment with U0126 in a concentration-dependent manner. At the highest concentration tested (10 μmol/L), inhibition of MEK1/2 kinase activity actually led to an enhanced RLuc activity. These results support the notion that IFN-γ exerts its antiviral effect against HCV through MEK1/2 or a MEK1/2-dependent pathway in Huh-7 cells.
To further test our hypothesis and establish the putative role of an MEK1/2-dependent pathway in negative regulation of HCV replication, we performed studies in which endogenous MEK1/2 in the RLuc-replicon system was either activated or blocked. For this purpose, a constitutive active (CA) or a dominant-negative (DN) mutant Ras was used, respectively. Transfection of CA, but not DN, Ras in RLuc-replicon resulted in significant down-regulation of RLuc activity (Fig. 3A), suggesting that activation of a Ras-MAPK pathway negatively modulates HCV replication. Transient expression of CA Ras activated MEK1/2 as revealed by ERK1/2 phosphorylation, but had no effect on the expression of total ERK1/2 proteins (Fig. 3B). Interestingly, introduction of DN Ras led to a significant increase in RLuc activity in a concentration-dependent manner (Fig. 3A), suggesting that inhibition of basally activated Ras-MAPK pathway is sufficient to promote HCV replication in the cell.
Based on these results, we reasoned that inhibition of endogenous MEK1/2 activity would be sufficient to promote HCV replication in the cell. Indeed, inhibition of MEK1/2 by U0126 increased RLuc-replicon activity (Fig. 4A). The effect of U0126 was specific to RLuc-replicon because the compound had no effect on the RLuc activity of a stable Huh-7 cell line, Gen-RLuc, carrying identical RLuc reporter (Fig. 4A).
Furthermore, similar stimulatory effect on RLuc-replicon activity was observed when the RLuc-replicon cells were treated with a structurally unrelated MEK1/2 inhibitor (SL327), indicating that the effect of U0126 is not scaffold-specific (Fig. 4B). To exclude potential clonal artifacts, we tested and found that U0126 also stimulated RLuc-replicon activity in an independent clone (Fig. 4C). As a further control, U0124, an inactive analog of the MEK1/2 inhibitor, did not enhance RLuc-replicon activity, strongly supporting a specific involvement of MEK1/2. Inhibition of other cellular protein kinases whose kinase activity has been suggested to be regulated by HCV, such as PI3K and Src, did not have any effect on the RLuc-replicon (Table 2).60–64 Interestingly, inhibition of JNK kinase activity had a negative impact on RLuc-replicon activity, suggesting that JNK and/or other protein kinase(s) inhibited by SP600125 may play a positive role in HCV replication. Taken together, these data suggest that a Ras-MAPK pathway has a specific role in limiting replication of the viral RNA.
|Inhibitor (10 μmol/L)||Target||Inhibition of RLuc-Replicon Activity||References|
|JAK inhibitor||JAK||−||Briscoe et al.55|
|PKR inhibitor||PKR||−||Gale et al.60; Tan et al.45|
|Wortmannin||PI3K||−||He et al.61; Street et al.62|
|Rapamycin||mTOR||−||Coito et al.25|
|SB203580||P38||−||He et al.63|
|SP600125||JNK||+||Park et al.64|
|Daidzein||II||−||Kim et al.51|
|PP2||SRC||−||McDonald et al.46|
|H89||PKA||−||Ide et al.50|
Next, we determined whether activation of the endogenous Ras-MAPK pathway in the HCV replicon system by a well-known stimulus of the Ras-MAPK pathway, namely, EGF, would also lead to inhibition of HCV replication. We first confirmed that EGF treatment could activate the Ras-MAPK pathway in the HCV replicon cells, resulting in phosphorylation of MEK1/2, Raf-1 and ERK1/2 (Fig. 5A, lane 1 vs. 2). Remarkably, the levels of HCV NS5A protein were significantly reduced within just 15 minutes of EGF treatment. Importantly, this effect was abolished when a parallel experiment was performed in the presence of U0126 (Fig. 5A, lane 2 vs. 3). Furthermore, U0126 treatment alone led to a slightly but noticeable enhanced expression of NS5A (Fig. 5A, lane 1 vs. 4), but not other viral NS proteins (data not shown). Longer treatment time with EGF (16 hours) led to significantly reduced levels of HCV proteins, namely, NS3 and NS5B (Fig. 5B, lane 1 vs. 2), but had no effect on MEK1 protein level. There was no detectable phosphorylated ERK and MEK after 16 hours because the EGF signaling is down-regulated after this point (data not shown). These effects were not seen when the experiments were done in the presence of U0126. The inhibition of NS5A, NS3, and NS5B proteins seen is unlikely to be attributable to a nonspecific effect on the EMCV IRES, which is driving the expression of these proteins, as inhibition of the Ras-MAPK pathway did not affect EMCV IRES-driven luciferase activity in Huh-7 cells (data not shown). Furthermore, previous studies have shown that EMCV IRES-mediated translation of viral transcripts is regulated by the p38 pathway, not by the Ras-MAPK pathway.23 EGF treatment did not induce IFN-γ-mediated transcriptional response, which could potentially offer an alternative explanation for its anti-HCV effect (Fig. 5C). Altogether, these results support a role for the Ras-MEK pathway in directly negatively modulating HCV replication in the replicon system.
NS5A is a key viral protein involved in the regulation of HCV replication whose function may be controlled by protein phosphorylation by cellular protein kinases.24 One of the cellular kinases that have been shown to phosphorylate NS5A, at least in vitro, is MEK1.25 To begin to identify the mechanisms by which activation of a MEK1/2-dependent pathway may be regulating HCV replication, we examined whether phosphorylation of NS5A is affected in the HCV replicon system. To this end, we generated a phospho-specific polyclonal antibody to the NS5A-derived phospho-peptide PP-pS-LASS-pS-A-pS, where pS indicates potential phosphorylation serine residues 2197, 2202, and 2204 on NS5A that were mutated in adapted HCV replicons (reviewed in Bartenschlager et al.26). We first characterized the phospho-NS5A antibody using lysates prepared from different kinase deletion mutant and parental yeast strains expressing wild-type NS5A. As shown in Fig. 6A, phosphorylated NS5A forms were readily detected in parental strain BY4742 (lane 3), but not in lysates treated with λ phosphatase (lane 4). There was no detectable phosphorylated NS5A in BY4742 cells transfected with vector control (lane 8). Importantly, gene deletion of yeast protein kinase PBS2 or STE7, which are highly related to human MEK kinases, led to a marked loss in NS5A phosphorylation (lanes 1 and 2). In contrast, phosphorylated NS5A proteins were detected in yeast strains devoid of other yeast protein kinases unrelated to MEK (lanes 5-7).
In line with recent reports that phosphorylation of NS5A negatively modulates HCV replication,27 the enhanced replication of HCV subgenomic replicon resulted from inhibition of MEK1/2 activity by U0126 corresponded with reduced NS5A phosphorylation in Huh-7 cells (Fig. 6B). We could barely detect the shifted form of hyperphosphorylated NS5A, p58, in our studies (Fig. 5 and 6B), which is consistent with previous studies that replicons containing the S1179I (S2204) point mutation have significantly less of the p58 form.6 U0126 did not appreciably affect NS5A protein levels. The role of MEK1/2 kinase activity on HCV replication is specific because inhibition of other protein kinases previously shown to interact/or phosphorylate NS5A, namely, PKR, mTOR, and CKII, did not significantly inhibit RLuc-replicon (Table 2). These data suggest that a Ras-MAPK pathway may limit HCV replication at least in part by either directly or indirectly modulating of NS5A protein phosphorylation, possibly through MEK1/2.
IFNs exert their in vivo biological effects predominantly through the Jak-STAT pathway, which regulates gene transcription for protein products that mediate antiproliferative, antiviral, and immunomodulatory responses.28 In the chimpanzee model of HCV infection, production of IFN-γ by immune effector cells in the liver is associated with viral clearance,29, 30 most likely attributable to natural killer cell homing and local secretion of IFN-γ in the HCV-infected liver.31 The direct antiviral effect of IFN-γ against HCV RNA replication has been demonstrated in cell culture,11, 12 but the mechanisms have not been elucidated. Here, we provide evidence that the direct anti-HCV effect of IFN-γ is at least in part mediated through activation of a Ras-MAPK pathway. We first confirmed that the kinase function of Jak is required for type II IFN response (Table 1). In contrast to previous findings,32–38 however, the kinase activity of PKR, PI3-K, mTOR, or p38 is dispensable for type II IFN response, at least in Huh-7 cells. Consistent with previous reports demonstrating that a cross-talk is present between the Jak-STAT and Ras-MAPK pathways (reviewed by Platanias39), we found that the MEK1/2 kinase activity is required for optimal type II response in Huh-7 cells (Table 1). However, inhibition of JAK enzymatic activity did not relieve the antiviral effect of IFN-γ in the HCV replicon system. On the contrary, our results indicate that MEK1/2 activity is required for the antiviral effect of IFN-γ in the replicon system. It will be interesting to examine whether Ras-MAPK inhibition could also prevent the inhibitory activity of IFN-γ against other viruses in future studies.
Inhibition of the basally activated Ras-MAPK pathway was sufficient to promote HCV replication in the replicon system (Fig. 4), suggesting that the Ras-MAPK pathway normally plays a negative role in controlling HCV replication. Indeed, activation of the Ras-MAPK pathway by either EGF or overexpression of activated Ras suppressed HCV replicon RNA replication (Figs. 3, 5). Although previous studies did not observe a loss in NS5A protein level upon EGF treatment,40, 41 another study recently showed that IL-1 also inhibited HCV subgenomic RNA replication and NS5A levels by activation of the Ras-MAPK pathway in an independent replicon clone.42 HBV replication also can be suppressed by activation of the Ras-MAPK pathway, although the mechanism has not been defined.43 Although the precise mechanisms by which activation of the Ras-MAPK pathway counteracts HCV replication remain to be determined, our results suggest that one potential means may involve a direct or indirect modulation of NS5A protein phosphorylation (Fig. 6). Activation of a Ras-MAPK pathway by external stimuli, including IFN-γ, IL-1 and EGF, may increase NS5A phosphorylation and subsequent inhibition of HCV replication. In this regard, it is noteworthy that HCV has been shown to possess mechanisms to down-modulate the MAPK signaling pathway,40, 44 which includes NS5A interaction with upstream adaptor protein Grb245 and Src tyrosine kinases.46 It is tempting to speculate that non- or basally phosphorylated NS5A functions to keep MAPK activation in check by binding to Grb2 or Src kinases.
HCV NS5A was initially implicated in the viral subversion of host IFN-induced antiviral response,47 but recent studies using the HCV replicon cell culture systems have provided an emerging role for the differential phosphorylation of HCV NS5A in the regulation of virus replication.27, 48, 49 Previous reports have shown that NS5A is tightly complexed with a cellular kinase(s).50–52 An initial biochemical analysis of the effects of a panel of kinase inhibitors on NS5A phosphorylation in vivo and in vitro has suggested that the kinase responsible for most NS5A phosphorylation may be a member of the CMGC family, which includes CKII and proline-directed kinases such as the cyclin-dependent kinases, MAPKs, and glycogen synthase kinase 3.53 Although CKII was demonstrated to be capable of phosphorylating NS5A in vitro,50–52 we found that inhibition of CKII kinase activity does not appear to affect viral replication in the HCV replicon system (Table 2). A more recent study demonstrated that several other kinases, including MEK1, MEK6, and MEK7, were also capable of phosphorylating NS5A in vitro,25 supporting our hypothesis that a protein kinase(s) in the Ras-MAPK pathway is involved in regulating HCV replication, although we have not ruled out the potential role of the other kinases reported in that study.
Inhibition of MEK1/2 kinase activity did not lead to complete abrogation of NS5A phosphorylation (Fig. 6). Basal phosphorylation of NS5A by another protein kinase(s) could also play a positive role in HCV replication. In this regard, inhibition of JNK kinase activity had a negative impact on RLuc-replicon activity, suggesting that JNK and/or other protein kinase(s) inhibited by SP600125 may play a positive role in HCV replication. Interestingly, activation of ERK has been reported to phosphorylate and activate MAPK phosphatase-7 (MKP-7), which is a JNK-specific phosphatase,54 offering a potential alternative mechanism by which MAPK activation could regulate HCV replication. Further studies will be necessary to test this hypothesis to determine the role of JNK in HCV replication.
In conclusion, we provide evidence suggesting that the direct antiviral effect of IFN-γ in the HCV replicon system is, at least in part, mediated through a Ras-MAPK signaling pathway. In addition, our data provide potential new insights into the mechanisms of host control of HCV replication, through Ras-MAPK-dependent modulation of phosphorylation of the viral NS5A protein. Further studies to ascertain the protein kinases and phosphatases that are responsible for regulating NS5A phosphorylation may provide new avenues for therapeutic interventions for HCV infection.
The authors thank Lifei Liu and Cong Ping Xie for technical assistance with TaqMan quantitative reverse transcription polymerase chain reaction analysis, Kirk Staschke for p426-NS5A plasmid, and Tina M. Myers for critical review of this manuscript.