|Home | About | Journals | Submit | Contact Us | Français|
Hepatitis C virus (HCV) entry is a complicated process that requires multiple host factors, such as CD81, scavenger receptor BI, claudin-1 (CLDN1), and occludin. The interaction of virus and cellular entry factors represents a promising target for novel anti-HCV drug development. In this study, we sought to identify peptide inhibitors for HCV entry by screening a library of overlapping peptides covering the four above-mentioned entry factors. An 18–amino acid peptide (designated as CL58) that was derived from the CLDN1 intracellular and first transmembrane region inhibited both de novo and established HCV infection in vitro. Unlike previously reported peptides corresponding to CLDN1 extracellular loops, CL58 did not alter the normal distribution of CLDN1 and was not cytotoxic in vitro at concentrations nearly 100-fold higher than the effective antiviral dose. The inhibitory effect of CL58 appeared to occur at a late step during viral entry, presumably after initial binding. Finally, overexpressed CL58 was able to interact with HCV envelope proteins.
We identified a novel CLDN1-derived peptide that inhibits HCV entry at a postbinding step. The findings expand our knowledge of the roles that CLDN1 play in HCV entry and highlight the potential for developing a new class of inhibitors targeting the viral entry process.
Hepatitis C virus (HCV) is an important human pathogen that infects more than 170 million people worldwide. Chronic infection of HCV causes severe liver disease, including hepatic cirrhosis and hepatocellular carcinoma.1,2 Despite the recent approval of boceprevir and telaprevir by the US Food and Drug Administration, successful treatment of HCV is expected to involve combination therapy with multiple inhibitors of different targets.3,4 Therefore, new antiviral drugs are urgently needed to treat HCV infection independently or in combination with current therapies.
Recent studies have demonstrated that HCV uses at least four cellular membrane proteins to gain entry: CD81, scavenger receptor B1 (SR-BI), claudin-1 (CLDN1), and occludin (OCLN).5–8 It has been postulated that infectious virions complete binding, endocytosis, and fusion processes through sequential interactions with SR-BI and CD81 earlier in the entry pathway, whereas two tight junction (TJ) proteins CLDN1 and OCLN play important roles during a postbinding step of HCV entry.9 With these advances in the field, researchers have an unprecedented opportunity to develop novel HCV inhibitors that target the entry process. Here we report the discovery of a novel peptide inhibitor derived from the N terminus of human CLDN1 that inhibits virus entry in a postbinding step.
The detailed information of cells, reagents, peptides, production of HCV pseudoviral particles (HCVpp) and cell culture–grown HCV (HCVcc), HCVpp infection assay, cytotoxicity assay, In-Cell Western assay (LI-COR), HCVcc binding assay, velocity sedimentation ultracentrifugation, time-of-addition assays, immunofluorescence staining, confocal microscopy, western blotting, and immunoprecipitation is described in the Supporting Information. For statistical analysis, bar graphs were plotted to show the mean ± SD of at least three independent experiments. Statistical analyses were performed using SigmaPlot 10 and Graphpad Prism 5. A P value <0.05 using the Student t test was considered statistically significant.
We initially postulated that peptides derived from host entry factors located on the cell surface may compete for incoming virions and hence block HCV entry. To test this hypothesis, we designed a peptide library of 121 overlapping peptides comprised of 18-mer peptides offset by 13 amino acids (aa) that covered the entire protein sequences of human CD81, SR-BI, CLDN1, and OCLN for the ability to inhibit HCVpp infection of Huh7.5.1 cells (Fig. 1). Thirty-two peptides were abandoned during the screening due to poor solubility. Among the remaining 89 peptides (sequences listed in Supporting Table 1), two overlapping peptides derived from the CLDN1 N terminus, CL58 (MANAGLQLLGFILAFLGW) and CL59 (AFLGWIGAIVSTALPQWR), inhibited HCVpp entry more than 80% at 50 μM (Fig. 1). Of note, other peptides in the library either failed to exert any effect or had only a marginal effect (± 2.5-fold).
Both CL58 and CL59 are derived from the first transmembrane region of CLDN1, but CL58 is more potent than CL59 in inhibiting HCV entry and was therefore selected for further analyses. In order to determine the length and sequence of CL58 for maximal inhibition, we first synthesized eight additional 18-mer peptides with a 3-aa offset (CL58.1–8). These peptides extended further into the first transmembrane and extracellular loop (EL) region of CLDN1. When tested, none of these peptides exerted inhibition to the same extent as the parental CL58 (Table 1). Next, we altered the length of the peptide by removing residues from or adding residues to the CL58 C terminus. We found that shortening the peptide by 2 or 4 aa drastically increased the 50% cell culture inhibitory concentration (IC50), and extending the peptide by 2, 4, or 6 aa slightly increased the IC50 as well. Lastly, a D-isomer of CL58 displayed a slightly lower IC50 (Table 1). Thus, CL58 appears to contain the essential length and sequence needed to inhibit HCVpp entry. In support, a scrambled peptide failed to inhibit HCV entry (Fig. 1).
The IC50 of CL58 was determined using HCV genotype 2a isolate from a patient with fulminant hepatitis in Japan (JFH-1) HCVcc and HCVpp carrying envelope proteins derived from major HCV genotypes. The IC50 of CL58 was 2 μM using JFH-1 HCVcc (Fig. 2A) and ranged from 0.6 to 5 μM using HCVpp, depending on the genotypic origin of the envelope proteins (Fig. 2B). CL58 did not inhibit entry of VSV-G–pseudotyped lentivirus (Fig. 2B) or group B coxsackievirus infection (Supporting Fig. 1). Importantly, CL58 was able to suppress HCVpp entry into primary human hepatocytes (Fig. 2C). In addition, CL58 inhibited HCVcc infection at multiple multiplicities of infection (MOIs) (Fig. 2D). To rule out any confounding effect due to cytotoxicity, the 50% cytotoxic concentration of CL58 was also determined and was estimated to be almost 100-fold higher than its IC50 (Fig. 2E). When used in combination with other known inhibitors, CL58 showed additive effect with interferon and cyclosporin A, but not with 2'-C-methylcytidine (Fig. 2F).
To determine whether CL58 suppresses persistent HCV infection, CL58 was added to hepatoma cells that have been inoculated with a very low amount of HCVcc (MOI 0.01) and the peptide was retained in medium during the entire treatment. As shown in Fig. 3, CL58 significantly suppressed the expansion of the virus in vitro. To explore the potential effect of CL58 on HCV RNA replication and release, we added CL58 to a HCV replicon cell line harboring a full-length genotype 1b genome or Huh7.5.1 cells that have been fully infected with JFH-1. In either case, the intracellular HCV RNA level or the supernatant viral RNA level was not altered by CL58 (Supporting Figs. 2 and 3), suggesting that CL58 does not inhibit postentry steps of HCV.
To gain more insight into the mechanism of CL58-dependent inhibition, we first sought to define the step of the HCV life cycle upon which CL58 acts. It was observed that CL58 inhibited infection when added to the cells together with the virus, but not when added 4 hours before or after infection (Fig. 4A). These results confirmed that CL58 blocked viral entry. To rule out the possibility that CL58 directly inactivated the virus, the concentrated HCVcc particles were treated with dimethyl sulfoxide (DMSO, vehicle) or 8 μM CL58 for 2 hours at 37°C and then loaded onto a 10%–50% sucrose gradient for rate zonal ultracentrifugation. Each fraction was weighed and then analyzed for HCV RNA by quantitative reverse-transcription polymerase chain reaction (qRT-PCR). It was observed that the DMSO and CL58-treated groups displayed similar profiles of peaks of viral RNA and nearly identical density in each fraction, suggesting CL58 did not disrupt the structural integrity of HCV (Fig. 4B). Alternatively, viruses pretreated with DMSO or CL58 were subjected to ultracentrifugation in order to remove the peptide, and purified viruses were then used to infect Huh7.5.1 cells. We found that DMSO-and CL58-treated viruses remained equally infectious (Fig. 4C). Together, these data suggest that CL58 was not directly virocidal to HCV.
Subsequently, we assessed the effect of CL58 on virus binding, and found that HCVcc binding to Huh7.5.1 cells was not affected by CL58 (Fig. 4D). To explore whether CL58 acts after virus binding, we synchronized the infection of cells by incubating virus with cells at 4° C for 2 hours followed by a temperature shift to 37° C and then added CL58 at different time points relative to the temperature shift to 37° C. As a positive control, we also used the anti-CD81 monoclonal antibody JS-81, which is known to block HCV entry at a post-binding stage.10 As shown in Fig. 5A, CL58 inhibition of HCV entry exhibited time dependence and greater than 50% sensitivity was achieved even when CL58 was added 1 hour after the temperature shift to 37° C, indicating that CL58 acts after initial viral attachment. Moreover, anti-CD81 and CL58 exhibited additive effect when added together (Fig. 5A).
In addition to CLDN1, CLDN6 and CLDN9 have been demonstrated to render 293T cells susceptible to HCVpp infection,11,12 whereas CLDN7, the CLDN family member that shares the highest homology with CLDN1, failed to do so.12 For these reasons, we synthesized 18-aa peptides derived from corresponding regions of CLDN6, CLDN7, and CLDN9. Direct comparison revealed that only CL58, but not counterpart peptides derived from the aforementioned CLDNs, was able to inhibit HCVcc infection (Fig. 5B). These results reinforce the idea that anti-HCV activity is a unique property of CL58. To determine the effect of CL58 on HCV envelope protein-mediated membrane fusion, we set up a cell-cell fusion assay in which 293T acceptor cells, containing a loxP-flanked STOP cassette that blocks transcription of the downstream luciferase reporter gene, were cocultured with Cre-expressing donor cells. In this assay, fusion between donor and acceptor cell membranes removes the STOP cassette and hence permits luciferase production. Shown in Fig. 5C, coculturing donor cells expressing HCV E1E2 with CLDN1-expressing acceptor cells resulted in a more than 10-fold increase in luciferase counts, indicating a successful achievement of cell-cell fusion between donor and receipt cells (Fig. 5C). The addition of CL58 consistently reduced the cell-cell fusion by more than 50%. By contrast, peptides derived from CLDN6, CLDN7, and CLDN9 failed to exert any effect (Fig. 5C). To gain insight into whether CL58 directly participates in the fusion process or acts at an earlier step that is required to enable subsequent fusion, we compared the inhibitory kinetics of CL58 with that of bafilomycin A1, which inhibits the final fusion, and found that sensitivity of HCVcc to CL58 declined slightly ahead of that of bafilomycin A1 (Fig. 5D). Altogether, these results support a model that CL58 interferes with a process that is just prior to the final intracellular fusion in endosomes.
TJs are major components of cell-cell adhesion complexes and are composed of integral membrane proteins, including OCLN and CLDNs. Recent studies demonstrate that several peptides corresponding to ELs of CLDN1 and OCLN disrupt TJ integrity by inducing rapid internalization of TJ proteins.13–17 To determine the effect of CL58 on TJ function, CL58 was added to Caco-2 or Huh7.5.1 cells for the indicated period. As shown in Fig. 6A, CL58 had no detectable effect on the cellular distribution of CLDN1 and ZO-1, nor did it significantly inhibit the transepithelial resistance conferred by polarized Caco-2 cell monolayers or the re-establishment of TJs after Ca2+ depletion/repletion (Fig. 6B,C). CL58 is therefore unlikely to inhibit HCV entry by interfering with TJ function at its antiviral dose.
In order to investigate the possible interaction between CL58 and HCV envelope proteins, we performed coimmunoprecipitation experiments using Flag-tagged CL58. Interestingly, CL58 tagged at its C terminus but not its N terminus was able to precipitate with both HCV glycoproteins (Fig. 7). Consistently, the antiviral effect of synthetic FLAG-tagged CL58 on HCVpp entry was confirmed (Supporting Fig. 4). This observation implies that CL58 might exert its effect via potential interaction with HCV glycoproteins. The topology/structure of the overexpressed fusion polypeptide is important for its association with HCV E1 and E2.
HCV entry is a multistep event involving a number of host factors, including heparan sulfate proteoglycan, LDLR, SR-BI, CD81, CLDN1, and OCLN,8,18 all of which are located on the plasma membrane of permissive cells. These cellular factors now offer promising targets for novel antiviral treatments because viral entry is necessary for disease initiation, spreading, and transmission. For example, antibodies against CD81, SR-BI, and CLDN1 extracellular regions have been shown to block viral entry.19,20 Further, compounds such as ITX 5061, a SR-BI antagonist, or atriazine compound called EI-1 also inhibit HCV entry at a postbinding step.21 More strikingly, a novel peptide derived from the N terminus of HCV NS5A protein exerts a broad spectrum of virocidal effect on HCV and several other enveloped viruses.22 Through peptide library screening and rational design, we obtained a novel peptide, CL58, derived from human CLDN1, which potently inhibited HCV entry at a postbinding step. Together, our findings provide a proof of principle that a new class of inhibitors that block virus-host interactions may be developed.
The finding that CL58 inhibits HCV entry is interesting for two reasons. First, a number of peptides derived from OCLN ELs have been reported to induce endocytosis of TJ proteins and interfere with TJ integrity.13–16,23 Similarly, addition of a CLDN1 EL1 peptide (residues 53–80) to polarized cells interferes with epithelial barrier function.17 These findings make CLDN1 a relatively less attractive target for anti-HCV therapies, because reagents targeting CLDN1/OCLN ELs will likely cause leakage of important cellular barriers due to disrupted TJs. In sharp contrast, CL58 contains the first 18 aa of the CLDN1 N terminus but has no effect on CLDN1/OCLN distribution and is noncytotoxic at doses that exert potent antiviral activity. Thus, CL58 can potentially be a lead peptide for further design of useful therapeutics.
Second, the observed inhibitory kinetics of CL58 suggests that CL58 acts at a postbinding stage of virus entry. Interestingly, another study has also implicated CLDN1 in a late step during in HCV entry, perhaps after viral engagement of CD81.6 In this study, CL58 retained its inhibitory activity when added at even later time points than anti-CD81 antibody. Interestingly, Flag-tagged CL58 immunoprecipitated with HCV E1E2. Therefore, it is possible that CL58 readily penetrates lipid membrane owing to its small size and hence becomes capable of interacting with HCV E1E2. However, what this interaction means to CL58-mediated inhibition remains unclear. It will be interesting if such interaction disrupts the yet-to-be confirmed interactions between HCV glycoproteins and endogenous CLDN1 or the CLDN1-CD81 complex.24,25 Although we are unable to nail down either possibility (data not shown), the observation that CL58 also inhibited cell-cell fusion mediated by HCV glycoprotein and CLDN1 warrants further investigation in its ability to inhibit intracellular fusion between HCV and cellular membranes. It is noteworthy that TJ was first depicted as a fusion of the outer lipid leaflets of adjacent cell membrane bilayers (hemifusion).26 Regardless of its direct target, the anti-HCV activity is unique to CL58, but not those peptides derived from the respective region of CLDN6, CLDN7, and CLDN9.
In conclusion, the identification of CL58 now adds new tools in developing novel antiviral drugs that target HCV entry. This reagent will also aid to dissect the molecular mechanisms of HCV entry. Although most small molecule inhibitors that have advanced to the clinic target viral components, the peptide inhibitor described here may offer advantages, because it targets cellular proteins that are required for HCV infection and hence reduce the likelihood of developing resistance. By virtue of its distinct mechanisms of inhibition, CL58 may be used in combination with other anti-HCV drugs for potential synergistic effects in treating HCV infections.
We thank T. Wakita, H. Greenberg, C. Rice, F. Chisari, F. Cosset, G. Luo, Y. Chen, R. Bartenschlager, G. Gao, J. Dubuisson, C. Coyne, and J. McKeating for providing cell lines, reagents, and technical assistance.
Wei Yang received grants from the Chinese Science and Technology Key Project (2012ZX10002007-003-003, 2008ZX10002-014, and 2009ZX10004-303), the National Natural Science Foundation of China (30970156), and the National Basic Research Program of China (2011CB504800). Tianyi Wang received grants from the National Institutes of Health (NIHR21AI083389 and R01DK088787).
Potential conflict of interest: Nothing to report.
Additional Supporting Information may be found in the online version of this article.