Department of Medicine, University of Massachusetts Medical School, Worcester, MA
Address reprint requests to: Gyongyi Szabo, M.D. Ph.D., Department of Medicine, University of Massachusetts Medical School, LRB208,364 Plantation St., Worcester, MA 01605. E-mail: Gyongyi.Szabo@umassmed.edu; fax: 508-856-4770.
Potential conflict of interest: Nothing to report.
Host cytoskeletal proteins of the ezrin-moesin-radixin (EMR) family have been shown to modulate single-stranded RNA virus infection through regulating stable microtubule formation. Antibody engagement of CD81, a key receptor for hepatitis C virus (HCV) entry, induces ezrin phosphorylation. Here we tested the role of EMR proteins in regulating HCV infection and explored potential therapeutic targets. We show that HCV E2 protein induces rapid ezrin phosphorylation and its cellular redistribution with F-actin by way of spleen tyrosine kinase (SYK). Therapeutically blocking the functional roles of SYK or F-actin reorganization significantly reduced Huh7.5 cell susceptibility to HCV J6/JFH-1 infection. Using gene regulation, real-time quantitative polymerase chain reaction, western blot, and fluorescent microscopy analysis, we found that proteins of the EMR family differentially regulate HCV infection in the J6/JFH-1/Huh7.5 cell system. Moesin and radixin, but not ezrin, expression were significantly decreased in chronic HCV J6/JFH-1-infected Huh7.5 cells and HCV-infected patient liver biopsies compared to controls. The decreases in moesin and radixin in HCV J6/JFH-1-infected Huh7.5 cells were associated with a significant increase in stable microtubules. Ezrin knockdown inhibited immediate postentry events in HCV infection. Overexpression of moesin or radixin significantly reduced HCV protein expression. In contrast, transient knockdown of moesin or radixin augmented HCV infection. Making use of the Con1 HCV replicon system, we tested the effect of EMR proteins on HCV replication. We found that transient knockdown of moesin increased HCV RNA expression while overexpression of EMR showed no significant effect on HCV replication. Conclusion: Our findings demonstrate the important role of EMR proteins during HCV infection at the postentry level and highlight possible novel targets for HCV treatment. (Hepatology 2013;58:1569–1579)
Hepatitis C virus (HCV) infection is a leading cause of liver disease, with at least 2%–3% of the world's population chronically infected. Virus elimination through therapy can be limited by several factors, including adverse side effects to current drugs, viral resistance, patient alcohol abuse, and high cost of therapy.[2-6] Chronic HCV infection progressively leads to liver fibrosis, cirrhosis, hepatocellular carcinoma (HCC), and ultimately death. Viruses, including HCV, exploit host factors and interact with cell surface or intracellular proteins to achieve effective infection and/or replication.[8-10] Recently, numerous host proteins/peptides have been identified that possess potent antiviral properties.[8, 11-13] Indeed, host proteins/peptides have emerged as alternatives to conventional antiviral agents and present advantages over currently used antiviral drugs such as selective cytotoxicity for the target virus or virus-infected cells, bypassing multidrug-resistance mechanisms and inducing minimal side effects.[11, 13]
The HCV virus, a single-stranded positive-sense RNA virus of the Flaviviridae family, was initially identified and distinguished from hepatitis A/B virus infections based on its characteristic induction of microtubule paracrystalline aggregates in infected hepatocytes and liver biopsies of HCV-infected patients.[16-21] HCV infects hepatocytes using host cell molecules including human CD81, scavenger receptor class B type I, claudin 1, NPC1-L1 cholesterol absorption receptor, epidermal growth factor receptor (EGFR), and occludin as receptors or coreceptors.[23, 24] Recently, studies in B cells have shown that antibody engagement of CD81, a key receptor for HCV infection, induces spleen tyrosine kinase (SYK) phosphorylation of ezrin that recruits F-actin to facilitate cytoskeletal reorganization. In addition, recent studies have found that inhibition of actin and/or microtubule functions markedly reduced HCV infection in vitro.[10, 26] Taken together, these studies implicate host cytoskeletal proteins in the pathophysiology of HCV infection.
Ezrin-moesin-radixin (EMR) represents a group of human cytoskeletal proteins that regulate lentiviral infection by modulating stable and dynamic microtubule formation.[8, 9] EMR also function as linkers between the actin cytoskeleton and plasma membrane proteins and as signal transducers in numerous signaling pathways.[27, 28] This family of proteins show >70% sequence homology among members and display a high degree of functional redundancy. Given that EMR proteins regulate human immunodeficiency virus (HIV) infection,[8, 9] it remains unknown if these proteins can regulate other positive sense RNA virus infections including HCV.
In the present study we performed a comprehensive molecular analysis to characterize the role of human EMR in HCV infection and replication. Using HCV J6/JFH-1 virus infection and the HCV Con1 replicon system we demonstrate that HCV infection and replication can be modulated by proteins of the EMR family. We found that chronic HCV infection resulted in a significant decrease in moesin and radixin expression in Huh7.5 cells and HCV-infected patients compared to controls. The significant decrease in moesin and radixin in HCV J6/JFH-1-infected Huh7.5 cells was associated with increased stable microtubule aggregate formation. Overexpression or transient knockdown of EMR proteins differentially modulated target cell susceptibility to HCV infection and replication. These experiments provided mechanistic insights into modulation of HCV infection by the EMR family of proteins and identified targets for development of new therapies against HCV infection.
Materials and Methods
Cell Lines and HCV J6/JFH-1 Virus
Huh7.5 and Con1 HCV FL replicon cells were cultured as described Infectious and replication competent HCV J6/JFH-1 virions were generated using pFL-J6/JFH-1 plasmid as described. Detailed protocols are described in the Supporting Materials and Methods.
Patient Liver Biopsy Samples
Liver biopsy specimens were obtained from the National Institutes of Health (NIH) liver tissue cell distribution system (LTCDS; Minneapolis, MN; Pittsburgh, PA; Richmond, VA), which was funded by NIH contract # N01-DK-7-004/HHSN26700700004C. Patient samples represented genotypes 1a and 3.
Additional methods are included in the Supporting Materials.
Chronic HCV J6/JFH-1 Infection of Huh7.5 Cell or Chronic HCV Infection in Human Livers Down-Regulates Moesin and Radixin Expression and Increases Stable Microtubule Formation
Based on the previously reported roles of the EMR proteins in regulating HIV infection,[8, 9] we hypothesized that EMR proteins might have a role in HCV infection. We found that chronic HCV J6/JFH-1 infection of Huh7.5 cells resulted in HCV RNA and protein expression (Supporting Fig. 1A,B) with over 90% of cells infected after 96 hours (Supporting Fig. 1C). Chronic HCV infection of Huh7.5 cells was associated with a significant decrease in moesin and radixin but not ezrin expression both at the messenger RNA (mRNA) (Supporting Fig. 2A-C)) and protein levels (Fig. 1A-C). Liver biopsies from chronic HCV-infected patients with confirmed HCV RNA expression (Supporting Fig. 2D) and elevated serum aspartate aminotransferase (AST) levels (Supporting Fig. 2E) also showed significant decreases in moesin (Fig. 1D) and radixin (Fig. 1E), but not in ezrin (Fig. 1F) protein expression. Fluorescent microscopy analysis of Glu-Tubulin revealed that the decrease in moesin and radixin after HCV J6/JFH-1 infection was associated with an increase in stable microtubule formation (Fig. 2A) and Glu-Tubulin protein expression (Supporting Fig. 3A). We found that transient knockdown of EMR proteins (Supporting Fig. 2B) significantly increased stable microtubule formations in Huh7.5 cells even in the absence of HCV infection (Fig. 2B; Supporting Fig. 3C). These observations demonstrate a direct role of EMR proteins in modulating stable microtubule formation.
HCV J6/JFH-1 Infection of Huh7.5 Cells Is Dependent on Spleen Tyrosine Kinase, Ezrin-Radixin Phosphorylation, and F-actin Reorganization
CD81 is a tetraspanin family member which is important for HCV infectivity. Recent reports indicated that CD81-engagement induced SYK activation, ezrin phosphorylation, and F-actin reorganization,[25, 33, 34] as well as expression of endogenous SYK in normal and HCV-infected hepatocytes.[35, 36] Based on these reports, we surmised that SYK phosphorylation of ezrin leads to its redistribution with F-actin and modulates postentry HCV trafficking towards the endoplasmic reticulum for efficient virus infection. In coculture experiments we found that HCV J6/JFH-1 infection induced time-dependent phosphorylation of SYK (Y323) in Huh7.5 cells (Fig. 3A). To decipher the likely viral component mediating SYK activation, we cocultured Huh7.5 cells with various HCV proteins and identified that HCV E2 protein engagement of CD81 induced SYK activation (Fig. 3B). Given that activated SYK phosphorylates ezrin leading to its redistribution with F-actin in B-cells, we tested if HCV J6/JFH-1 engagement of CD81 led to a similar occurrence. We found that the activated SYK led to a time-dependent phosphorylation of ezrin (pY354 and pThr567) and radixin (pThr564) (Fig. 3C,D). SYK was responsible for ezrin and radixin phosphorylation, given that a specific inhibitor of SYK phosphorylation (Bay 61-3606) inhibited ezrin/radixin phosphorylation upon HCV J6/JFH-1 virus engagement of CD81 in Huh7.5 cells (Fig. 3D). We confirmed by coimmunoprecipitation western blot experiments that phospho-SYK interacts with phospho-ezrin, thus demonstrating that phospho-SYK most likely phosphorylates ezrin (Supporting Fig. 4). Given that CD81 engagement by HCV E2 protein induced SYK phosphorylation (Fig. 3B), we tested the functional impact of these signaling events in HCV infection. Using the HCV J6/JFH-1 and Huh7.5 experimental system, we found that transient knockdown of SYK by small interfering RNA (siRNA), or use of a potent and reversible SYK inhibitor, BAY 61-3606, significantly reduced HCV core and NS3 protein expression in Huh7.5 cells, suggesting the involvement of SYK in HCV infection (Fig. 3E,F). Because SYK activation and ezrin phosphorylation result in F-actin reorganization, use of a specific F-actin reorganization inhibitor, cytochalasin B, resulted in a dose-dependent inhibition of HCV infectivity at the HCV RNA (Fig. 4A) and NS3 protein levels (Fig. 4B). The chemical agents used showed no cellular toxicity (Supporting Fig. 5A,B).
Proteins of the EMR Family Differentially Regulate HCV Infection and Replication
The HCV life cycle involves multiple events including cell entry, postentry trafficking, intracellular replication of viral RNA and proteins, assembly, and release. To determine the role of EMR proteins in HCV infectivity and replication we took advantage of the HCV J6/JFH-1, HCV E1/E2 pseudo-particles (HCVpp), and HCV Con1 replication systems. Because chronic HCV infection resulted in decreased moesin and radixin expression, we asked if decreases in moesin or radixin prior to infection could modulate target cell susceptibility to infection. Indeed, siRNA knockdown of moesin (Fig. 5A) and radixin (Fig. 5B) prior to infection with HCV J6/JFH-1 virus led to significantly higher HCV NS3 protein (Fig. 5A,B) and HCV RNA expression (Supporting Fig. 6). In contrast, overexpression of moesin or radixin prior to HCV J6/JFH-1 infection significantly reduced Huh7.5 cell susceptibility to infection demonstrated by reduced HCV NS3 protein levels (Fig. 5C,D). Given that SYK inhibition decreased HCV infection via ezrin, we tested the role of ezrin in regulating HCV infection. Transient knockdown of ezrin prior to HCV J6/JFH-1 infection resulted in significantly lower HCV NS3 (Fig. 5E) protein and RNA (Supporting Fig. 6) in Huh7.5 hepatoma cells compared to controls. These observations suggest that ezrin, which is the only EMR protein that has been shown to associate and redistribute with F-actin, can be exploited by HCV to mediate postentry trafficking within the cell, similar to observations with other viruses for effective infection.[38, 39] However, overexpression of ezrin prior to HCV J6/JFH-1 infection of Huh7.5 hepatoma cells had no significant effect on HCV NS3 protein expression (Fig. 5F), suggesting that in the presence of excess ezrin, the virus multiplicity of infection (MOI) determines the level of virus infection.
Next, we assessed at which level in the HCV life cycle EMR proteins exerted their effect using HCVpp. We found that transient knockdown of moesin and radixin resulted in increased HCVpp infection of Huh7.5 cells (Fig. 6A). Alternatively, overexpression of moesin and radixin or transient knockdown of ezrin proteins significantly decreased HCVpp infection of Huh7.5 cells (Fig. 6A,B). These observations suggest that EMR proteins mostly likely regulate HCV infection postvirus entry.
We also tested the Con1 full-length replicon cells, which are capable of HCV RNA replication without producing infectious virions. Compared to parent Huh7.5 cells, Con1 full-length replicons expressed significantly higher ezrin (Fig. 7A), lower moesin (Fig. 7B), and comparable radixin (Fig. 7C) levels. We observed that transient knockdown of moesin in the HCV Con1 replicon system (Fig. 7D) markedly increased HCV RNA expression (Fig. 7E), while ezrin or radixin knockdown (Fig. 7D) had no effect (Fig. 7E). Overexpression of EMR using nongreen fluorescent protein (GFP)-tagged EMR expression vectors in Con1 replicon cells had no significant effect on HCV replication (Fig. 7F). Taken together, these findings suggest that only moesin plays a role in HCV RNA replication in Con1 FL replicon (genotype 1b) cells.
Treatment of Chronic HCV J6/JFH-1-Infected Huh7.5 and FL Con1 Replicon Cells Restores Moesin/Radixin to Preinfection Levels
As chronic HCV J6/JFH-1 infection of Huh7.5 cells or Con 1 FL replicon cells resulted in a significant decrease in radixin and or moesin, we evaluated whether treatment with antiviral drugs could restore moesin and or radixin expression. We found that a combination of recombinant human interferon-alpha and telaprevir over a course of 10 days significantly decreased HCV NS3 proteins in chronic HCV J6/JFH1-infected Huh7.5 cells and Con1 FL replicon cells (Fig. 8A-C). This was associated with a significant restoration of radixin and or moesin protein expression to preinfection levels (Fig. 8A-C).
HCV infection is a multistep process involving viral glycoproteins E1/E2 and host factors including heparan sulfate proteoglycans, CD81, SR-BI, LDL-R, CLDN1, occludin, EGFR, NPLC1-L1 cholesterol receptor, DC-SIGN, and L-SIGN.[23, 24] After successful binding to a target cell, HCV must penetrate the cell membrane and traverse the dense cytoplasm to the endoplasmic reticulum, where virus replication occurs. The presence of a dense cytoskeletal network and cellular organelles greatly impedes diffusion of macromolecules including viruses. As such, viruses have developed functional ways of hijacking host actin and microtubules for short- and long-distance transport, respectively, within host cells. Here we demonstrate the role of EMR proteins as important players modulating efficient HCV infection.
EMR are closely related cytoskeletal proteins containing an N-terminal FERM (Band Four-point one) domain which interacts with the Ig-like EW-2 and EWI-F. EW-2 and EWI-F proteins that have been shown to limit HCV infection and form a direct link between EMR with the tetraspanin CD81. Upon cellular activation, the highly conserved N-terminal domain of EMR proteins binds to other cellular proteins while the C-terminal domain binds to F-actin filaments. Recent reports suggest that activation of EMR proteins can be mediated by the Rho family of GTPases. Phospho-SYK has also been shown to specifically activate ezrin upon CD81 engagement. These observations suggest that EMR proteins do act as cross-linkers between the plasma membrane CD81, cellular actin filaments, and molecular signaling transducers within cells. In concert with this suggestion, a recent report showed that antibody engagement of CD81 in B-cells induced phospho-SYK dependent ezrin phosphorylation and its cellular redistribution with filamentous actin. From our kinetic immunoprecipitation and western blot studies, we found that HCV J6/JFH-1 virus E2 protein engagement of CD81 induced a time-dependent SYK activation and ezrin phosphorylation in Huh7.5 cells. Additionally, we found that F-actin coupling/redistribution with ezrin following ezrin phosphorylation was crucial for effective HCV infection, given that cytochalasin-B pre-treatment of Huh7.5 cells prior to HCV J6/JFH-1 infection resulted in decreased HCV infection. These findings identified F-actin reorganization and coupling as an important step during HCV infection.
Moesin and radixin expression was significantly decreased both in vitro and in chronic HCV-infected patient liver biopsies including genotype 1a, 1b, and 3 as well as in the J6/JFH-1 system (genotype 2a) of the HCV genotype, suggesting that the role of EMR proteins are most likely conserved and consistent between HCV genotypes. The decrease in moesin and radixin was associated with a significant increase in stable microtubule expression in chronic HCV J6/JFH-1-infected Huh7.5 cells. This scenario hypothetically creates microtubule “rail-roads” facilitating postentry HCV trafficking and enhancing effective virus infection. This hypothesis was confirmed using gene regulation approaches where transient knockdown of moesin or radixin in Huh7.5 cells prior to HCV J6/JFH-1 or HCVpp infection resulted in increased HCV infection. On the contrary, transient ezrin knockdown significantly reduced HCV J6/JFH-1 and HCVpp infection of Huh7.5 cells. Alternatively, overexpression of moesin or radixin proteins abrogated J6/JFH-1 HCV or HCVpp infection in Huh7.5 cells. Ezrin overexpression showed no significant effect on Huh7.5 cell susceptibility to infection. These observations suggest that ezrin functions at the level of immediate virus entry, while increased microtubules, as a result of decreased moesin and radixin, modulate postentry events facilitating virus transport. Our observations are in concert with recent reports where EMR proteins were involved in efficient vesicular stomatitis virus (VSV-G) pseudotyped lentivirus infections,[8, 9] given that the HCVpp has an HIV (lentiviral) core.
Our data in HCV Con1 full-length replicon cells that mimic HCV RNA replication without producing infectious virions indicate that reduced moesin affected replication. Transient knockdown of moesin but not ezrin or radixin led to a significant increase in HCV RNA levels, suggesting the first effect of moesin may be mediated at the level of HCV RNA replication. Overexpression of EMR had no effect on HCV RNA replication, suggesting that EMR proteins have a limited role in HCV RNA replication. Therapeutically, we found that interferon alpha and telaprevir over 10 days restored moesin and radixin to preinfection levels. This observation indicates that the significant decrease in liver moesin and radixin expression associated with chronic HCV can be restored by HCV elimination.
Taken together, our results show for the first time a direct link between EMR proteins and the induction of microtubule aggregate formation observed during chronic HCV infection in patients and in in vitro culture systems.[16-21] We demonstrated that EMR proteins exert differential roles in HCV infectivity and replication and identified novel signaling regulators in HCV infection. In conclusion, our findings, illustrated in Supporting Fig. 7, reveal mechanistic and signaling events regulating HCV postentry and trafficking within target cells involving SYK, F-actin, stable microtubules, and EMR proteins, thereby providing novel targets for anti-HCV therapies.
The authors thank Dr. Charles M. Rice and Dr. Takaji Wakita for kindly providing reagents and Dr. S. Shaw for the Ezrin overexpression plasmid. We thank Drs. W. Thomas and Molrine (Mass Biologics) for providing the HCV pseudo-virus and anti-E2 antibody. The following reagent was obtained through the NIH AIDS Research and Reference Reagent Program, Division of AIDS, NIAID, NIH: pNL4-3.Luc.R–E– from Dr. Nathaniel Landau.