Impaired expression and function of toll-like receptor 7 in hepatitis C virus infection in human hepatoma cells

Authors


  • Potential conflict of interest: Nothing to report.

Abstract

Hepatitis C virus (HCV) interferes with interferon (IFN)-mediated innate immune defenses. Toll-like receptor (TLR) 7 agonists robustly inhibit HCV infection. We hypothesize that HCV infection may interfere with the expression and/or function of TLR7, a sensor of single-stranded RNA. We identified reduced TLR7 RNA and protein levels in hepatoma cells expressing HCV (full-length, BB7-subgenomic, and JFH-1 clone) compared with control HCV-naïve cells. The biological relevance of this finding was confirmed by the observation of decreased TLR7 RNA in livers of HCV-infected patients compared with controls. HCV clearance, by IFN-α treatment or restrictive culture conditions, restored the decreased TLR7 expression. Treatment with RNA polymerase inhibitors revealed a shorter TLR7 half-life in HCV-replicating cells compared with controls. Downstream of TLR7, an increased baseline IRF7 nuclear translocation was observed in HCV-positive cells compared with controls. Stimulation with the TLR7 ligand R837 resulted in significant IRF7 nuclear translocation in control cells. In contrast, HCV-replicating cells showed attenuated TLR7 ligand-induced IRF7 activation. Conclusion: Reduced TLR7 expression, due to RNA instability, directly correlates with HCV replication and alters TLR7-induced IRF7-mediated cell activation. These results suggest a role for TLR7 in HCV-mediated evasion of host immune surveillance. (HEPATOLOGY 2009.)

Chronic hepatitis C virus (HCV) infection affects around 2% of the world's population.1 With limited therapeutic options and no effective vaccine, chronic HCV infection is a growing public health burden.

HCV is a positive single-stranded RNA virus.2 Some HCV-derived products, including HCV RNA and several HCV proteins, trigger host defense.3, 4 Immune responses, including cell-mediated immunity and type 1 IFNs, are vital in controlling and clearing HCV infection.5 A number of host receptors, including TLR3, TLR7, TLR8, and TLR9 and helicases RIG-I and MDA-5, lead to a type 1 IFN-mediated response upon stimulation with viral RNA.6, 7 HCV has developed mechanisms to bypass the immune defense and facilitate viral persistence: NS3/4A protein actively disrupts TLR3- and RIG-I/MDA-5–mediated signaling pathways, thus interfering with the IFN production and bypassing the immunomediated response.8, 9 The effect of HCV on other IFN-inducing pathways is currently unknown.

Viral recognition receptors, including TLR7, activate IRF7 through a series of phosphorylation events starting with recruitment of MyD88, the common TLR adaptor protein. Once the receptor is engaged, a complex is formed between MyD88, TRAF6, IRAK4, and IRAK1 that allows for activation of IRF7. Both TRAF6 and IRAK1 have been reported to be responsible for phosphorylating IRF7, leading to its homodimerization and subsequent nuclear translocation.10, 11 Recent reports have indicated that robust TLR7 agonists decrease HCV RNA in HCV-infected patients12 and HCV RNA and NS5A protein in HCV-replicating hepatoma cells.13 TLR7 has also been reported to recognize other single-stranded RNA viruses such as influenza14 and dengue.15 Based on these reports, we postulated that TLR7 may play a role in HCV infection and analyzed the hypothesis that HCV, in an effort to bypass immune surveillance, interferes with TLR7 expression and/or function. Our results demonstrate that HCV infection produces instability of TLR7 RNA, thus leading to low TLR7 RNA and protein expression and impaired TLR7-mediated activation of the IRF7 signaling pathway.

Abbreviations

GAPDH, glyceraldehyde 3-phosphate dehydrogenase; HCV, hepatitis C virus; IFN, interferon; mRNA, messenger RNA; qRT-PCR, quantitative real-time polymerase chain reaction; TLR, Toll-like receptor.

Materials and Methods

Reagents.

Dulbecco's modified Eagle's medium and nonessential amino acids Geneticin and TrypLE Express were from Gibco (Grand Island, NY); fetal bovine serum was obtained from Hyclone (Logan, UT). Ciprofloxacin was from CellGro (Herndon, VA); IFN-α, actinomycin D, and α-amanitin were from Sigma Aldrich (St. Louis, MO); and R837 (Imiquimod) was from Invivogen (San Diego, CA).

Cells.

All cells were grown in 5% CO2 at 37°C and passed at 75% confluency. Hepatoma cell lines Huh7 and Huh7.5 were maintained in Dulbecco's modified Eagle's medium with 10% heat-inactivated fetal bovine serum, 1% nonessential amino acids, and 500 μg/mL ciprofloxacin. Stable HCV cell lines FL on Huh 7.5 and BB7 on Huh7 background were maintained in Dulbecco's modified Eagle's medium with 10% fetal bovine serum, 1% nonessential amino acids, 500 μg/mL ciprofloxacin, and Geneticin (750 μg/mL or 500 μg/mL, respectively). The JFH-1 cell line on Huh7.5 background was propagated as described.16

RNA Quantification.

For analysis with quantitative real-time polymerase chain reaction (qRT-PCR), cellular RNA was extracted with an RNeasy Kit (Qiagen, Valencia, CA) and on column DNase (Qiagen) treatment, and complementary DNA was generated using a reverse transcription system (Promega, Madison, WI). Host gene quantification was performed by way of qRT-PCR with SybrGreen MasterMix (Eurogentec, San Diego, CA), and HCV was assessed with Taqman MasterMix (Applied Biosystems, Foster City, CA), all using Eppendorf Realplex4 Mastercycler (Eppendorf, Westbury, NY). The sequences of the specific primers are listed in Supporting Table 1.

Protein Detection and Quantification.

All proteins were quantified by way of western blotting, as described.4 We used anti-NS5A (1:500) or anti–β-actin (1:10,000, Abcam, Cambridge, MA), and secondary horseradish peroxidase–labeled antibody (1:10,000, Santa Cruz Biotechnology, Santa Cruz, CA) and enhanced chemiluminescence (LumiGlo kit, Cell Signaling Technology (Danvers, MA) to identify the immunoreactive bands.

For flow cytometry analysis, cells were detached from plates with TrypLE Express (Invitrogen, Carlsbad, CA), washed, fixed, and permeabilized with BD CytoFix/CytoPerm Kit (BD Biosciences). Cells were stained with anti-human TLR7 antibody (1:100, Imgenex, San Diego, CA) for 1 hour in CytoPerm buffer, washed with CytoPerm/CytoFix buffer, and stained with secondary isotype-phycoerythrin antibody (1:1,000, Invitrogen) for 1 hour; secondary antibody alone was used as a control. Cell fluorescence was analyzed using an LSRII (BD Biosciences) flow cytometer and FlowJo fluorescence-activated cell sorting analysis program (TreeStar, Ashland, OR). The specificity of anti-TLR7 antibody was confirmed using TLR7-specific small interfering RNA (Supporting Fig. 1) and fluorescence activated cell sorting analysis.

For small interfering RNA transfection, cells were plated in 6-well plates at 104/mL. Small interfering RNAs (final concentration, 300 nM/well; Ambion, Austin, TX) were combined with Mirus TransIT transfection reagent and Opti-MEM medium for 15 minutes and added to the cells; 72 to 96 hours later, the cell were analyzed as indicated.

IRF7 Nuclear Translocation Analysis.

Cells were plated at 8 × 105/mL, grown overnight in 5% CO2 at 37°C, and stimulated with R837 (10 μg/mL) for a time course of up to 4 hours. Cells were fixed and permeabilized as described for the flow cytometric analysis and stained for IRF7 (1:50) for 1 hour on ice. Cells were counterstained with secondary AlexaFluor488 antibody (1:1,000) for 1 hour on ice, followed by DRAQ5 (5 μM; Biostatus Limited, Leicestershire, UK) for nuclear staining. Image files of 10,000 events were collected for each sample using the ImageStream system (Amnis, Seattle, WA) and analyzed using IDEAS software (Amnis). Single cells were identified by way of in-focus gating on IRF7 or DRAQ5 fluorescent events with high nuclear aspect ratios (minor to major axis ratio, a measure of circularity) and high nuclear contrast (as measured by the Gradient Max feature). Nuclear localization of IRF7 was measured using the similarity score, which quantifies the correlation of pixel values of the nuclear and IRF7 images on a per-cell basis.17 If the IRF7 is localized in the nucleus, the two images are similar and have large positive values. If the IRF7 is cytoplasmic, the two images are dissimilar and have large negative values.17 Events with positive values had visually apparent nuclear distributions of IRF7 and were gated to quantify the percentage of cells within the hepatoma population with nuclear-localized IRF7.

Results

HCV Replicating Cells Show Reduced TLR7 Levels.

Expression of TLRs is not restricted to immune cells, and recent studies have suggested that human hepatocytes express all known TLRs.18 Based on previous reports connecting TLR7 with the reduction of HCV infection,10, 13 we postulated that TLR7 may play a role in anti-HCV immunity. Specifically, we analyzed whether HCV may interfere with TLR7 expression and/or functions in a disruptive manner.

To address our hypothesis, we used Huh7 and Huh7.5 hepatoma control cell lines and the following HCV-replicating stable cell lines: (1) FL, which contains the full-length HCV genome transfected into the Huh7.5 parental line; (2) BB7, which contains the subgenomic HCV genome transfected into Huh7; and (3) JFH-1, which contains the infectious full-length HCV genome transfected into Huh7.5 (Fig. 1A). HCV NS5A protein expression was confirmed by way of western blot analysis for all HCV-expressing cell lines (Fig. 1B).

Figure 1.

Hepatoma cells actively replicating HCV show reduced TLR7 mRNA and protein. (A) Genomic sequences of HCV-replicating cell lines JFH-1, full-length HCV genotype 2a, FL, full-length HCV genotype 1a, and BB7, subgenomic HCV genotype 1b. (B) HCV-replicating cells and hepatoma control whole cell extracts (5 μg) immunoblotted against antibodies for HCV NS5A or β-actin used as a loading control. Shown is a representative blot of n = 3. (C) RNA extracts of HCV-replicating and hepatoma control cells used to amplify TLR7 mRNA by way of qRT-PCR normalized to glyceraldehyde 3-phosphate dehydrogenase (GAPDH) mRNA. (D) Hepatoma and HCV-replicating cells were fixed, permeabilized and stained for TLR7 receptor expression; data were analyzed by way of flow cytometry. One representative set of histograms out of n = 3 is shown. (E,F) RNA extracts of liver samples from controls (E,F), patients with HCV (E), and those with HBV infection or of nonviral nonalcoholic steatohepatitis (NASH) (F) were analyzed for TLR7 mRNA qRT-PCR. HCV average indicates the average TLR7 mRNA of the six HCV patient samples. (C,E,F) Data are shown as the mean ± standard deviation (n = 3).

We identified detectable TLR7 RNA (Fig. 1C) and protein levels (Fig. 1D) in control Huh7 and Huh7.5 cells. However, there was a significant decrease of TLR7 RNA (Fig. 1C) and protein (Fig. 1D) in all HCV-positive cell lines. These data suggest that TLR7 levels are impaired in the presence of HCV in vitro. Expression of TLR7 RNA was significantly reduced in livers of patients with chronic HCV infection compared with healthy, HCV-naïve controls (Fig. 1E). In contrast, there was no decrease in the levels of TLR7 expression in the livers of patients infected with HBV or those with nonalcoholic liver disease (Fig. 1F) compared with controls. These data confirmed the in vivo significance of reduced TLR7 expression and allowed us to conclude that TLR7 decrease during HCV infection is HCV-specific and is not a result of nonspecific inflammatory response.

HCV Clearance Restores TLR7 Expression.

Reduced TLR7 expression in HCV-replicating hepatoma cells and in the livers of HCV-infected patients indicated a novel phenotype possibly caused by HCV infection. Thus, we hypothesized that the reduced TLR7 expression was due to HCV replication. To test this hypothesis, we reduced HCV replication using three different approaches: (1) IFN-α treatment, (2) culture of stable HCV-expressing cell lines in selection antibiotic-depleted media, and (3) self-restriction of viral replication through high passage number of JFH-1 cells.

First, treatment of HCV-replicating cells with IFN-α over a 48-hour time course led to time-dependent reduction of HCV replication in all three HCV cell models including FL, BB7, and JFH-1 (Fig. 2A). In contrast to the significant decrease in HCV levels (Fig. 2A), TLR7 RNA significantly increased upon IFN-α treatment in all three HCV-replicating cells (Fig. 2B).

Figure 2.

IFN-α curing of HCV replication increases TLR7 mRNA expression. (A) HCV mRNA was amplified by Taqman qRT-PCR and (B) TLR7 was amplified by SybrGreen qRT-PCR in HCV-replicating cells, FL, JFH-1, and BB7; all data were normalized to GAPDH mRNA levels. IFN-α (1 ng/mL) was administered to cells for a 48-hour time course, and samples were taken at indicated intervals. Data are shown as the mean ± standard deviation (n = 3).

Second, culturing FL and BB7 HCV-replicating cells without selection media, Geneticin, for 7 days, resulted in loss of HCV and restoration of TLR7 RNA (Fig. 3A). The increasing messenger RNA (mRNA) levels upon decreasing HCV RNA expression correlated with an increase in TLR7 protein levels (Fig. 3B).

Figure 3.

Reduction of HCV replication increases TLR7 mRNA and protein expression. (A) RNA was extracted from FL and BB7 cell lines. HCV was amplified by Taqman qRT-PCR, TLR7 mRNA was amplified by SybrGreen qRT-PCR, and all values were normalized to GAPDH. (B) Cell samples were stained with TLR7 antibody representing TLR7 protein and analyzed by flow cytometry. Cells stained with isotype-specific immunoglobulin G were used as controls. HCV-replicating cells were passed and maintained in Geneticin-deficient media on day 0, and RNA and protein samples were taken on the indicated days over a period of 1 week. Data are shown as the mean ± standard deviation (n = 4).

Third, in JFH-1 cells, we examined TLR7 protein levels in parallel with HCV expression over a series of passages. JFH-1 replication was depleted after 14 serial passages; in contrast, TLR7 protein levels were restored (Fig. 4). These data demonstrate a direct correlation between active HCV replication and decreased TLR7 expression.

Figure 4.

Diminished HCV infection from serial passages of JFH-1 cells increases TLR7 protein expression. JFH-1 cells were passed every 3 to 4 days, with the first passage performed after electroporated cell recovery.16 RNA extracts were taken from JFH-1–expressing cells from the indicated passage number. HCV mRNA was amplified using Taqman qRT-PCR and normalized with GAPDH. TLR7 protein levels were quantified by way of flow cytometry from cells fixed, permeabilized, and stained for TLR7.

HCV Expression Leads to TLR7 mRNA Instability.

Deficient production of RNA coding for a gene, which can lead to impaired protein expression, is often caused by modifications at the level of gene transcription. More importantly, some viruses manipulate host mRNA stability to achieve delayed and/or impaired immune responses, thus leading to viral persistence.19 Considering our results in HCV-replicating cells showing decreased TLR7 due to HCV replication, and recent reports of HCV NS5A protein interference with the TLR7-mediated signaling pathway resulting in impaired cytokine production,20 we predicted that HCV may affect the stability of TLR7 mRNA. To investigate this hypothesis, we used actinomycin D, an inhibitor of mRNA transcription achieved by binding the DNA at the transcription initiation complex and thus preventing elongation by RNA polymerase,21 and quantified the TLR7 mRNA half-life levels using qRT-PCR over a time course. The half-life of TLR7 mRNA was significantly shorter in HCV-replicating cells (FL, P < 0.0387 [Fig. 5A,E]; BB7, P < 0.0092 [Fig. 5B,E]) compared with controls. A similar pattern of decreased TLR7 mRNA stability in HCV-replicating cells was observed using α-amanitin, another inhibitor of mRNA transcription that interacts with the bridge helix in RNA polymerase II, allowing it to disrupt the translocation of RNA required to vacate the area for the next round of RNA synthesis22 (Fig. 5B,D,F). Reduction in mRNA stability during HCV replication was specific to TLR7, because the half-life of TLR5 mRNA (TLR5 does not recognize viruses or viral-derived products) yielded no significant changes in control cells and HCV-replicating cells (Supporting Fig. 2). Furthermore, actinomycin D and α-amanitin had no effect on HCV replication (Supporting Fig. 3). These results indicate that HCV replication can decrease TLR7 expression by way of TLR7 mRNA instability.

Figure 5.

RNA polymerase inhibitors reveal TLR7 mRNA shortened half-life in HCV-replicating cells compared with control cells. (A,C,E) Cells were treated with RNA polymerase inhibitor actinomycin D (2 μg/mL) or (B,D,F) α-amanitin (10 μg/mL) as indicated. RNA extracts were taken at the indicated time points and TLR7 mRNA was quantitated by qRT-PCR normalized with GAPDH. Half-life is indicated on the y-axis as relative quantities (A-D) or as absolute numbers (E,F). Data are shown as the mean ± standard deviation (n = 3).

HCV Expression Leads to Increased Baseline and Impaired TLR7 Ligand-Induced IRF7 Activation.

Our results so far showed a correlation between HCV replication and impaired TLR7 expression. Others have reported an increased expression of type 1 IFN-inducible genes in HCV-infected patients,23–25 but no significant increase in the IFN-dependent genes ISG-15 and OAS in HCV-infected patients when given isatoribine, a TLR7 agonist.12 These data together prompted us to explore the effect of HCV replication on the TLR7-mediated type 1 IFN signaling pathway. IRF7 is vital to TLR7-induced type I IFN production.26 Using ImageStream technology, we analyzed the activation of IRF7 by investigating IRF7 nuclear translocation in control and HCV-replicating hepatoma cells. Unstimulated control cells showed IRF7 localized in the cytoplasm separate from the nuclear compartment (untranslocated; Fig. 6A), whereas TLR7-stimulated cells demonstrated IRF7 colocalized with the nucleus (translocated; Fig. 6B). The majority of unstimulated control cells showed minimal IRF7 nuclear translocation and baseline activation of IRF7 (Fig. 6C). Compared with control cells, HCV-replicating cells FL and BB7 had a significantly higher baseline activation of IRF7 (Fig. 6C). Upon stimulation with the TLR7 ligand R837, control cells showed a significant up-regulation of IRF7 activation by 2 to 4 hours (Fig. 6D). The robust IRF7 activation in control cells was distinct from the minimal and insignificant activation in HCV-replicating cells (Fig. 6D). The reduced IRF7 activation in HCV-replicating cells was not due to delayed IRF7 activation because these cells failed to show stimulation-induced IRF7 nuclear translocation even at later timepoints (6-24 hours) (data not shown). The observation that stimulation of HCV-replicating cells with R837 triggered only minimal activation of IRF7 compared with the existing high baseline activation was in agreement with low expression of TLR7 (Fig. 1). These results indicate that the presence of HCV modulates the baseline expression of IRF7; however, the extent of TLR7-induced activation of HCV-replicating cells is impaired.

Figure 6.

Type 1 IFN-mediated pathway by way of IRF7 nuclear translocation is elevated in HCV-replicating cells yet is less responsive to TLR7 stimulation compared with controls. Cells were unstimulated or stimulated with R837 (10 μg/mL), fixed/permeabilized, and stained with anti-IRF7 antibody (green), and DRAQ5 (nucleus, red). Cells were analyzed using the ImageStream system, which individually excites fluorescently stained cells with a 488-nm laser and a brightfield light source while collecting cell images equivalent to ×40-60 magnification and fluorescent intensity data through a custom charge-coupled device camera. A representative brightfield of (A) unstimulated and (B) TLR7 (R837)-stimulated cells is shown (n = 3). (C) Unstimulated control and HCV-replicating cells were fixed, permeabilized, and stained for IRF7 and nucleus then analyzed using the ImageStream system. Percent of IRF7 nuclear translocation was calculated with IDEAS software using data collected from 10,000 events in each sample. (D) Percent of IRF7 nuclear translocation after R837 stimulation for up to 4 hours. Data are shown as the mean ± standard deviation (n = 3).

Discussion

Immune evasion by HCV has been documented in several different host cell types, and it has been suggested to play a key role in viral persistence and development of chronic infection.5, 8, 9 HCV interferes with host defense at multiple levels3–5, 8, 9, 20; here we show that HCV employs a novel mechanism for immune evasion by specifically targeting TLR7 expression, mRNA stability, and function.

We identified a significant decrease of TLR7 expression in the presence of HCV infection both in vitro in hepatoma cells and in vivo in human HCV-infected livers. We established the direct effect of HCV replication on TLR7 expression, as indicated by restoration of TLR7 levels upon viral suppression. Among others, the most frequently encountered mechanisms of viral-dependent impaired expression of a cellular receptor include interference with gene transcription or protein translation, posttranscriptional and posttranslational receptor modifications, and disruption of the downstream signaling events.8, 19 We determined that HCV infection interferes with at least two of these events regarding TLR7.

First, HCV infection interfered with the TLR7 gene during transcriptional regulation as shown by TLR7 mRNA instability in HCV-replicating cells. Induction of host mRNA instability is a common evasion strategy employed by other viruses, including Herpesviruses, Vaccinia, and Influenza to bypass immune surveillance.19, 27 Among the mechanisms of induction of host mRNA instability are association of viral proteins with several members of the mammalian exosome, activation of the exosomal enzymes and/or recruitment of the exosome to mRNAs, usage of the exosome as a vehicle to access cellular mRNAs if the viral protein possesses RNase activity, and enhancement of exosomal activity whether or not the viral protein functions as an RNase; most often viruses employ more than one mechanism to achieve total or partial host shutoff.27–30 The detailed mechanisms of HCV-induced TLR7 mRNA instability are yet to be discovered; it is unknown whether HCV uses the ubiquitin machinery to target TLR7 mRNA for degradation, produces small interfering RNAs targeting TLR7 mRNA, exploits host regulatory factors that bind to TLR7 mRNA or uses an HCV protein to aid the TLR7 mRNA instability. Nevertheless, impaired TLR7 mRNA stability results in low levels of TLR7 protein expression during HCV infection.

It is not impossible that HCV-induced TLR7 mRNA instability is not solely responsible for the observed impaired TLR7 protein levels. Several viruses affect different steps of the host protein expression, including transcriptional, posttranscriptional, and posttranslational modifications29, 30; such possibility with regard to TLR7 remains to be analyzed in the case of HCV. Reduced TLR7 levels in cells during HCV RNA replication may also occur through processes that can be attributed directly or indirectly to HCV products through intracellular host interactions. This scenario remains to be analyzed using complex immune cells and liver parenchymal cell cocultures in the future.

Second, we identified an up-regulation of IRF7 activation in HCV-replicating cells compared with control, noninfected hepatoma cells. This observation suggests that whereas the relative TLR7 expression is decreased by HCV infection, HCV infection results in activation of the IRF7 pathway in hepatoma cells. Our results also revealed that although TLR7-induced IRF7 nuclear translocation was significantly increased in control cells, it failed to appreciably enhance further from elevated background activation of IRF7 in HCV-infected cell lines. These results indicated that the remaining TLR7 receptors cannot fully compensate for the diminished TLR7 levels. This novel observation suggests that HCV impairs not only TLR3 and RIG-I/MDA-5 signaling but also interferes with TLR7-mediated signaling pathways. Because the baseline IRF7 activation in HCV-replicating cells was higher compared with controls, we concluded that TLR7-independent activation of IRF7 signaling pathways was intact in the presence of HCV infection. Baseline IRF7 activation has been correlated with endogenous IFN levels in some cells, whereas other ligands such as lipopolysaccharide have been shown to activate IRF7 also. In addition to TLR7, several other receptors can activate IRF7, including TLR8, TLR9, or TLR3. It is unknown whether these receptors play a role in the basal activation of both control hepatoma and HCV-replicating cells. Our novel data identified TLR7 as a specific target of HCV based on the fact that only TLR7—but not TLR4 or TLR5—mRNA stability was targeted during HCV infection, suggesting the specific influence.

Recent reports suggested a substantial cellular activation in HCV-replicating cells and HCV-infected patients without external stimuli.12, 18, 23–25, 31 In chronically infected HCV patients, several groups reported an increase of IFN-stimulated genes before therapy and regardless of stimulation.24, 25, 31 These reports support our findings of the initially high IRF7 nuclear translocation in the absence of stimulation in HCV-replicating cells and suggest preactivation of IRF7 nuclear translocation due to HCV infection. A lower extent of induction of IRF7 nuclear localization after stimulation with TLR7 ligand may be also due to a higher baseline level, indicative of an already induced state of IRF7 activation, and therefore only remotely related to any altered function of TLR7 by HCV. The complexity of the IRF7-involving signaling pathways in HCV infection remain the target of future research.

In conclusion, we determined that HCV interferes with TLR7 expression and function. Our findings aid the understanding of the HCV-induced immune evasion mechanisms.

Acknowledgements

We thank Dr. Charlie Rice (Rockefeller University, New York, NY) for the generous gift of all cell lines and anti-NS5A antibody, Dr. Takaji Wakita (National Institute of Infectious Disease, Tokyo, Japan) for the the JFH-1 plasmid, and Dr. Angela Dolganiuc (University of Massachusetts Medical School) for constructive criticism of the manuscript.

Ancillary