ABSTRACT
The hepatitis C virus (HCV) co-opts a number of cellular elements – including proteins, lipids, and microRNAs – to complete its viral life cycle. The cellular RNA-binding protein poly(rC)-binding protein 1 (PCBP1) had previously been reported to bind the HCV genome 5’ untranslated region (UTR), but its importance in the viral life cycle has remained unclear. Herein, we aimed to clarify the role of PCBP1 in the HCV life cycle. Using the HCV cell culture (HCVcc) system, we found that endogenous PCBP1 knockdown decreased viral RNA accumulation yet increased extracellular virus titers. To dissect PCBP1’s specific role in the viral life cycle, we carried out assays for viral entry, translation, genome stability, RNA replication, virion assembly and egress. We found that PCBP1 did not affect viral entry, translation, RNA stability, or RNA replication in the absence of efficient virion assembly. To specifically examine virion assembly and egress, we inhibited viral RNA replication with an RNA-dependent RNA polymerase inhibitor and tracked both intracellular and extracellular viral titers over time. We found that when viral RNA accumulation was inhibited, knockdown of PCBP1 still resulted in an overall increase in HCV particle secretion. We therefore propose a model where endogenous PCBP1 limits virion assembly and egress, thereby indirectly enhancing viral RNA accumulation in infected cells. This model furthers our understanding of how cellular RNA-binding proteins modulate HCV genomic RNA utilization during the viral life cycle.
IMPORTANCE Hepatitis C virus (HCV) is a positive-sense RNA virus, and as such, its genome must be a template for multiple mutually exclusive steps of the viral life cycle, namely translation, RNA replication, and virion assembly. However, the mechanism(s) that regulate how the viral genome is used throughout the viral life cycle still remain unclear. A cellular RNA-binding protein – PCBP1 – had previously been reported to bind the HCV genome, but its precise role in the viral life cycle was not known. In this study, we found that depleting PCBP1 decreased viral RNA accumulation but increased virus secretion. We ruled out a role for PCBP1 in virus entry, translation, genome stability or RNA replication, and demonstrate that PCBP1 knockdown enhances virus secretion when RNA replication is inhibited. We conclude that PCBP1 normally prevents virus assembly and egress, which allows more of the viral genomic RNA to be available for translation and viral RNA replication.
INTRODUCTION
The hepatitis C virus (HCV) is an enveloped virus of the Flaviviridae family (genus: hepacivirus) that typically causes a persistent liver infection (1). Its ~9.6 kb single-stranded, positive-sense RNA genome contains a single open reading frame flanked by 5’ and 3’ untranslated regions (UTR). A highly structured, type 3 internal ribosomal entry site (IRES) in the 5’ UTR drives the translation of the viral polyprotein, which is subsequently processed into 10 mature viral proteins: 3 structural proteins (core, E1 and E2 glycoproteins), and 7 nonstructural (NS) proteins (p7, NS2, NS3, NS4A, NS4B, NS5A and NS5B) (2, 3). While the structural proteins form the nucleocapsid and viral envelope, the NS3-5B form the viral replicase, required for viral RNA replication (4). The p7, NS2, NS3 and NS5A proteins have also been implicated in viral genome packaging and the assembly of new virion particles (5–8). As a positive-sense RNA virus, the HCV genome itself must serve as a template for viral translation, genome replication, and packaging; however, the mechanisms that determine which process each viral RNA is engaged in at any given time have not been defined. Due to their limited coding capacity, viruses are highly dependent on the molecular machinery of the host cell; thus, it is likely that cellular components participate in regulation of the viral RNA during the HCV life cycle. While a number of cellular proteins and RNAs have been shown to interact with the HCV genome, their precise role(s) in the viral life cycle have yet to be defined.
The poly(rC)-binding protein 1 (PCBP1) is one of the three most abundant cellular RNA-binding proteins with a strong affinity for poly(rC), along with its paralogs hnRNP K and PCBP2 (9). These multifunctional proteins can regulate translation and enhance the stability of their cellular mRNA targets, which they interact with through their hnRNP K homologous (KH) domains (10). Notably, all three paralogs have been reported to bind to the HCV 5’ UTR (11–13). However, the degree to which each protein has been studied in the context of HCV infection varies significantly – while hnRNP K and PCBP2 have been fairly extensively studied and reported to play markedly different roles in the viral life cycle, the role of PCBP1 in the HCV life cycle has not been investigated in detail (14, 15). Beyond its interactions with the 5’ UTR, previous reports suggested that PCBP1 was not necessary for HCV IRES-mediated translation, but that knockdown of PCBP1 decreases HCV RNA accumulation during infection (16, 17).
Herein, we sought to clarify the role of PCBP1 in the HCV life cycle. Using a cell culture-adapted strain of HCV, we found that PCBP1 knockdown decreased viral RNA accumulation, yet led to an increase in virus secretion. By examining individual steps of the viral life cycle, we ruled out a role for PCBP1 in viral entry, translation, genome stability and viral RNA replication. Further analysis of the assembly step revealed that, similarly to its paralog hnRNP K, endogenous PCBP1 limits HCV virion assembly and release.
MATERIALS AND METHODS
Cell culture
Huh-7.5 human hepatoma cells were obtained from Charlie Rice (Rockefeller University) and maintained in complete media: Dulbecco’s Modified Eagle Media (DMEM) supplemented with inactivated 10% fetal bovine serum (FBS), 2 mM L-glutamine, and 1X MEM non-essential amino acids. Human embryonic kidney (293T) cells were kindly provided by Martin J. Richer (McGill University, Montreal, QC, Canada) and were maintained in DMEM supplemented with 10% FBS. All cell lines were maintained at 37°C/5% CO2 and were routinely screened for mycoplasma contamination.
Plasmids and viral RNAs
The pJFH-1T plasmid, encoding a cell culture-adapted Japanese Fulminant Hepatitis (JFH-1; HCV genotype 2a) with three adaptive mutations that increase viral titers in cell culture, was a gift from Rodney Russell (Memorial University of Newfoundland) (18). Plasmids pJ6/JFH1 FL RLuc WT (“RLuc-wt”) and pJ6/JFH-1 FL RLuc GNN (“RLuc-GNN”) bear full-length viral sequence derived from the J6 (structural genes) and JFH-1 (NS genes) isolates of HCV, with a Renilla luciferase (RLuc) reporter (5). The pJ6/JFH-1 mono RLuc-NS2 plasmid (“Δcore-p7”) – a truncated version of the Renilla reporter virus with a deletion of the structural genes through p7 – was a gift from Joyce Wilson (University of Saskatchewan) (19). The pJ6/JFH-1 FL RLuc-NS5A-GFP (“NS5A-GFP”) was created via overlapping PCR and subcloned using the AvrII and XbaI restriction sites, as previously described (20).
To make full-length uncapped viral RNAs, all plasmid templates were linearized and in vitro transcribed as previously described (21). The firefly luciferase (FLuc) mRNA was transcribed from the Luciferase T7 Control DNA plasmid (Promega) linearized using XmnI and in vitro transcribed using the mMessage mMachine T7 Kit (Life Technologies) according to the manufacturer’s instructions.
Generation of infectious HCV stocks
To generate viral stocks, 30 μg of in vitro transcribed JFH-1T RNA was transfected into Huh-7.5 cells using the DMRIE-C reagent (Life Technologies) according to the manufacturer’s instructions. Four days post-transfection, infectious cell supernatants were passed through a 0.45 μm filter and infectious viral titers were determined by focus-forming unit assay (18). Infectious virus was amplified for two passages through Huh-7.5 cells at a MOI of 0.1. Viral stocks were aliquoted and stored at −80°C until use.
Focus-forming unit (FFU) assays
One day prior to infection, 8-well chamber slides (Lab-Tek) were seeded with 4 × 105 Huh-7.5 cells/well. Infections were performed with 10-fold serial dilutions of viral samples in 100 μL for 4 h, after which the supernatant was replaced with fresh media. Three days post-infection, slides were fixed in 100% acetone and stained with anti-HCV core antibody (1:100, clone B2, Anogen), and subsequently with the AlexaFluor-488-conjugated anti-mouse antibody (1:200, ThermoFisher Scientific) for immunofluorescence analysis. Viral titers are expressed as the number of focus-forming units (FFU) per mL.
Extracellular virus titers were determined directly from cell supernatants, while intracellular virus titers were determined after cell pellets were subjected to lysis via four freeze-thaw cycles, removal of cellular debris via centrifugation, and recovery of virus-containing supernatants.
MicroRNAs and siRNA sequences
siGL3 (siCTRL): 5’-CUUACGCUGAGUACUUCGAUU-3’, siGL3* : 5’-UCGAAGUACUCAGCGUAAGUU-3’, miR122p2-8 (siCTRL for luciferase experiments): 5’- UAAUCACAGACAAUGGUGUUUGU-3’, miR122p2-8*: 5’-AAACGCCAUUAUCUGUGAGGAUA-3’ (22), siPCBP1: 5’- CUGUGUAAUUUCUGGUCAGUU-3’, siPCBP1*: 5’- CUGACCAGAAAUUACACAGUU-3’ (17) were all synthesized by Integrated DNA Technologies.
All microRNA and siRNA duplexes were diluted to a final concentration of 20 μM in RNA annealing buffer (150 mM HEPES pH 7.4, 500 mM potassium acetate, 10 mM magnesium acetate), and annealed at 37°C for 1 h and stored at −20°C. For all knockdown experiments, 50 nM siRNA transfections were conducted 2 days prior to infection or electroporation of viral RNAs. Transfections were conducted using the Lipofectamine RNAiMAX (Invitrogen) according to the manufacturer’s instructions with the modification that 20 μL of reagent were used to transfect a 10-cm dish of cells.
HCV and VSV pseudoparticles (HCVpp and VSVpp)
HCVpp consisting of a FLuc reporter lentiviral vector pseudotyped with the HCV E1 and E2 glycoprotein (from the H77, a genotype 1a strain) were a kind gift from John Law (University of Alberta) (23). To generate lentiviral vectors pseudotyped with the VSV-G glycoprotein (VSVpp), a 90% confluent 10-cm dish of 293T cells were transfected with 10 μg pPRIME-FLuc, 5 μg psPAX.2, and 2.5 μg pVSV-G plasmid with 10 μL Lipofectamine 2000 (Invitrogen) diluted in 4 mL Opti-MEM. Media was changed 4, 20, and 28 h post-transfection. At 48 h post-transfection, the cell culture media was passed through a 0.45 μm filter and stored at −80°C.
To assay for cell entry, HCVpp and VSVpp were diluted 1/3 in dilution media (1X DMEM, 3% FBS, 100 IU penicillin and 100 μg/mL streptomycin) with 20 mM HEPES and 4 μg/μL polybrene, and then introduced to Huh-7.5 cells by spinoculation at 1,200 rpm for 1 h at room temperature. The cells were left to recover at 37°C for at least 5 h before the pseudoparticle-containing media was changed for fresh complete Huh-7.5 media. In parallel, cells seeded in a 6-well plate were transfected with 1 μg of pPRIME-FLuc plasmid using Lipofectamine 2000 (Invitrogen) according to the manufacturer’s instructions. Three days post-spinoculation and transfection, cells were lysed in passive lysis buffer (Promega) and FLuc activity was assayed using the Dual Reporter Luciferase kit (Promega).
Electroporations
For each electroporation, 400 μL of resuspended cells (1.5 × 107 cell/mL) were mixed with 2 μg of FLuc mRNA and 5 μg of replicating (WT, Δcore-p7 or NS5A-GFP J6/JFH-1 RNA) or 10 μg GNN J6/JFH-1 RNA, and electroporated in 4-mm cuvettes at 270 V, 950 μF, and infinite resistance, optimized for the Bio-Rad GenePulser XCell (Bio-Rad). Electroporated cells were resuspended in complete Huh-7.5 media and transferred to 6-well plates for luciferase assays and protein analysis.
Inhibition of RNA replication by 2’CMA
Two days post-siRNA transfection, Huh-7.5 cells were infected with JFH-1T at a MOI of 0.05. Four to five hours post-infection, each plate of infected cells was split into 6-well plates. Three days post-infection, the media on these cells was changed for complete Huh-7.5 media with 5 μM 2’CMA (2’C-methyladenosine, Carbosynth), an HCV NS5B polymerase inhibitor, or DMSO control (24). Total RNA and intracellular virus samples were collected at 0, 6 and 12 h post-treatment, while cell culture supernatants were collected 6 and 12 h post-treatment. Protein samples were collected from untreated plates to assess PCBP1 knockdown efficiency by Western blot.
Western blot analysis
To collect total intracellular protein samples, cells were lysed in RIPA buffer (150 mM sodium chloride, 1% NP-40, 0.5% sodium deoxycholate, 0.1% SDS, 50 mM Tris pH 8.0, Complete Protease Inhibitor Cocktail (Roche)) and frozen at −80°C. Cellular debris was pelleted by centrifugation at 16,000 x g for 30 min at 4°C, and the supernatant was quantified by BCA Protein Assay (ThermoScientific). Ten micrograms of sample were loaded onto 10-12% SDS-PAGE gels. Samples were transferred onto Immobilon-P PVDF membranes (Millipore), blocked in 5% milk, and incubated overnight with primary antibodies diluted in 5% BSA: rabbit anti-PCBP1 (clone EPR11055, Abcam ab168378, 1:10,000); rabbit anti-actin (A2066, Sigma, 1:20,000); mouse anti-HCV core (clone B2, Anogen MO-I40015B, diluted 1:7,500); mouse anti-JFH-1 NS5A (clone 7B5, BioFront Technologies, 1:10,000). Blots were incubated for 1 hour with HRP-conjugated secondary antibodies diluted in 5% skim milk: anti-mouse (HAF007, R&D Systems, 1:25,000); anti-rabbit (111-035-144, Jackson ImmunoResearch Laboratories, 1:50,000) and visualized using enhanced chemiluminescence (ECL Prime Western Blotting Detection Reagent, Fisher Scientific).
RNA isolation and Northern blot analysis
Total RNA was harvested using TRIzol Reagent (ThermoFisher Scientific) according to the manufacturer’s instructions. Ten micrograms of total RNA were separated on a 1% agarose gel containing 1X 3-(N-morpholino)propanesulfonic acid (MOPS) buffer and 2.2 M formaldehyde and transferred to a Zeta-probe membrane (Bio-Rad) by capillary transfer in 20X SSC buffer (3 M NaCl, 0.3 M sodium citrate). Membranes were hybridized in ExpressHyb Hybridization Buffer (ClonTech) to random-primed 32P-labeled DNA probes (RadPrime DNA labelling system, Life Technologies) complementary to HCV (nt 84-374) and γ-actin (nt 685-1171). Autoradiograph band densities were quantified using Fiji (25).
RT-qPCR analysis
The iTaq Universal Probes One-Step kit (Bio-Rad) was used to perform duplex assays probing for the HCV genome (NS5B-FW primer: 5’-AGACACTCCCCTATCAATTCATGGC-3’; NS5B-RV primer: 5’-GCGTCAAGCCCGTGTAACC-3’; NS5B-FAM probe: 5’-ATGGGTTCGCATGGTCCTAATGACACAC-3’) and the GAPDH loading control (PrimePCR Probe assay with HEX probe, Bio-Rad). Each 20 μL reaction contained 500 ng of total RNA, 1.5 μL of the HCV primers and probe, and 0.5 μL of the GAPDH primers and probe. RT-PCR reactions were conducted in a CFX96 Touch Deep Well Real-Time PCR system (Bio-Rad). Genome copies were calculated using a standard curve and fold-differences in gene expression were calculated using the 2−ΔΔCt method (26).
Luciferase assays
For translation and replication assays, cells were washed in PBS and harvested in 100 μL of 1X passive lysis buffer (Promega). The Dual-Luciferase Assay Reporter Kit (Promega) was used to measure both Renilla and firefly luciferase activity according to the manufacturer's instructions with the modification that 25 μL of reagent were used with 10 μL of sample. All samples were measured in triplicate.
Data analysis
Statistical analyses were performed using GraphPad Prism v9 (GraphPad, USA). Statistical significance was determined by paired t-test to compare results obtained from multiple experiments, and by two-way ANOVA with Geisser-Greenhouse and Bonferroni corrections when more than two comparisons were applied at once. To calculate half-lives, a one-step decay curve using the least-squares regression was used, and error was reported as the asymmetrical (profile likelihood) 95% confidence interval of the half-life. To calculate virus accumulation and virus secretion rates, a simple linear regression was performed using the least squares regression method. The slope and standard error calculated for each regression represents the rate of virus accumulation or secretion.
RESULTS
PCBP1 plays a role in the HCV life cycle
The PCBP1 protein was previously reported to directly interact with the 5’ UTR of the HCV genome, and in an siRNA screen PCBP1 knockdown resulted in a reduction in HCV RNA accumulation (11, 17). Thus, we sought to further characterize the role of PCBP1 in the HCV life cycle. We began by assessing how PCBP1 knockdown affected the accumulation of cell culture-derived HCV (HCVcc), using the cell-culture adapted JFH-1T strain. Compared to the parental JFH-1 strain, JFH-1T has three adaptive mutations in the E2, p7 and NS2 coding-region, which enable it to produce higher viral titers in cell culture (18). We found that knockdown of endogenous PCBP1 resulted in an approximately 2.2-fold decrease in viral protein and RNA accumulation in Huh-7.5 cells (Figure 1A-C and Supplementary Figure 1). Interestingly, when we quantified intracellular and extracellular virions, we found that while intracellular titers were not significantly different between the PCBP1 knockdown and control conditions, extracellular titers were elevated, with an average increase of approximately 3.90-fold in the PCBP1 knockdown condition (Figure D-E). Thus, in line with previous findings, we found that PCBP1 knockdown decreased viral protein expression and intracellular viral RNA accumulation. Yet, despite this overall decrease, we observed an increase in extracellular (secreted) virus titers. These results imply that endogenous PCBP1 indeed plays a role in the HCV life cycle, although the precise step(s) influenced by PCBP1 remain unclear.
PCBP1 knockdown has no impact on HCV entry
Firstly, we explored whether PCBP1 knockdown had any effect on virus entry. To this end, we made use of the HCV pseudoparticle (HCVpp) system, which consists of lentiviral vectors with a firefly luciferase reporter gene pseudotyped with the HCV E1 and E2 glycoproteins (23). HCVpp enter cells by clathrin-mediated endocytosis after engaging with HCV-specific entry receptors; thus, to account for any changes in clathrin-mediated endocytosis, we used a vesicular stomatitis virus (VSV) pseudoparticle system (VSVpp) as a control. In addition, to verify that PCBP1 knockdown did not affect luciferase reporter gene expression, we assessed firefly luciferase expression from cells directly transfected with a FLuc reporter plasmid. In all cases, we found that depleting endogenous PCBP1 had no impact on luciferase activity (Figure 2). This suggests that PCBP1 knockdown does not affect FLuc reporter expression, clathrin-mediated endocytosis, or the HCV entry process.
PCBP1 knockdown has no impact of HCV translation or genome stability
PCBP1 was previously reported to bind the HCV 5’ UTR, which contains the viral IRES that drives translation of the viral polyprotein in the absence of most canonical translation initiation factors (2, 3, 11, 27). Furthermore, PCBP1 has been reported to contribute to the IRES-mediated translation of some cellular mRNAs such as Bag-1 and c-myc (28–30). Thus, it was plausible that PCBP1’s interactions with the viral 5’ UTR could affect viral protein expression by altering IRES-mediated translation. To assess PCBP1’s impact on HCV translation, we used full-length J6/JFH-1 RLuc reporter RNAs containing an inactivating mutation in the NS5B polymerase gene (GNN). The RLuc activity thus served as a direct measure of HCV IRES-mediated translation, and over time, this signal also served as a proxy measure for viral RNA stability. We found that the siPCBP1 and siCTRL conditions had similar RLuc activity at all timepoints (Figure 3). Moreover, the luciferase signal half-lives were nearly identical, with 2.68 h (95% CI 1.64 – 4.64) for the siPCBP1 condition and 2.71 h (95% CI 1.43 – 5.50) for the siCTRL condition. Thus, PCBP1 knockdown does not appear to affect either viral IRES-mediated translation or genome stability.
PCBP1 knockdown does not affect viral RNA replication
PCBP1 has been reported to promote viral RNA replication by binding to the 5’ UTR of several RNA viruses (31, 32). Thus, to specifically assess whether PCBP1 knockdown has an effect on HCV RNA replication, we assessed RLuc expression of replication-competent but assembly-deficient reporter RNAs (Figure 4). This includes the full-length WT J6/JFH-1 RLuc RNA, a subgenomic J6/JFH-1 replicon RNA containing a deletion of the structural protein genes (Δcore-p7), and a full-length J6/JFH-1 genome with a GFP insertion in the NS5A gene (NS5A-GFP), the latter previously shown to impair virion assembly without impairing viral RNA replication (7, 20). In all cases, we did not observe any significant differences in viral RNA accumulation between the siPCBP1 and siCTRL conditions (Figure 4A-C). This is seemingly in contrast to our findings using the JFH-1T system, where PCBP1 depletion decreased viral RNA accumulation (Figure 1B-C and Supplementary Figure 1). However, it is notable that all of the viral RNAs used in Figure 4 are defective in viral packaging, including the full-length WT J6/JFH-1 RLuc RNA (Figure 4A), likely due to the large RLuc reporter gene insertion (Supplementary Figure 2). Thus, taken together, our data suggests that PCBP1 has no effect on the viral RNA replication step of the HCV life cycle.
PCBP1 knockdown limits virus egress
Finally, since knockdown of PCBP1 did not seem to have a significant impact on HCV entry, translation, genome stability or viral RNA replication, we reasoned that PCBP1 was likely to be exerting an effect on virion assembly and/or secretion. We had previously established that PCBP1 knockdown did not have a significant effect on intracellular viral titers yet resulted in an overall increase in extracellular (secreted) virions (Figure 1D-E). Thus, to further investigate whether PCBP1 plays a role in viral packaging and/or egress, we used a nucleoside analog, 2’C-methyladenosine (2’CMA), to block viral RNA synthesis by the HCV NS5B RNA-dependent RNA polymerase and monitored viral titers over time (24). Similar to related studies using Zika virus, we reasoned that blocking viral RNA synthesis could allow us to assess viral packaging and egress in the absence of genomic RNA production (33).
To this end, we infected cells with JFH-1T and, three days post-infection, we replaced the culture media with media containing 5 μM 2’CMA (or DMSO, as a control) and collected total RNA, intracellular virions and extracellular virions over the next 12 h (Figure 5A). In agreement with our previous findings, we observed an overall reduction in JFH-1T viral RNA accumulation by day 3 post-infection in the siPCBP1 condition (0 h timepoint, Figure 5B). In addition, the 2’CMA treatment efficiently blocked viral RNA accumulation, which continued to increase under the control (DMSO) condition (6-12 h timepoints, Figure 5B). Additionally, we observed similar initial intracellular titers as well as similar 2’CMA-induced decreases in intracellular titers under both the siPCBP1 and siCTRL conditions (Figure 5C and D). In contrast, we observed a continued rise in extracellular titers as packaged viruses continued to egress out of the cell, with the PCBP1 knockdown resulting in a >3-fold greater virus secretion rate than the siCTRL condition during 2’CMA treatment (Figure 5E and F). Notably, since inhibiting viral RNA synthesis equalized the intracellular virus accumulation rates yet failed to equalize the virus secretion rates, these results suggest that PCBP1 modulates the virion assembly and egress steps of the HCV life cycle.
DISCUSSION
Herein, we investigated the role of PCBP1 in the HCV life cycle. We found that PCBP1 knockdown did not directly affect virus entry, translation, genome stability or viral RNA replication; but resulted in an increase in infectious particle secretion. Furthermore, since PCBP1 knockdown does not alter intracellular infectious particle accumulation yet reduces total intracellular viral RNA, our results suggest that PCBP1 knockdown promotes virion assembly and egress.
Previous studies had found that PCBP1 interacts with the HCV 5’ UTR in vitro (11, 16). Notably, the full 5’ UTR appeared to be critical for PCBP1 binding, as truncated fragments of the 5’ UTR were unable to bind to PCBP1 (11, 16). Somewhat more recently, PCBP1 was one of many host factors evaluated in a siRNA screen targeting cellular proteins predicted to interact with HCV (17). This screen revealed that PCBP1 knockdown decreased viral RNA accumulation approximately 2.3-fold, a result consistent with our observation that PCBP1 knockdown decreased JFH-1T viral RNA accumulation by approximately 2.2-fold (Figure 1D) (17). However, our systematic evaluation of the effect of PCBP1 knockdown on viral translation and viral RNA replication revealed that PCBP1 does not modulate either of these steps in the viral life cycle. Moreover, we did not observe any significant effects on viral entry or genome stability.
Initially, we were quite puzzled to find that PCBP1 did not affect viral RNA accumulation of a full-length WT J6/JFH-1 RLuc RNA; however, this RNA generates far fewer infectious viral particles than the JFH-1T strain we used for our infection experiments, likely due to the increased genome size due to the RLuc insertion resulting in reduced packaging efficiency (Supplementary Figure S2) (14). The fact that PCBP1 knockdown failed to have a significant impact on the luciferase reporter RNA is therefore consistent with a model whereby PCBP1 primarily acts at the level of virion assembly and egress (Figure 6). As such, knockdown of PCBP1 results in an overall decrease in the translating/replicating pool of viral RNAs and a concomitant increase in extracellular (secreted) virions. Since intracellular viral titers remained similar in PCBP1 knockdown and control conditions, a greater proportion of the intracellular viral RNA must represent viral RNA packaged into virions, implying that endogenous PCBP1 normally limits virion assembly. This model is further supported by our finding that 2’CMA equalized the rate of intracellular virus accumulation yet resulted in a virus secretion rate >3-fold greater during PCBP1 knockdown, suggesting that endogenous PCBP1 limits virion egress. Moreover, this model is consistent with the observed decreases in viral protein expression and negative-strand replicative intermediate accumulation (Figure 1B and Supplementary Figure 1), since increasing viral genome packaging into new virions would sequester the RNA from the translation and RNA replication machinery. Thus, endogenous PCBP1 normally limits viral assembly and egress, and PCBP1 knockdown therefore indirectly impairs viral protein synthesis and viral RNA replication by liberating the RNA for virion assembly and egress.
Notably, PCBP1 has been previously implicated in turnover of MAVS, a signal transduction protein directly downstream of RIG-I, which is known to limit HCV infection (34). However, all of our conclusions were drawn from experiments conducted in the Huh-7.5 cell line, which has well-documented defects in the RIG-I antiviral signalling pathway, and our own brief explorations of interferon induction and MAVS turnover did not reveal any significant differences between PCBP1 knockdown and control conditions (data not shown) (35). Moreover, the HCV NS3-4A protease also inactivates MAVS during infection to reduce antiviral signaling and recognition of the viral RNA (36, 37). However, as hepatocytes typically express RIG-I and MAVS, we cannot rule out the possibility that PCBP1 may modulate this pathway during infection in vivo.
Finally, PCBP1’s closely related paralogs, hnRNP K and PCBP2, have been more extensively studied during HCV infection. While the hnRNP K protein was shown to restrict infectious virion production, the PCBP2 protein has been suggested to modulate viral translation and RNA replication (14, 15). Since the PCBP1 amino acid identity is far more similar to PCBP2 (~80% identity) than hnRNP K (~24% identity), we were initially surprised that our findings for PCBP1 in HCV infection closely matched those reported for hnRNP K. Yet, while PCBP1 and PCBP2 have been shown to perform similar functions, they have also been reported to play distinct roles during poliovirus, VSV, and human immunodeficiency virus infection (31, 38, 39). In addition to the similarities we observed with PCBP1 and those previously reported with hnRNP K during HCV infection, our results and conclusions also echo those reported for the IGF2BP2/YBX-1 complex, METTL3/METTL14 N6-methyladenosine (m6A) writers, and the YTHDF m6A-binding proteins (14, 40, 41). These have all been reported to inhibit HCV infectious particle production, with no effect on viral translation or RNA replication – with the exception of the IGF2BP2/YBX-1 complex, which plays an additional role in facilitating viral RNA replication (40). It is currently unclear if PCBP1, hnRNP K, the YBX-1 complex or m6A modifications of the HCV genome inhibit virion assembly as components of a common pathway, or through distinct and/or additive mechanisms. Interestingly, high-throughput affinity capture and proximity ligation studies have found that PCBP1 interacts with hnRNP K, IGF2BP2, YBX-1, METTL3 and METTL14; although this has been demonstrated only in non-hepatic cell lines to date (42–44). Should these interactions be conserved during HCV infection, it seems plausible that PCBP1 could participate in these virion assembly inhibition pathways – or, conversely, that it may have an inhibitory role to promote one pathway over another. Future investigations will reveal whether these proteins function in an overlapping or a distinct manner; and are likely to improve our understanding of HCV virion assembly and how this process is regulated by cellular RNA binding proteins.
Taken together, our results support a model where endogenous PCBP1 limits HCV infectious particle production. By preventing virion assembly and egress, PCBP1 indirectly enhances viral RNA accumulation. The model presented herein helps to inform understanding of how cellular RNA-binding proteins modulate HCV genomic RNA utilization during the viral life cycle, specifically as it pertains to virion assembly. While the precise molecular mechanism employed by PCBP1 to inhibit HCV assembly and egress remains to be characterized, similar phenotypes reported for hnRNP K, IGF2BP2/YBX-1 and m6A modifications of the HCV genome offer promising leads for future investigations.
FUNDING INFORMATION
This research was supported by the Canadian Institutes for Health Research (CIHR) [MOP-136915 and PJT-169214]. S.E.C. was supported by the Canadian Network on Hepatitis C (CanHepC) training program, as well as a Vanier Canada Graduate Scholarship. In addition, this research was undertaken, in part, thanks to the Canada Research Chairs program (S.M.S.).
AUTHOR CONTRIBUTIONS
S.E.C. and S.M.S. designed the study; S.E.C. performed the experiments and analyzed the data, and S.E.C and S.M.S. wrote and edited the manuscript.
ACKNOWLEDGEMENTS
We would like to acknowledge Charlie Rice (Rockefeller University) for kindly providing the Huh-7.5 cells, pJ6/JFH FL RLuc WT and GNN plasmids; Rodney Russell (Memorial University) for providing JFH-1T; Mamata Panigrahi and Joyce Wilson (University of Saskatchewan) for the pJ6/JFH mono RLuc NS2 plasmid; Martin J Richer (McGill University) for the 293T cells; and John Law (University of Alberta) for supplying the HCVpp used herein. We are also grateful to Nathan Taylor, Julie Magnus and Carolina Camargo (McGill University) for technical support.