|Home | About | Journals | Submit | Contact Us | Français|
Nonstructural protein 5A (NS5A) of the hepatitis C virus (HCV) possesses multiple and diverse functions in RNA replication, interferon resistance, and viral pathogenesis. Recent studies suggest that NS5A is involved in the assembly and maturation of infectious viral particles; however, precisely how NS5A participates in virus production has not been fully elucidated. In the present study, we demonstrate that NS5A is a prerequisite for HCV particle production as a result of its interaction with the viral capsid protein (core protein). The efficiency of virus production correlated well with the levels of interaction between NS5A and the core protein. Alanine substitutions for the C-terminal serine cluster in domain III of NS5A (amino acids 2428, 2430, and 2433) impaired NS5A basal phosphorylation, leading to a marked decrease in NS5A-core interaction, disturbance of the subcellular localization of NS5A, and disruption of virion production. Replacing the same serine cluster with glutamic acid, which mimics the presence of phosphoserines, partially preserved the NS5A-core interaction and virion production, suggesting that phosphorylation of these serine residues is important for virion production. In addition, we found that the alanine substitutions in the serine cluster suppressed the association of the core protein with viral genome RNA, possibly resulting in the inhibition of nucleocapsid assembly. These results suggest that NS5A plays a key role in regulating the early phase of HCV particle formation by interacting with core protein and that its C-terminal serine cluster is a determinant of the NS5A-core interaction.
Hepatitis C virus (HCV) infection is a major public health problem and is prevalent in about 200 million people worldwide (27, 40, 42). Current protocols for treating HCV infection fail to produce a sustained virological response in as many as half of treated individuals, and many cases progress to chronic liver disease, including chronic hepatitis, cirrhosis, and hepatocellular carcinoma (15, 31, 35, 43).
HCV is a positive-strand RNA virus classified in the Hepacivirus genus within the Flaviviridae family (55). Its approximately 9.6-kb genome is translated into a single polypeptide of about 3,000 amino acids (aa), in which the structural proteins core, E1, and E2 reside in the N-terminal region. A crucial function of core protein is assembly of the viral nucleocapsid. The amino acid sequence of this protein is well conserved among different HCV strains compared to other HCV proteins. The nonstructural (NS) proteins NS3-NS5B are considered to assemble into a membrane-associated HCV RNA replicase complex. NS3 possesses the enzymatic activities of serine protease and RNA helicase, and NS4A serves as a cofactor for NS3 protease. NS4B plays a role in the remodeling of host cell membranes, probably to generate the site for the replicase assembly. NS5B functions as the RNA-dependent RNA polymerase. NS5A is known to play an important but undefined role in viral RNA replication.
NS5A is a phosphoprotein that can be found in basally phosphorylated (56 kDa) and hyperphosphorylated (58 kDa) forms (49). Comparative sequence analyses and limited proteolysis of recombinant NS5A have demonstrated that NS5A is composed of three domains (52). Domain I is relatively conserved among HCV genotypes compared to domains II and III. Analysis of the crystal structure of the conserved domain I that immediately follows the membrane-anchoring α-helix localized at the N terminus revealed a dimeric structure (53). The interface between protein molecules is characterized by a large, basic groove, which has been proposed as a site of RNA binding. In fact, its RNA binding property has been demonstrated biochemically (17). Domains II and III of NS5A are far less understood. Domain II contains a region referred to as the interferon sensitivity determining region, and this region and its C-terminal 26 residues have been shown to be essential for interaction with the interferon-induced, double-stranded RNA-dependent protein kinase (6-10, 38, 39, 48). Domain III includes a number of potential phosphoacceptor sites and is most likely involved in basal phosphorylation. This domain tolerates insertion of large heterologous sequences such as green fluorescent protein (GFP) and is not required for function of NS5A in HCV RNA replication (1, 34). However, a study with the recently established productive HCV cell culture system using genotype 2a isolate JFH-1 (28, 56, 58) demonstrated that while insertion of GFP within the NS5A region does not affect RNA replication, it does produce marked decreases in the production of infectious virus particles (41). This suggests that the C-terminal region of NS5A may affect virus particle production independent of RNA replication. Recently, Miyanari et al. reported that the association of core protein with the NS proteins and replication complexes around lipid droplets (LDs) is critical for producing infectious viruses (33).
In the present study, we demonstrated that NS5A is a prerequisite for HCV particle production via its interaction with core protein, and we identified serine residues in the C-terminal region of NS5A that play an important role in virion production. Substitution of the serine residues with alanine residues inhibited not only the interaction of NS5A with core protein but also HCV RNA-core association and led to a decrease in HCV particle production with no effect on RNA replication.
Plasmids pJFH1, which contains the full-length JFH-1 cDNA downstream of the T7 RNA promoter sequence, and pSGR-JFH1/Luc, in which the neomycin resistance gene of pSGR-JFH1 has been replaced by the firefly luciferase reporter gene, have been previously described (24, 56). To generate the fluorochrome gene-tagged full-length JFH-1 plasmid, pJFH1/NS5A-GFP, the region encompassing the RsrII site of NS5A and the BsrGI site of NS5B was amplified by PCR, the amplification product was cloned into pGEM-T Easy vector (Promega, Madison, WI), and the resultant plasmid was designated pGEM-JFH1/RsrII-BsrGI. A GFP reporter gene was amplified by PCR from pGreen Lantern-1 (Invitrogen, Carlsbad, CA) with primers containing the XhoI sequence and inserted, after restriction digestion with XhoI, into the XhoI site of pGEM-JFH1/RsrII-BsrGI. The resulting plasmid was digested by RsrII and BsrGI and ligated into pJFH1 similarly digested by RsrII and BsrGI to produce pJFH1/NS5A-GFP. For generation of the fluorochrome gene-tagged subgenomic reporter plasmid, pJFH1/NS5A-GFP was digested by RsrII and SnaBI and ligated into pSGR-JFH1/Luc similarly digested by RsrII and SnaBI. The mutations in the NS5A gene were generated by oligonucleotide-directed mutagenesis (57). To construct plasmids expressing N-terminally FLAG-tagged HCV core protein or hemagglutinin (HA)-tagged NS5A, DNA fragments encoding core protein or NS5A (wild type or mutants) were generated from the full-length JFH-1 cDNA by PCR. The core protein coding sequence, together with a FLAG sequence linked to its N terminus, was cloned into the pCAGGS vector (37). The coding sequences of NS5A, together with an HA sequence linked to their N termini, were also cloned into pCAGGS vectors. All PCR products were confirmed by automated nucleotide sequencing with an ABI Prism 3130 Avant Genetic Analyzer (Applied Biosystems, Tokyo, Japan).
The human hepatoma cell line, Huh-7, and JFH1/4-1 cells, which are Huh-7 cells carrying a subgenomic replicon of JFH-1 (32), were maintained in Dulbecco's modified Eagle's medium (DMEM) supplemented with minimal essential medium nonessential amino acids (Invitrogen), 100 units/ml of penicillin, 100 μg/ml of streptomycin, and 10% fetal bovine serum (FBS) at 37°C in a 5% CO2 incubator. Huh/c-p7 cells, which are Huh-7 cells stably expressing the proteins core to p7 derived from the JFH-1 strain (18), were incubated in DMEM containing 300 μg/ml of zeocin (Invitrogen). HCV particles derived from JFH-1 were produced by transient transfection of Huh-7 cells with in vitro transcribed RNA, as described previously (56, 58). Recombinant vaccinia virus strain DIs, which expresses the bacteriophage T7 RNA polymerase under the control of the vaccinia virus early/late promoter P7.5, was generated and propagated as previously described (19).
For coexpression of FLAG-tagged core protein and HA-tagged NS5A, cells were seeded onto 35-mm wells of a six-well cell culture plate and cultured overnight. Plasmid DNAs (2 μg) were transfected into cells using TransIT-LT1 transfection reagent (Mirus, Madison, WI). Cells were harvested at 48 h posttransfection, washed three times with 1 ml of ice-cold phosphate-buffered saline (PBS), and suspended in 0.25 ml lysis buffer (20 mM Tris-HCl [pH 7.4] containing 135 mM NaCl, 1% Triton X-100, 0.05% sodium dodecyl sulfate [SDS], and 10% glycerol) supplemented with 50 mM NaF, 5 mM Na3VO4, 1 μg/ml leupeptin, and 1 mM phenylmethylsulfonyl fluoride (PMSF). Cell lysates were sonicated at 4°C for 5 min, incubated for 30 min at 4°C, and centrifuged at 14,000 × g for 5 min at 4°C. After preclearing, the supernatant was immunoprecipitated with 10 μl of anti-FLAG M2-agarose beads (Sigma, St. Louis, MO). For expression of the full-length HCV polyprotein, Huh-7 cells transfected with 10 μg of in vitro transcribed RNAs by electroporation were resuspended in 20 or 30 ml of culture medium, and 10-ml aliquots were seeded into 100-mm culture dishes. At 72 h posttransfection, the cells were incubated in 0.5 ml of lysis buffer (20 mM Tris-HCl [pH 7.4] containing 135 mM NaCl, 1% Triton X-100, 0.5% sodium deoxycholate, and 10% glycerol) supplemented with 50 mM NaF, 5 mM Na3VO4, 1 μg/ml leupeptin, and 1 mM PMSF. After preclearing, the supernatant was immunoprecipitated with 5 μg of polyclonal anti-NS5A antibody (34a) or polyclonal anti-C/EBPβ antibody (Santa Cruz Biotechnology, Santa Cruz, CA), and 20 μl of protein G-agarose beads (Invitrogen). The immunocomplex was precipitated with the beads by centrifugation at 800 × g for 30 s and then was washed five times with lysis buffer by centrifugation. The proteins binding to the beads were boiled in 20 μl of SDS sample buffer and then subjected to SDS-12.5% polyacrylamide gel electrophoresis (PAGE). The proteins were transferred onto a polyvinylidene difluoride membrane (Immobilon; Millipore, Bedford, MA) and then reacted with a primary antibody and a secondary horseradish peroxidase-conjugated antibody. The immunocomplexes were visualized with an ECL Plus Western Blotting Detection System (GE Healthcare, Buckinghamshire, United Kingdom) and detected using an LAS-3000 imaging analyzer (Fujifilm, Tokyo, Japan).
Plasmid DNAs were digested with XbaI and treated with mung bean nuclease (New England Biolabs, Ipswich, MA) to remove the four terminal nucleotides, resulting in the correct 3′ end of the HCV cDNA. Digested DNAs were purified and used as templates for RNA synthesis. HCV RNA was synthesized in vitro using a MEGAscript T7 kit (Ambion, Austin, TX). Synthesized RNA was treated with DNase I (Ambion), followed by acid guanidinium thiocyanate-phenol-chloroform extraction to remove any remaining template DNA. Synthesized HCV RNAs were used for electroporation. Trypsinized Huh-7 cells were washed with Opti-MEM I reduced-serum medium (Invitrogen) and resuspended at 3 × 106 cells/ml with Cytomix buffer (54). RNA was mixed with 400 μl of cell suspension and transferred into an electroporation cuvette (Precision Universal Cuvettes; Thermo Hybaid, Middlesex, United Kingdom). Cells were then pulsed at 260 V and 950 μF using a Gene Pulser II unit (Bio-Rad, Hercules, CA). Transfected cells were immediately transferred onto six-well culture plates or 100-mm culture dishes.
Cells were harvested at different time points posttransfection of subgenomic reporter replicons and lysed in passive lysis buffer (Promega). The luciferase activity in cells was determined using a luciferase assay system (Promega).
HCV core protein in transfected cells or cell culture supernatants was quantified using a highly sensitive enzyme immunoassay (Ortho HCV antigen ELISA Kit; Ortho Clinical Diagnostics, Tokyo, Japan). To determine intracellular core protein amounts, cell lysates were prepared as described previously (41). To determine the efficiency of core protein release, the ratio of extracellular core protein to total core protein (the sum of intra- and extracellular core protein amounts) was calculated.
Culture supernatants were harvested 72 h posttransfection, and virus titers were determined by a 50% tissue culture infectious dose (TCID50) assay as described previously (28, 46). Virus titration was performed by seeding naïve Huh-7 cells in 96-well plates at a density of 1 × 104 cells/well. Samples were serially diluted fivefold in complete growth medium and used to infect the seeded cells (six wells per dilution). At 72 h after infection, the inoculated cells were fixed and immunostained with a mouse monoclonal anti-core protein antibody (2H9) (56), followed by an Alexa Fluor 488-conjugated anti-mouse immunoglobulin G (IgG) (Invitrogen). Wells that showed at least one core protein-expressing cell was counted as positive. Cell-associated infectivity was determined essentially as described previously (12, 47). Briefly, cells were extensively washed with PBS, scraped, and centrifuged for 3 min at 120 × g. Cell pellets were resuspended in 1 ml of DMEM containing 10% FBS and subjected to four cycles of freezing and thawing using dry ice and a 37°C water bath. Samples were then centrifuged at 2,400 × g for 10 min at 4°C to remove cell debris, and cell-associated infectivity was determined by TCID50 assay.
Metabolic labeling of cells and radioimmunoprecipitation analysis were performed as described by Huang et al. (17) with some modifications. A total of 4 × 105 Huh-7 cells were seeded onto each well of six-well cell culture plates and cultured overnight. A 2-μg amount of subgenomic replicon DNAs carrying defined NS5A mutations was transfected into cells using TransIT-LT1 transfection reagent, and at 12 h posttransfection the cells were then infected at a multiplicity of infection of 10 with recombinant vaccinia viruses expressing the T7 RNA polymerase. After 40 h of transfection, cells were incubated in methionine- and cysteine-deficient DMEM (Invitrogen) or phosphate-deficient DMEM (Invitrogen) for 2 h and labeled for 6 h with [35S]methionine and [35S]cysteine (200 μCi/well; GE Healthcare) or [32P]orthophosphate (250 μCi/well; GE Healthcare). The cells were then washed twice with cold PBS and lysed with SDS lysis buffer (50 mM Tris-HCl [pH 7.6], 0.5% SDS, 1 mM EDTA, 20 μg/ml of PMSF). The cell lysates were passed through a 27-gauge needle several times to shear cellular DNA. After a 10-min incubation at 75°C, the lysates were clarified by centrifugation and diluted fivefold with HNAET buffer (50 mM HEPES [pH 7.5], 150 mM NaCl, 0.67% bovine serum albumin, 1 mM EDTA, 0.33% Triton X-100). After preclearing by incubation with 20 μl of protein G-agarose beads for 1 h at 4°C, the supernatant was incubated with 2 μg of rabbit polyclonal anti-NS5A antibody overnight at 4°C. A 20-μl aliquot of protein G agarose beads was further added and incubated for 2 h at 4°C. The cell pellets were washed three times with 0.5 ml of HNAETS buffer (HNAET containing 0.5% SDS), followed by washing once with 0.5 ml of HNE buffer (50 mM HEPES [pH 7.5], 150 mM NaCl and 1 mM EDTA). After treatment with or without λ protein phosphatase (New England Biolabs), the cell pellets were suspended in 20 μl of SDS sample buffer and boiled for 10 min. The proteins were resolved on 10% SDS-polyacrylamide gels and analyzed by autoradiography.
All steps were carried out at 4°C in the presence of a protease inhibitor cocktail (Complete; Roche, Mannheim, Germany) as described previously (20), with some modifications. Cells were suspended in four cell volumes of homogenization buffer (50 mM NaCl, 10 mM triethylamine [pH 7.4], 1 mM EDTA), snap frozen in liquid nitrogen, stored at −80°C, and thawed in a water bath at room temperature. Supernatants (0.4 ml) were layered on linear 10-ml iodixanol gradients from 2.5 to 25% and centrifuged at 37,000 rpm for 3.5 h in an SW41 rotor (Beckman, Fullerton, CA), followed by collection of 0.8-ml fractions from the top. Each fraction was concentrated by Centricon YM30 (Millipore), separated by SDS-PAGE, and immunoblotted with a rabbit polyclonal anti-calnexin antibody (Stressgen Biotechnologies, Victoria, Canada), a mouse monoclonal anti-adipose differentiation-related protein (ADRP) antibody (Progen Biotechnik, Heidelberg, Germany), or a rabbit polyclonal anti-NS5A antibody. The core protein amount in each fraction was also determined by enzyme-linked immunosorbent assay (ELISA).
The process of cell lysis to RNA purification was carried out essentially as described by Johnson et al. (21) with some modifications. A total of 3 × 106 Huh-7 cells were transfected with 10 μg of in vitro transcribed HCV RNAs and resuspended in 20 or 30 ml of culture medium, after which 10-ml aliquots were seeded into 100-mm culture dishes. At 72 h posttransfection, the cells were scraped and incubated in 500 μl of hypotonic buffer (10 mM HEPES [pH 7.6], 1.5 mM MgCl2, 10 mM KCl, 0.2 mM PMSF) per dish. The cells were passed through a 20-gauge needle several times, lysed with Nonidet P-40 at a final concentration of 1%, and incubated on ice for an additional 10 min. After centrifugation at 4,000 × g at 4°C for 15 min, glycerol was added to the supernatants at a final concentration of 5%. The cell lysates were incubated with 20 μl of protein G-agarose beads for 30 min at room temperature. After the cell lysates were removed from protein G-agarose beads, 5 μg of mouse monoclonal anti-core protein antibody or normal mouse IgG (Sigma) as a negative control was added, and samples were incubated for an additional 1 h at room temperature. A 20-μl aliquot of protein G-agarose beads per sample was added to the cell lysates and incubated for 1 h. After incubation, the beads were washed three times with wash buffer (10 mM Tris-HCl [pH 7.6], 100 mM KCl, 5 mM MgCl2, and 1 mM dithiothreitol) and eluted in 100 μl of elution buffer (50 mM Tris-HCl [pH 8.0], 1% SDS, and 10 mM EDTA) at 65°C for 10 min. After treatment with 100 μg of proteinase K at 37°C for 30 min, the RNAs in immunocomplexes were isolated by acid guanidinium thiocyanate-phenol-chloroform extraction. Reverse transcriptase PCR (RT-PCR) was carried out using random hexamer and Superscript II RT (Invitrogen), followed by nested PCR with LA Taq DNA polymerase (TaKaRa, Shiga, Japan) and primer sets amplifying the fragments of nucleotides (nt) 129 to 2367 and nt 7267 to 9463 of the JFH-1 genome. To amplify the fragment of nt 129 to 2367, the sense primer 5′-CTGTGAGGAACTACTGTCTT-3′ and the antisense primer 5′-TCCACGATGTTCTGGTGAAG-3′ were used for first-round PCR; the sense primer 5′-CGGGAGAGCCATAGTGG-3′ and the antisense primer 5′-CATTCCGTGGTAGAGTGCA-3′ were used for second-round PCR. To amplify the fragment of nt 7267 to 9463, the sense primer 5′-GTCCAGGGTGCCCGTTCTGGACT-3′ and the antisense primer 5′-GCGGCTCACGGACCTTTCAC-3′ were used for first-round PCR; the sense primer 5′-CACCGTTGCTGGTTGTGCT-3′ and the antisense primer 5′-GTGTACCTAGTGTGTGCCGCTCTA-3′ were used for second-round PCR.
Cells incubated for 3 days after transfection with JFH-1 RNAs were seeded in an eight-well chamber slide (BD Biosciences, San Jose, CA) and cultured overnight. The adherent cells were washed twice with PBS and fixed with 4% paraformaldehyde at room temperature. After a washing step with PBS, the cells were permeabilized with PBS containing 0.3% Triton X-100 and 2% FBS for 1 h at room temperature and stained with a rabbit polyclonal anti-NS5A antibody and a mouse monoclonal anti-core protein antibody. The fluorescent secondary antibodies were Alexa Fluor 488- or Alexa Fluor 555-conjugated anti-rabbit or anti-mouse IgG antibodies (Invitrogen). Analyses of JFH-1 were performed on a Zeiss confocal laser scanning microscope LSM 510 (Carl Zeiss, Oberkochen, Germany).
As demonstrated in a previous study, insertion of GFP into the NS5A C terminus does not significantly affect viral RNA replication but reduces the generation of infectious HCV particles (41). The C-terminal region of NS5A contains highly conserved serine residues that are involved in basal phosphorylation (1, 23, 49). To examine the involvement of the serine clusters (cluster 3-A [CL3A] and cluster 3-B [CL3B]) in the C-terminal region of NS5A in HCV particle production, we created mutated HCV genomes as well as subgenomic replicons carrying alanine substitutions for the conserved serine residues at aa 2384, 2388, 2390, and 2391 (residues are numbered according to the positions within the original JFH-1 polyprotein) (CL3A/SA); at aa 2428, 2430, and 2433 (CL3B/SA); or an in-frame deletion spanning aa 2384 to 2433 (Δ2384-2433) (Fig. (Fig.1).1). A construct with an in-frame insertion of GFP (NS5A-GFP) was also generated as described previously for the Con1 isolate (34).
First, we analyzed the effects of the NS5A mutations on HCV RNA replication using a transient RNA replication assay using subgenomic luciferase reporter replicons (Fig. (Fig.2A)2A) and found that the serine-to-alanine substitutions (CL3A/SA and CL3B/SA) did not affect viral RNA replication. NS5A-GFP and Δ2384-2433 slightly reduced RNA replication, indicating that the mutations of the NS5A C terminus tested in this study do not critically affect RNA replication, which is consistent with previous reports (1, 34, 51).
Next, the phosphorylation status of the mutated NS5A was analyzed as described in Materials and Methods (Fig. (Fig.2B).2B). NS5A was isolated from radiolabeled cells by IP and analyzed either directly by SDS-PAGE or after treatment with λ protein phosphatase. Analysis of 32P-radiolabeled proteins revealed that the CL3A/SA, CL3B/SA, and Δ2384-2433 mutations resulted in marked reduction of basal phosphorylation (Fig. (Fig.2B,2B, compare lane 1 with lanes 3, 5, and 7 in the top panel). All 32P-labeled NS5A proteins were sensitive to treatment with phosphatase (lanes 2, 4, 6, and 8). The possibility that loss of signal after dephosphorylation was due to contaminating proteases present in the phosphatase preparations can be ruled out because no degradation of the 35S-labeled proteins was observed (Fig. (Fig.2B,2B, bottom panel). These results suggest that mutations in the C-terminal serine cluster of NS5A impair basal phosphorylation but have no significant effect on viral RNA replication.
To analyze HCV particle production from cells transfected with the in vitro transcribed viral genomic RNAs, we harvested supernatants and cells at 4, 24, 48, 72, and 96 h posttransfection and measured the amounts of core protein. As shown in Fig. Fig.3A,3A, comparable amounts of core proteins were detected in all transfected cells 4 h after transfection, reflecting unchanged transfection efficiencies, and the kinetics of intracellular core protein levels was similar among transfectants. By contrast, core protein released from cells transfected either with the mutated genome of CL3B/SA, Δ2384-2433, or NS5A-GFP was more than 10-fold lower than that for the wild-type JFH-1 or CL3A/SA (Fig. (Fig.3B).3B). Figure Figure3C3C shows the efficiency of core protein release from each transfectant, which is expressed as a percentage of the extracellular core protein level relative to the amount of total core protein (the sum of intra- and extracellular core protein). Core protein release efficiency with the wild type and CL3A/SA was 2 to 13% at 48 to 96 h after transfection, while only 1% or less of core protein was released in the cases of CL3B/SA, Δ2384-2433, and NS5A-GFP strains.
To further investigate production and release of infectious virus particles, naïve Huh-7 cells were infected with culture supernatants of cells harvested 72 h posttransfection, and infectious virus titers were determined by TCID50 assay at 72 h after infection. Figure Figure3D3D shows that release of infectious virus particles from cells transfected with the genome of CL3B/SA or Δ2384-2433 mutants was markedly reduced (about 10,000-fold) compared to that from wild-type- or CL3A/SA-transfected cells (white bars). To examine whether such a decrease in infectious HCV in the culture supernatants was attributable to defective virion assembly or impaired release of virions, we determined cell-associated infectivity (Fig. (Fig.3D).3D). Production of intracellular infectious virions in CL3B/SA- and Δ2384-2433-transfected cells was strongly impaired in comparison with that in wild-type-transfected (~1,000-fold) and CL3A/SA-transfected (~100-fold) cells. Thus, the results suggest a potential role for the serine cluster at aa 2428, 2430, and 2433 of NS5A in assembly of infectious HCV particles. Among the NS5A mutations tested, CL3B/SA is of particular interest because this mutation leads to a marked reduction in HCV production with no impact on viral RNA replication.
Miyanari et al. reported that the association of core protein with NS proteins is critical for infectious HCV production and that mutations of the core protein and NS5A that cause these proteins to fail to associate with each other impair the production of infectious virus (33). Based on these observations and the findings noted above, we hypothesize that NS5A plays a key role in recruiting viral RNA, which is synthesized at the viral replication complex, to nucleocapsid formation via interaction between the NS5A C-terminal region and the core protein. To prove this, we analyzed the interaction of NS5A with the core protein by coimmunoprecipitation experiments. HA-tagged NS5A constructs carrying defined mutations were generated (Fig. (Fig.1)1) and coexpressed with the FLAG-tagged core protein in Huh-7 cells. As shown in Fig. Fig.4A,4A, coimmunoprecipitation of NS5A with the core protein was observed in cells expressing the wild-type NS5A and the CL3A/SA-mutated NS5A, but the amount of immunoprecipitated NS5A in the CL3A/SA-expressing cells was slightly lower than that in the wild-type-expressing cells. In contrast, the CL3B/SA- or the Δ2384-2433-mutated NS5A coimmunoprecipitated with the core protein only slightly or not at all.
We further examined the interaction of NS5A with core protein in cells expressing HCV genomes. At 72 h posttransfection with the wild type or CL3B/SA, cells were harvested and immunoprecipitated with an anti-NS5A antibody or an anti-C/EBPβ antibody as a negative control, followed by immunoblotting. Under these experimental conditions, the amount of extracellular core protein released from cells transfected with the CL3B/SA genome was about 10-fold lower than that for the wild type, although comparable amounts of intracellular core protein were observed in both transfectants (Fig. (Fig.4B,4B, left panels). As shown in the right panels of Fig. Fig.4B,4B, the core protein was specifically coimmunoprecipitated with NS5A in cells expressing the wild-type JFH-1 genome but not with the mutated NS5A in cells expressing the CL3B/SA genome. These results demonstrate that NS5A interacts with the core protein in cells producing infectious particles and that serine residues at aa 2428, 2430, and 2433 are important to the success of this interaction.
To further determine the critical residues in the C-terminal serine cluster of NS5A responsible for HCV particle production, we replaced one or two serine residues in the region with alanine (Fig. (Fig.1)1) and investigated which serine-to-alanine substitution influenced HCV particle production. Core protein levels in cells transfected with any construct were comparable over 4 days after transfection, indicating similar efficiencies of transfection and RNA replication from each construct (data not shown). As shown in Fig. Fig.5A,5A, we observed a slight delay in the kinetics of core protein release from cells transfected with the single-substitution genomes, S2428A, S2430A, and S2433A, up to 48 or 72 h posttransfection. However, core protein release from these cells reached comparable levels to that for the wild type at 96 h after transfection. In the cases of the double-substitution mutants (Fig. (Fig.5B),5B), core protein release from cells transfected with the double-substitution genomes was markedly reduced, with 10- to 30-fold decreases compared to that for wild type observed. The kinetics of core protein release were similar to that for CL3B/SA.
Interaction of NS5A carrying single or double serine-to-alanine substitutions with the core protein was investigated by coimmunoprecipitation analysis using HA-tagged NS5A constructs. NS5A mutants carrying a single substitution were coimmunoprecipitated with the core protein (Fig. (Fig.5C),5C), while none of the double-substitution NS5A mutants or the triple-substitution mutant, CL3B/SA, coimmunoprecipitated with the core protein (Fig. (Fig.5D).5D). These results suggest that at least two serine residues in the C-terminal serine cluster of NS5A (aa 2428, 2430, and 2433) are necessary for the interaction between NS5A and the core protein as well as for regulation of HCV particle production and that there is positive correlation between their interaction and the amount of core protein released.
A consequence of phosphorylation is the addition of negative charge to a protein. In some cases, phosphoserine can be mimicked by glutamic or aspartic acid (14). To determine whether the introduction of negative charges into aa 2428, 2430, and 2433 instead of phosphoserines positively regulates the interaction of NS5A with the core protein and virus production, we replaced the serine residues with glutamic acid residues and constructed the CL3B/SE and S2428/2430E mutants (Fig. (Fig.6A).6A). Cells transfected with the double-glutamic acid substitution, S2428/2430E, exhibited similar kinetics to the wild-type-transfected cells and released ~22-fold more core protein than S2428/2430A-transfected cells by 96 h posttransfection (Fig. (Fig.6B).6B). In contrast, the transfectant with the triple glutamic acid substitution, CL3B/SE, showed similar trends to that of CL3B/SA. In the coimmunoprecipitation experiments with FLAG-tagged core protein and HA-tagged NS5A constructs (Fig. (Fig.6C),6C), S2428/2430E, but not S2428/2430A, restored the ability of NS5A to interact with the core protein up to a similar level to that of wild type. As expected, neither CL3B/SE nor CL3B/SA coimmunoprecipitated with the core protein. Taken together, these results indicate that negative charges at aa 2428 and 2430 preserve the ability of NS5A to interact with the core protein and positively regulate virus production. However, the data of the CL3B/SE mutant indicate that it is likely that negative charges alone are not sufficient to enhance either the interaction of NS5A with the core protein or virus production.
The coimmunoprecipitation experiments described above indicate that the wild-type NS5A but not the CL3B/SA mutant interacts with the core protein. To evaluate the NS5A-core protein interaction in intact cells, we examined the subcellular localization of NS5A with the core protein by immunofluorescence analysis. NS5A colocalized with the core protein in cells transfected with the JFH-1 wild type (Fig. (Fig.7A),7A), whereas their colocalization was rarely observed in cells transfected with the CL3B/SA RNA (Fig. (Fig.7B7B).
To further analyze the subcellular compartments for the localization of NS5A and core protein in cytoplasmic membrane structures, including the endoplasmic reticulum (ER) and LDs, we performed subcellular fractionation studies as described in Materials and Methods. The iodixanol gradient was collected from the top to the bottom into 12 fractions (fractions 1 to 12). As shown in Fig. Fig.7C,7C, an ER marker, calnexin, was found in fractions 7 to 12 and was localized primarily in fractions 11 and 12. In contrast, ADRP, a cellular marker for LDs, was mainly observed in fractions 4 to 7. These two markers were equally distributed among cells analyzed (data not shown). The distribution of the wild-type NS5A was found in fractions 4 to 7, which was parallel to the fractionation profile of ADRP. The CL3B/SA-mutated NS5A was more broadly distributed and was also observed in heavier fractions than the wild-type NS5A, which was analogous to distribution of NS5A expressed in JFH1/4-1 cells bearing subgenomic replicons. The core protein in cells expressing the JFH-1 wild type, the CL3B/SA mutant, and in Huh/c-p7 cells that express JFH-1 structural proteins was distributed in a similar fashion, indicating that the distribution of core protein is not affected by NS5A mutation. The fractionation profile of the core protein, with a peak in fraction 4 or 5, was similar to that of the wild-type NS5A or ADRP but not to that of the CL3B/SA-mutated NS5A or calnexin, suggesting that core protein interacts with the wild-type NS5A in LD fractions, which is consistent with previous reports (33, 44, 45).
To further address our hypothesis regarding involvement of NS5A in recruiting viral RNA to nucleocapsid formation, we analyzed the association of the core protein with HCV RNA in wild-type- or CL3B/SA-expressing cells by IP-RT-PCR (Fig. (Fig.8).8). Both cell lysates were immunoprecipitated with an anti-core protein antibody or a negative control, mouse IgG. Total RNA prepared from each immunoprecipitate was subjected to RT-PCR in order to detect HCV RNA. The amounts of immunoprecipitated core protein (Fig. (Fig.8,8, lower panel) as well as the expression of HCV RNA (Fig. (Fig.8,8, upper panels, Input) were comparable in both cells. In cells expressing the wild-type JFH-1 genome, the viral RNAs covering the 5′ terminal 2.2-kb as well as the 3′ terminal 2.2-kb regions were detected in immunoprecipitates obtained with the anti-core protein antibody but not with the mouse IgG. In contrast, in cells expressing the CL3B/SA genome, HCV RNA was not detected in the immunoprecipitates with either antibody. These results demonstrate that HCV RNA associates with the core protein in cells where NS5A interacts with core protein (JFH-1 wild type) but not in cells where their interaction is impaired (CL3B/SA).
In the present study, we demonstrated the involvement of NS5A in the production of HCV particles via the interaction of NS5A with the core protein and identified its C-terminal serine cluster 3-B (aa 2428, 2430, and 2433), which is implicated in basal phosphorylation, as a key element for the interaction of NS5A with the core protein and for infectious virus production. Serine-to-alanine substitutions at the cluster, which have no impact on viral RNA replication, inhibit the interaction between NS5A and the core protein, thereby indicating that there is a connection between NS5A-core protein association and virus production. Finally, CL3B mutation leads to impairment of the association of the core protein with HCV RNA and, therefore, possibly RNA encapsidation.
Several reports have indicated that viral NS proteins are involved in the virion assembly of Flaviviridae viruses (25, 29, 30, 33). For instance, mutations in yellow fever virus NS2A block production of infectious virus, and this perturbation can be released by a suppressor mutation in NS3 (25), while the hydrophobic residues of Kunjin virus NS2A required for virus assembly have been mapped (26). Miyanari et al. have shown that HCV core protein recruits NS proteins to the LD-associated membranes and that the NS proteins around the LDs participate in the assembly of infectious viral particles (33). Furthermore, during preparation of the current article, two studies regarding participation of NS5A in the assembly of HCV particles were published. Appel et al. have demonstrated the essential role of domain III of NS5A in the formation of infectious particles, and deletions in this domain that disrupt colocalization of NS5A and the core protein abrogate virion production (2). Tellinghuisen et al. identified a serine residue in domain III as a key determinant for viral particle production (50). However, the mechanism by which NS proteins participate in virus assembly or the role of the interaction between structural and NS proteins in virus life cycles has not been fully elucidated. Here, we have clearly demonstrated that HCV NS5A interacts with the core protein in coimmunoprecipitation experiments not only with coexpression of each epitope-tagged protein but also with cells expressing the viral genome; and by using immunofluorescence and subcellular fractionation analysis, we have confirmed that mutations in CL3B abolish colocalization of NS5A and the core protein, presumably around LDs. In addition, the intracellular infectivity assay and IP-RT-PCR strongly suggest that impairment of the NS5A-core protein interaction results in disruption of virus production at an early stage of virion assembly. On the basis of the present results and findings in accompanying articles, one may infer the following events: newly synthesized HCV RNAs bound to NS5A are released from the replication complex-containing membrane compartment and can be captured by the core protein via interaction with domain III of NS5A at the surface of LDs or LD-associated membranes. Consequently, the viral RNAs are encapsidated, and virion assembly proceeds in the local environment. Recruitment of newly synthesized viral RNAs to the core protein could be important for efficient nucleocapsid formation in cells, where concentrations of the viral genome and the structural proteins are typically low, and may contribute to the selection of the viral genome to be packaged. Interaction between NS5A and the core protein has been previously reported, and the NS5A region containing an interferon sensitivity determining region and the PKR-binding sequence (aa 2212 to 2330) has been mapped to that required for binding with core protein by yeast two-hybrid and in vitro pull-down assays (13). However, involvement of domain III in the NS5A-core protein interaction was not analyzed in detail, and a role for the NS5A-core protein interaction in the HCV life cycle was not examined in that study.
A growing body of evidence points to phosphorylation of NS5A as being important in controlling HCV RNA replication. Although the degree and the requirement for its hyperphosphorylation diverge between different HCV isolates, mutations that are associated with increased replicative fitness of HCV replicons frequently lead to a reduced level of NS5A hyperphosphorylation (1, 5, 36). Inhibitors of serine/threonine protein kinases that block NS5A hyperphosphorylation facilitate replication of a non-culture-adapted replicon (3, 36). One model that has been proposed suggests that NS5A hyperphosphorylation negatively regulates HCV RNA replication by disrupting the interaction between NS5A and the vesicle-associated membrane protein-associated protein subtype A, a cellular factor considered necessary for efficient RNA replication (5). However, the regulatory role of the basal phosphorylation of NS5A in the viral life cycle is poorly understood. It has been reported that the C-terminal region of NS5A (aa 2350 to 2419) is involved in basal phosphorylation (23). There are highly conserved serine residues in this region, and alanine substitutions or in-frame deletion of the serine residues has been shown to impair basal phosphorylation but not to affect RNA replication in the genotype 1b isolate (1). Consistently, a metabolic 32P labeling experiment in the present study demonstrated that NS5A mutants of the JFH-1 isolate in the region impair the basal phosphorylation. Nevertheless, Tellinghuisen et al. noted that the serine at aa 2433 of JFH-1 is involved in generating hyperphosphorylated NS5A, as shown by Western blotting (50). The basis for this difference is uncertain. To date, there is no clear evidence to determine which serine residues located in domain III are phosphoacceptor sites or whether these residues influence NS5A phosphorylation in an indirect fashion. Future study to map phosphoacceptor sites in the NS5A domain III by biochemical approaches is needed.
We found that two of the three serine residues at CL3B are responsible for regulating the interaction of NS5A with the core protein as well as for infectious virus production. To further evaluate the effect of constitutive serine phosphorylation at the cluster, we replaced the serine residues with glutamic acid, which mimics the presence of phosphoserines. The S2428/2430E mutant led to restoration of the interaction of NS5A with the core protein and virus production up to levels similar to the wild type. Somewhat unexpectedly, the triple glutamic acid substitution (CL3B/SE) exhibited only a slight restoration effect or none at all. It is considered that the degree of negative charge on the glutamic acid residue is not completely equivalent to that of phosphoserine. It is likely that the range of acidity at the local environment of the NS5A domain III that will allow interaction with the core protein is rather narrow. Induction of a conformational change in NS5A by the incorporation of phosphate may also be important for its interaction with the core protein. Tellinghuisen et al. reported that a single serine-to-alanine substitution at aa 2433 blocks the production of infectious virus and that casein kinase II likely phosphorylates the residue (50). Although this seems inconsistent with our results, these investigators also showed that deletions producing a lack of all three serine residues in the cluster inhibited virus production more severely than a single mutation. We observed that a single substitution of S2428A, S2430A, or S2433A resulted in a moderate decrease in the virus released from the transfected cells; however, more evident perturbation was obtained from double or triple substitutions (Fig. 5A and B). Tellinghuisen et al. determined the HCV production at 48 h after RNA transfection and found a marked inhibition by the single substitution S2433A. In our study, as indicated in Fig. Fig.5A,5A, the reduction caused by the S2433A mutant was approximately 90% at 48 h after transfection; however, the virus production from the mutant reached a similar level to that of the wild type at 96 h posttransfection.
Several previous studies have found that apolipoproteins B (apoB) and E (apoE), microsomal triglyceride transfer protein, and HCV p7 protein are key factors for production of the infectious HCV particles (4, 11, 16, 22, 47). Assembly and maturation of the viral particles appear to depend on the formation of very-low-density lipoprotein, a large particle containing apoB, apoE, and large amounts of neutral lipids in hepatic cells. p7 protein is primarily involved in a late step of virus particle production, and the findings support the idea that p7 acts as viroporin, which has the capacity to compromise cell membrane integrity and thus favors the release of viral progeny. How the early step in virion production regulated by the NS5A-core protein interaction links with the later step(s) involved in the very-low-density lipoprotein assembly or p7 function remains an interesting question to be addressed.
In summary, we demonstrated that the C-terminal serine cluster of NS5A (aa 2428, 2430, and 2433), which is involved in generating the basal phosphorylated form, is a determinant of NS5A interaction with the core protein and the subcellular localization of NS5A. Mutation of this cluster blocks the NS5A-core protein interaction, resulting in perturbation of association between the core protein and HCV RNA. It is thus tempting to consider that NS5A plays a key role in transporting the viral genome RNA synthesized by the replication complex to the surface of LDs or LD-associated membranes, where the core protein localizes, leading to facilitation of nucleocapsid formation. Structural analysis of the NS5A domain III-core protein complex should provide greater insight into the mode of interaction between these viral proteins. Identification of residues at the interface that are involved in important interactions will be of significant value in designing novel structure-based inhibitors to block the early step of HCV particle formation.
We are grateful to Francis V. Chisari (The Scripps Research Institute) for providing Huh-7 cells. We thank M. Matsuda, S. Yoshizaki, T. Shimoji, M. Kaga, and M. Sasaki for technical assistance and T. Mizoguchi for secretarial work.
This work was supported by Grants-in-Aid from the Ministry of Health, Labor and Welfare; by the Program for Promotion of Fundamental Studies in Health Sciences of the Organization for Drug ADR Relief, R&D Promotion and Product Review of Japan (grant ID:01-3); by the Japan Society for the Promotion of Science; and by Research on Health Sciences focusing on Drug Innovation from the Japan Health Sciences Foundation, Japan. T.M. is the recipient of a Research Resident Fellowship from the Foundation for Promotion of Cancer Research in Japan.
Published ahead of print on 4 June 2008.