|Home | About | Journals | Submit | Contact Us | Français|
Hepatitis C virus (HCV) nonstructural protein 5A (NS5A) is required for the replication of the viral genome and is involved in several host signaling pathways. To gain further insight into the functional role of NS5A in HCV replication, we screened human cDNA libraries by a yeast two-hybrid system using NS5A as the bait and identified human butyrate-induced transcript 1 (hB-ind1) as a novel NS5A-binding protein. Endogenously and exogenously expressed hB-ind1 was coimmunoprecipitated with NS5A of various genotypes through the coiled-coil domain of hB-ind1. The small interfering RNA (siRNA)-mediated knockdown of hB-ind1 in human hepatoma cell lines suppressed the replication of HCV RNA replicons and the production of infectious particles of HCV genotype 2a strain JFH1. Furthermore, these reductions were canceled by the expression of an siRNA-resistant hB-ind1 mutant. Among the NS5A-binding host proteins involved in HCV replication, hB-ind1 exhibited binding with FKBP8, and hB-ind1 interacted with Hsp90 through the FxxW motif in its N-terminal p23 homology domain. The impairment of the replication of HCV RNA replicons and of the production of infectious particles of JFH1 virus in the hB-ind1 knockdown cell lines was not reversed by the expression of an siRNA-resistant hB-ind1 mutant in which the FxxW motif was replaced by AxxA. These results suggest that hB-ind1 plays a crucial role in HCV RNA replication and the propagation of JFH1 virus through interaction with viral and host proteins.
Hepatitis C virus (HCV) infects approximately 170 million people worldwide and induces serious chronic hepatitis that results in steatosis, cirrhosis, and ultimately hepatocellular carcinoma (7, 64). More than two-thirds of the HCV-positive population in Western countries and Japan face chronic infection by genotypes 1a and 1b. The current combination therapy using pegylated alpha interferon (IFN) plus ribavirin has achieved a sustained virological response in 50% of individuals infected with HCV genotypes 1a and 1b (37, 53).
HCV belongs to the genus Hepacivirus of the family Flaviviridae and has a single-stranded, positive-sense RNA genome of approximately 9.6 kb, encoding a large polyprotein composed of approximately 3,000 amino acid residues. The polyprotein is cleaved by host and viral proteases, resulting in viral structural proteins (core, E1, and E2), a putative ion channel-forming protein (p7), and nonstructural proteins (NS2, NS3, NS4A, NS4B, NS5A, and NS5B) (40, 55). Highly structured untranslated regions are flanked at both the 5′ and 3′ ends of the open reading frame. The initiation of translation of the viral RNA is dependent on an internal ribosome entry site (IRES) localized in the 5′ untranslated region (28, 58).
The HCV RNA is suggested to replicate in a replication complex composed of the viral nonstructural proteins and several host proteins. An HCV replicon system established as a representative functional system was composed of an antibiotic gene for selection and HCV genomic RNA for autonomous replication in the intracellular compartments of human hepatoma cell line Huh7 without production of infectious particles (34). Recently, cell culture systems for production of an infectious HCV have been established based on HCV genotype 2a (32, 62, 74). Furthermore, a mouse model consisting of an immunodeficient mouse xenotransplanted with human liver fragments has been established for the study of in vivo replication of HCV (38). These in vitro and in vivo systems have enabled us to investigate the life cycle of HCV and to develop antiviral drugs for chronic hepatitis C.
NS5A is a phosphoprotein that possesses multiple functions in viral replication, IFN resistance, and pathogenesis (35). Adaptive mutations to increase RNA replication are frequently mapped to the coding region of NS5A, indicating that NS5A is critical for HCV replication (1, 71). NS5A has been shown to be associated with a range of cellular proteins involved in cellular signaling pathways, such as IFN-induced kinase PKR (14), growth factor receptor-binding protein 2 (Grb2) (56), p53 (36, 48), and the phosphoinositide-3-kinase p85 subunit (18), and with proteins involved in protein trafficking and membrane morphology, such as karyopherin b3 (8), apolipoprotein A1 (52), amphiphysin II (73), F-box and leucine-rich repeat protein 2 (FBL2) (26, 63, 70), and vesicle-associated membrane protein-associated protein subtype A (VAP-A) (59). We have previously reported that the host proteins VAP-B and FKBP8, a member of the FK506-binding protein (FKBP) family, interact with NS5A and that these interactions are required for efficient replication of HCV (16, 45), further supporting the hypothesis that NS5A is a pivotal component of the HCV replication complex.
To gain a better understanding of the functional role of NS5A in HCV replication, we screened human libraries by employing a yeast two-hybrid system and using NS5A as the bait. We thus identified human butyrate-induced transcript 1 (hB-ind1) as an NS5A-binding protein. Murine B-ind1 has been identified as a transcript induced by treatment with sodium butyrate in BALB/c BP-A31 mouse fibroblasts (10). hB-ind1 is a multiple-membrane-spanning protein, consisting of 362 amino acids, that possesses significant homology with protein tyrosine phosphatase-like, member A (PTPLA), and cochaperone p23 and is suggested to be involved in the Rac1 signaling pathway (10). In this study we examine the biological effects of the interaction of hB-ind1 with NS5A and other host proteins on the replication of HCV.
The plasmids encoding NS5A, FKBP8, VAP-A, VAP-B, and heat shock protein 90 (Hsp90) have been described previously (45). The human FBL2 gene was amplified from the total cDNA of Huh7 by PCR. A cDNA clone containing hB-ind1 cDNA was isolated from a human fetal brain library (Clontech, Palo Alto, CA) by the advanced yeast two-hybrid system Matchmaker Two-Hybrid System 3 (Clontech) using an HCV NS5A protein as bait. Each cDNA of N-terminally FLAG-tagged hB-ind1 and its mutants was generated by cloning into pEF FlagGs pGKpuro (23). pSilencer-hB-ind1, carrying a short hairpin RNA (shRNA) targeted to hB-ind1 under the control of the U6 promoter, was constructed by cloning of the oligonucleotide pair 5′-GATCCGGAAAAGCGACCACTGTTTCTCAAGAGAAAACAGTGGTCGCTTTTCCTTTTTTGGAAA-3′-5′-AGCTTTTCCAAAAAAGGAAAAGCGACCACTGTTTTCTCTTGAGAAACAGTGGTCGCTTTTCCG-3′ between the BamHI and HindIII sites of pSilencer 2.1-U6 hygro (Ambion, Austin, TX). A plasmid encoding a mutant hB-ind1 resistant to shRNA was prepared by introduction of five silent mutations (nucleotides were changed from A to G, G to A, A to C, A to T, and C to T at positions 291, 294, 297, 300, and 301, respectively) into hB-ind1 cDNA by the method of splicing by overlap extension (19). The pSilencer negative-control plasmid (Ambion) has no homology to any human gene. The pFK-I389 neo/NS3-3′/NK5.1 plasmid (46) was kindly provided by R. Bartenschlager, and the neomycin-resistant gene was replaced with a firefly luciferase gene. The resulting plasmid was designated pFK-I389 FL/NS3-3′/NK5.1. The plasmids used in this study were confirmed by sequencing with ABI Prism genetic analyzer 3130 (Applied Biosystems, Tokyo, Japan).
All cell lines were cultured at 37°C under a humidified atmosphere with 5% CO2. Human embryo kidney 293T cells were maintained in Dulbecco's modified Eagle's medium (DMEM) (Sigma, St. Louis, MO) supplemented with 100 U/ml penicillin, 100 μg/ml streptomycin, and 10% fetal calf serum (FCS). The human hepatoma cell line Huh7.5.1 was kindly provided by F. Chisari (74). The Huh7 and Huh7.5.1 cell lines were maintained in DMEM containing nonessential amino acids (NEAA), 100 U/ml penicillin, 100 μg/ml streptomycin, and 10% FCS. The Huh9-13 cell line, an Huh7-derived cell line harboring a subgenomic HCV replicon (34), was maintained in DMEM containing 10% FCS, NEAA, and 1 mg/ml G418 (Nacalai Tesque, Kyoto, Japan). Huh7.5.1 cells were transfected with pSilencer-hB-ind1 or an empty plasmid, and drug-resistant clones were selected by treatment with hygromycin (Wako, Tokyo, Japan) at a final concentration of 10 μg/ml. Plasmids encoding a full-length or truncated (amino acid residues 101 to 277) version of hB-ind1 were transfected into Huh7.5.1 cells, and the cells surviving after selection with 0.1 μg/ml of puromycin for 1 week were used for virus infection. The viral RNA of JFH1 was introduced into Huh7.5.1 cells according to the method of Wakita et al. (62). The supernatant was collected at 7 days posttransfection and used as HCV particles that are infectious in cell culture (HCVcc).
A rabbit anti-hB-ind1 antibody was prepared by immunization with synthetic peptides corresponding to amino acid residues 106 to 117 of hB-ind1. A mouse monoclonal antibody to influenza virus hemagglutinin (HA) was purchased from Covance (Richmond, CA). The mouse anti-FLAG M2 antibody that was conjugated with a horseradish peroxidase and a mouse anti-β-actin monoclonal antibody were purchased from Sigma. The mouse monoclonal antibody to HCV NS5A was obtained from Austral Biologicals (San Ramon, CA).
A human fetal brain library prepared with pAct2 was purchased from Clontech and was screened by the yeast two-hybrid system Matchmaker GAL4 Two-Hybrid System 3 (Clontech) according to the manufacturer's protocol. The NS5A cDNA fragment encoding amino acid residues 1973 to 2419 of HCV strain Con1 was amplified by PCR and cloned into pGBKT7 (Clontech); the resulting plasmid was designated pGBKT7 HCV NS5A. The yeast Saccharomyces cerevisiae strain AH109, which secretes α-galactosidase under the control of the MEL1 region, was transformed with pGBKT7 HCV NS5A and grown on a medium lacking tryptophan. The clone including the bait plasmid was transformed with the library plasmids. The transformed yeast cells were grown on 2% agar plates of a dropout medium lacking tryptophan, leucine, histidine, and adenine. The resulting colonies grown on the dropout plate were inoculated again on a new dropout plate containing 20 μg/ml X-α-Gal (5-bromo-4-chloro-3-indolyl-α-O-galactopyranoside) and incubated at 30°C for 7 days. The total DNA was prepared from all blue colonies and then introduced into Escherichia coli strain JM109. The prey plasmids were recovered from the clones grown on LB agar plates containing 10 μg/ml ampicillin. One positive clone was isolated from among 2 million colonies of the human fetal brain library, and the nucleotide sequence of this clone includes the complete cDNA of hB-ind1 in its frame.
Transfection and immunoprecipitation analyses were carried out as described previously (16, 45). Immunoprecipitates boiled in loading buffer were subjected to 12.5% sodium dodecyl sulfate-polyacrylamide gel electrophoresis. The proteins were transferred to polyvinylidene difluoride membranes (Millipore, Bedford, MA) and were reacted with the appropriate antibodies. The immune complexes were visualized with SuperSignal West Femto substrate (Pierce, Rockford, IL) and detected by an LAS-3000 image analyzer system (Fujifilm, Tokyo, Japan).
The short interfering RNAs (siRNAs) Target-4 (5′-GCUGAGUGACGUACAGAAC-3′) and Target-6 (5′-GGAAAAGCGACCACUGUUU-3′) were obtained for knockdown of endogenous hB-ind1 (Ambion, Austin, TX). The negative control, siCONTROL Non-Targeting siRNA 2, which exhibits no downregulation of any human genes, was purchased from Dharmacon (Buckinghamshire, United Kingdom). Huh9-13 cells harboring a subgenomic HCV replicon grown on 6-well plates were transfected with 20 nM siRNA by using siFECTOR (B-Bridge International, Sunnyvale, CA) according to the manufacturer's protocol. The transfected cells were incubated in DMEM supplemented with 10% FCS and were then harvested at 96 h posttransfection.
The HCV RNA level was estimated by the method described previously (16, 45). Total RNA was prepared from cells by using the RNeasy minikit (Qiagen, Tokyo, Japan). First-strand cDNA was synthesized using an RNA LA PCR kit (Takara Bio Inc., Shiga, Japan) and random primers. Each cDNA was estimated by Platinum Sybr green qPCR SuperMix UDG (Invitrogen, Carlsbad, CA.) according to the manufacturer's protocol. Fluorescent signals were analyzed by an ABI Prism 7000 system (Applied Biosystems). The HCV IRES, glyceraldehyde-3-phosphate dehydrogenase (GAPDH), and hB-ind1 genes were amplified using primer pairs 5′-GAGTGTCGTGCAGCCTCCA-3′-5′-CACTCGCAAGCACCCTATCA-3′, 5′-GAAGGTGAAGGTCGGAGTC-3′-5′-GAAGGTGAAGGTCGGAGTC-3′, and 5′-CACCTGGAGTTCTTAGACCTTGTG-3′-5′-CAGTCGGAGTTTATTTAGGCGCTC-3′, respectively. The values for HCV genomic RNA and hB-ind1 mRNA were normalized to that for GAPDH mRNA. Each PCR product was detected as a single band of the correct size by agarose gel electrophoresis (data not shown).
Plasmids pFK-I389 neo/NS3-3′/NK5.1 and pFK-I389 FL/NS3-3′/NK5.1 were linearized at the ScaI site and then transcribed in vitro using the MEGAscript T7 kit (Ambion) according to the manufacturer's protocol. To generate capped mRNA encoding Renilla luciferase, pRL-CMV was cleaved with BamHI and then transcribed using the mMESSAGE mMACHINE kit (Ambion) according to the manufacturer's protocol. These in vitro-transcribed RNAs were introduced into Huh7.5.1 cells at 4 million cells/0.4 ml by electroporation at 270 V and 960 μF using Gene Pulser (Bio-Rad, Hercules, CA).
The colony formation assay has been described previously (45). Briefly, in vitro-transcribed RNA was electroporated into Huh7 cells and plated in DMEM containing 10% FCS and NEAA. The medium was replaced with fresh DMEM containing 10% FCS, NEAA, and 1 mg/ml G418 at 24 h posttransfection. The remaining colonies were fixed with 4% paraformaldehyde and stained with crystal violet at 4 weeks after electroporation.
Transfected cells were seeded in a 12-well plate and then lysed in 200 μl of passive lysis buffer (Promega, Madison, WI) at 24 h posttransfection. Luciferase activity was measured in 20-μl aliquots of cell lysates using the Dual-Luciferase reporter assay system (Promega). Firefly luciferase activity was standardized to that of Renilla luciferase, and the results are expressed as the increases in relative luciferase units (RLU).
Results are expressed as means ± standard deviations. The significance of differences between the means was determined by Student's t test.
NS5A derived from the genotype 1b strain Con1 was used as bait to screen the human fetal brain cDNA library by a yeast two-hybrid system, and one clone including a gene encoding the open reading frame of the hB-ind1 gene was isolated. To examine whether hB-ind1 could interact with NS5A in mammalian cells, HA-tagged NS5A (HA-NS5A) was coexpressed with FLAG-tagged hB-ind1 (FLAG-hB-ind1) in 293T cells and immunoprecipitated with an antibody to the HA or the FLAG tag. FLAG-hB-ind1 and HA-NS5A were coimmunoprecipitated by either antibody (Fig. (Fig.1A).1A). To determine the interaction of various genotypes of NS5A with hB-ind1, HA-NS5A of the genotype 1a strain H77C, the genotype 1b strain J1, or the genotype 2a strain JFH1 was coexpressed with FLAG-hB-ind1 and immunoprecipitated with the anti-FLAG antibody. An empty plasmid was used as a negative control. FLAG-hB-ind1 was immunoprecipitated with the anti-FLAG antibody at similar levels in cells coexpressing FLAG-hB-ind1 and HA-NS5A of all genotypes. HA-NS5A of various genotypes was coprecipitated with FLAG-hB-ind1 by the anti-FLAG antibody, whereas the anti-FLAG antibody did not precipitate any HA-NS5A of the various genotypes used in this study (Fig. (Fig.1B).1B). To further confirm the interaction between hB-ind1 and HCV NS5A in the functional setting, lysates of Huh9-13 cells harboring subgenomic HCV replicon RNA were subjected to immunoprecipitation analysis with a rabbit polyclonal antibody raised against hB-ind1. NS5A was coimmunoprecipitated with endogenous hB-ind1 in the lysates of replicon cells (Fig. (Fig.1C).1C). These results indicate that hB-ind1 interacts with NS5A of various HCV genotypes in mammalian cells.
hB-ind1 is composed of 362 amino acid residues and has domains homologous with p23 and PTPLA in the regions from Pro8 to Asp112 and from Gln196 to Leu346, respectively (Fig. (Fig.2A).2A). To determine the region responsible for the interaction with NS5A, various deletion mutants of FLAG-hB-ind1 were constructed (Fig. (Fig.2B).2B). Each of the mutants was coexpressed with Con1 HA-NS5A in 293T cells and immunoprecipitated with an anti-HA antibody. An empty plasmid was used as a negative control in the immunoprecipitation analyses. HA-NS5A was coimmunoprecipitated with full-length hB-ind1 and with mutants possessing amino acid residues 114 to 134, corresponding to the coiled-coil domain, which generally participates in protein-protein interactions (Fig. 2B and C), whereas HA-NS5A was not coimmunoprecipitated with hB-ind1 mutants lacking the coiled-coil domain. The anti-HA antibody did not coprecipitate FLAG-hB-ind1 or its mutants. These results indicate that hB-ind1 interacts with HCV NS5A through the coiled-coil domain.
To investigate the role(s) of endogenous hB-ind1 in the replication of HCV RNA, an siRNA targeted to hB-ind1 or a control siRNA was transfected into Huh9-13 cells harboring subgenomic HCV replicon RNA. Total RNA was extracted from the transfected cells, and levels of hB-ind1 mRNA and HCV RNA were determined by real-time PCR. At 72 h posttransfection, hB-ind1 mRNA and HCV subgenomic RNA levels in cells transfected with each of the hB-ind1 siRNAs were reduced more than 60% from the levels in cells treated with the control siRNA (Fig. (Fig.3A).3A). The levels of expression of hB-ind1 and the HCV NS5A protein were decreased in HCV replicon cells transfected with the hB-ind1 siRNA but not in those transfected with the control siRNA (Fig. (Fig.3B3B).
To examine the effects of the knockdown of hB-ind1 on the replication of HCV RNA and the propagation of HCVcc, we established Huh7.5.1 cell lines stably expressing an shRNA targeted to hB-ind1. Dozens of colonies were obtained from cells transfected with a plasmid encoding the cDNA of the shRNA to hB-ind1 after selection with hygromycin. Although the levels of mRNA and expression of endogenous hB-ind1 were not changed in cells bearing a nonspecific shRNA, they were reduced in the clones bearing shRNAs targeted to hB-ind1, except for clone 1 (Fig. 3C and D). There was no significant difference in growth among the cell lines (Fig. (Fig.3E3E).
The replicon RNA transcribed from pFK-I389 neo/NS3-3′/NK5.1 was transfected into the hB-ind1 knockdown cell lines Huh-si2 and Huh-si5, which were cultured for 4 weeks in the presence of G418. The numbers of colonies in the knockdown cell lines were less than one-fourth of those in the control cell line (Huh-c) (Fig. (Fig.4A).4A). A FLAG-tagged hB-ind1 wobble mutant (FLAG-rB-ind1), which is resistant to the shRNA targeted to hB-ind1 due to the introduction of silent mutations, was capable of expressing an siRNA-resistant hB-ind1 upon introduction into cells at a level similar to that of the endogenous hB-ind1 (eB-ind1) detected in the control cell line (Fig. (Fig.4B).4B). The reduction of colony formation by the knockdown of eB-ind1 in the hB-ind1 knockdown cell lines Huh-si2 and Huh-si5 was canceled by the expression of FLAG-rB-ind1 (Fig. (Fig.4A).4A). To further examine the involvement of hB-ind1 in the replication of HCV, a chimeric HCV RNA encoding a firefly luciferase gene under the control of HCV IRES (Fig. (Fig.4C)4C) was transfected into the knockdown cell lines. Knockdown of hB-ind1 reduced the RLU in Huh-si2 and Huh-si5 cells by 40% and 70%, respectively, and this reduction was also canceled by the expression of FLAG-rB-ind1. To further examine the effect of hB-ind1 knockdown on the production of HCV infectious particles, HCVcc were inoculated into the hB-ind1 knockdown cell lines. Both virus titers, determined by focus-forming units at 72 h postinfection in culture supernatants, and HCV RNA levels in Huh-si2 and Huh-si5 cells were significantly reduced, and these reductions were canceled by the expression of FLAG-rB-ind1 (Fig. (Fig.4D).4D). These results suggest that hB-ind1 is involved in the replication of HCV RNA and the propagation of HCVcc.
To examine the involvement of hB-ind1 in the replication of HCV in greater detail, deletion mutants of hB-ind1 retaining or lacking the binding region to NS5A were expressed in Huh9-13 cells harboring subgenomic HCV replicon RNA (Fig. (Fig.5A).5A). Although the hB-ind1 mutant possessing the NS5A binding region (101-277) and full-length hB-ind1 were detected at similar levels in replicon cells transfected with the expression plasmids (Fig. (Fig.5B),5B), HCV RNA replication was reduced only in cells expressing the mutant retaining the binding region to NS5A, not in those expressing full-length hB-ind1 or the mutant lacking the binding region to NS5A (101-277 Δ114-134) (Fig. (Fig.5C).5C). However, no significant difference in NS5A expression was observed in Huh9-13 cells transfected with the expression plasmids (Fig. (Fig.5B).5B). Production of the infectious HCV particles was also reduced in the culture supernatants of Huh7.5.1 cells expressing the hB-ind1 mutant retaining the binding region to NS5A (101-277) but not in those expressing full-length hB-ind1 or the hB-ind1 101-277 Δ114-134 mutant (Fig. (Fig.5D).5D). These dominant-negative effects of the hB-ind1 mutant retaining the binding region to NS5A on the replication of HCV RNA in Huh9-13 cells and on the production of infectious particles in Huh7.5.1 cells further support the notion that hB-ind1 regulates the replication of HCV RNA and the propagation of HCVcc.
Previous reports have suggested that HCV NS5A interacts with several host proteins such as FBL2 (63), VAP-A (59), VAP-B (16), and FKBP8 (45) and that these interactions participate in the replication of HCV. To determine the interplay of the NS5A-binding proteins, FLAG-tagged hB-ind1 was coexpressed with HA-tagged FBL2, VAP-A, VAP-B, or FKBP8 in 293T cells and immunoprecipitated with an anti-FLAG antibody, and FKBP8 was shown to specifically interact with hB-ind1 (Fig. (Fig.6A).6A). We have previously shown that FKBP8 is capable of binding to both NS5A and Hsp90 through the tetratricopeptide repeat (TPR) domain and that the recruitment of Hsp90 to the replication complex plays a crucial role in the replication of HCV (45). Hsp90 is a molecular chaperone and requires various cochaperone proteins such as p23 for efficient chaperone activity. hB-ind1 shows homology to p23 (Fig. (Fig.2A),2A), and the FxxW motif, essential for the binding to Hsp90, is conserved in residues Phe107xxTrp110 of hB-ind1 (11, 27, 68). To determine whether hB-ind1 interacts with Hsp90 through the FxxW motif as reported for p23, FLAG-tagged hB-ind1 or an hB-ind1 mutant in which Phe107 and Trp110 had been replaced with Ala (FLAG-hB-ind1AxxA) was coexpressed with HA-tagged Hsp90 in 293T cells and immunoprecipitated with an anti-FLAG antibody. Hsp90 was coimmunoprecipitated with wild-type hB-ind1 but not with the mutant hB-ind1, indicating that hB-ind1 interacts with Hsp90 through the FxxW motif (Fig. (Fig.6B6B).
Previously, we showed that the amino acid residues of the carboxylate clump position in the TPR domain of FKBP8 attach to the C-terminal MEEVD motif of Hsp90 (45). To examine the interaction of hB-ind1 with Hsp90 in the absence of association with FKBP8, FLAG-tagged hB-ind1 was first coexpressed with HA-tagged Hsp90 or mutant Hsp90 lacking the MEEVD motif in 293T cells and then immunoprecipitated with an anti-FLAG antibody. Similar levels of hB-ind1 were coprecipitated with Hsp90 irrespective of the deletion of the MEEVD motif of Hsp90 (Fig. (Fig.6C),6C), suggesting that hB-ind1 alone is capable of binding to Hsp90 through the FxxW motif irrespective of the association of FKBP8. To further clarify the interplay among hB-ind1, FKBP8, and Hsp90, FLAG-tagged hB-ind1 was coexpressed with HA-tagged Hsp90 and/or FKBP8 and then immunoprecipitated with an anti-FLAG antibody. Coprecipitation of Hsp90 with hB-ind1 was increased by additional expression of FKBP8 (Fig. (Fig.6D).6D). These results suggest that hB-ind1 interacts with Hsp90 through the FxxW motif and that FKBP8 also participates in the complex formation to enhance the interaction.
Next, to examine the role of the interaction of hB-ind1 with Hsp90 in the replication of HCV RNA, the replicon RNA transcribed from pFK-I389 neo/NS3-3′/NK5.1 was transfected into hB-ind1 knockdown Huh-si5 cells expressing siRNA-resistant FLAG-rB-ind1 or FLAG-rB-ind1AxxA, in which the Hsp90 binding motif FxxW was changed to AxxA. The colony formation in Huh-si5 cells transfected with an empty plasmid was 10% of that in Huh-c cells. The expression of FLAG-rB-ind in Huh-si5 cells recovered the colony formation in Huh-si5 cells to 98% of that in Huh-c cells, although that of FLAG-rB-ind1 AxxA in Huh-si5 cells exhibited only 40% recovery (Fig. (Fig.7A).7A). To further examine the role of the interaction between hB-ind1 and Hsp90 in the production of HCVcc, Huh-si5 cells expressing either FLAG-rB-ind1 or FLAG-rB-ind1AxxA were infected with HCVcc, and the virus titer in the culture supernatants and the intracellular HCV RNA level at 72 h postinfection were determined. Virus production was reduced in the culture supernatants, and viral RNA replication in the hB-ind1 knockdown cells was restored by the expression of FLAG-rB-ind1 but not by that of FLAG-rB-ind1AxxA, as seen in colony formation by the replicon RNA (Fig. (Fig.7B).7B). Collectively, these results suggest that the interaction of hB-ind1 with Hsp90 through the FxxW motif is required for genomic RNA replication and particle production of HCV.
In this study we have shown that hB-ind1 participates in HCV RNA replication and particle production through interaction with NS5A, FKBP8, and Hsp90. hB-ind1 was initially identified as a downstream transducer of Rac1, a member of the small GTP-binding proteins, in mouse fibroblasts treated with sodium butyrate, a multifunctional agent known to inhibit cell proliferation and to induce differentiation by modulating transcription (6, 10). Rac1 possesses diverse biological functions, including cytoskeletal dynamics, membrane ruffling, cell cycle progression, gene transcription, and cell survival (4, 31, 49). Previous studies have suggested that hB-ind1 mediates Rac1 and Jun N-terminal protein kinase-NF-κB signaling and is involved in the regulation of gene expression (6, 10). Inhibition of Rac1 function leads to disruption of cytoskeleton dynamics, resulting in impairment of cell growth (17, 69).
Inhibition of cell growth downregulates HCV RNA replication in the replicon cell line (41, 51), and cell cycle regulation affects HCV IRES-mediated translation (20, 61). Furthermore, cytoskeletal regulation is required for HCV RNA synthesis (3). However, knockdown of hB-ind1 and expression of the deletion mutants exhibited neither morphological change nor suppression of cell growth, suggesting that the suppression of HCV replication by dysfunction of hB-ind1 is not due to cell growth arrest or cytoskeletal disruption. Murine B-ind1 has been reported to be expressed in all mouse tissues examined, with abundant expression detected in the testis, kidney, brain, and liver (10). Significant levels of endogenous hB-ind1 expression have been detected in the human hepatic cell lines Huh7, HepG2, Hep3B, and FLC4 and in the nonhepatic human cell lines HeLa, 293T, and THP-1 (data not shown); therefore, the tissue specificity of HCV replication could not be explained by the expression of hB-ind1.
Combination therapy with IFN and cyclosporine A has been shown to be effective for patients infected with a high viral load of HCV genotype 1b (24), and cyclosporine A has been shown to suppress HCV RNA replication in vitro through deactivation of the interaction between NS5B and cyclophilin B (66). Cyclophilin and FKBP are classified as immunophilins capable of binding to immunosuppressants cyclosporine A and FK506, respectively (33). The immunophilins do not share a homologous domain with each other, based on their amino acid sequences, substrate specificities, and inhibitor sensitivities. We have recently reported that NS5A binds specifically to FKBP8 but not to other homologous immunophilins such as FKBP52 and cyclophilin D. FKBP8 forms both a homomultimer and a heteromultimer with the chaperone protein Hsp90. Mutation analyses of FKBP8 and Hsp90 suggest that FKBP8 acts as an intermediate between NS5A and Hsp90 via the different position of the TPR domain in FKBP8 and regulates HCV genome replication (45).
The molecular chaperone Hsp90 is one of the most abundant proteins in unstressed cells and generally requires various cochaperone proteins in multiple steps to promote the folding, functional maturation, and stability of its client proteins. Newly synthesized unfolded client proteins are delivered to the Hsp70 complex via Hsp40. In most cases, Hsp70 is able to process the client proteins on its own. Certain substrates require Hsp90 for proper folding or activation. In this case, the scaffold protein Hop connects elements of the Hsp70 and Hsp90 machineries to form an intermediate complex (2, 12, 13, 47). In the late stage, the Hsp70 component dissociates, and at the same time, p23 and immunophilins enter the complex (44, 54) and the client proteins are refolded by Hsp90 chaperone activity to achieve the mature form. After that, p23 enhances the dissociation of the mature client protein from the final complex, and the released Hsp90 enters in the next chaperone cycle (72). It has been reported that Hsp90 cochaperone frequencies differ among client proteins (50). FKBP8 interacts with the C-terminal MEEVD motif of Hsp90 through the carboxylate clump position in the TPR domain of FKBP8 (45).
The C-terminal region of hB-ind1 shares homology with PTPLA (60). Protein tyrosine phosphatases are generally involved in the signaling pathways regulating metabolism, cell growth, differentiation, and cytoskeletal dynamics through the conserved HC(x)5R motif (57). NS5A also interacts with signal transducer and activator of transcription 1 (STAT1) and impairs IFN signaling through the suppression of STAT1 phosphorylation (30). In addition, intracellular uptake of apoptotic cells expressing NS5A by dendritic cells leads to an increase in the secretion of CXCL-8 and impairment of IFN-induced tyrosine phosphorylation of STAT1 and STAT2 (67). Although hB-ind1 lacks the conserved active motif, the interaction of NS5A with the coiled-coil domain in the central region of hB-ind1 may have an effect on the phosphorylation of host proteins involved in the replication of HCV.
Hsp90 has been shown to be involved in the enzymatic activity and intracellular localization of several viral polymerases, including those of influenza virus (39, 42), herpes simplex virus type 1 (5), and Flock house virus (25). Knockdown and treatment with an Hsp90 inhibitor have revealed that Hsp90 activity is important for the rapid growth of negative-strand RNA viruses (9). Furthermore, Hsp90 has been shown to be required for the activity of hepatitis B virus reverse transcriptase (21, 22). Although the precise mechanisms by which Hsp90 and FKBP8 cooperate with NS5A to improve the in vivo replication of HCV have not been clarified yet, treatment with Hsp90 inhibitors in combination with IFN reduced HCV replication in mice xenotransplanted with human liver fragments (43).
In this study, hB-ind1 was shown to interact with Hsp90 through the FxxW motif in the N-terminal p23 homology domain, and the interaction of hB-ind1 with Hsp90 was shown to be further intensified by the expression of FKBP8, suggesting that FKBP8 and hB-ind1 cooperatively recruit Hsp90 to the HCV replication complex. Furthermore, hB-ind1 was shown to be involved in HCV genomic RNA replication and particle production through the interaction with NS5A and Hsp90. These results suggest that hB-ind1 may be involved in the Hsp90 chaperone pathway in a function similar to that of p23 in cooperation with immunophilins such as FKBP8 and that it plays a crucial role in HCV replication in terms of the correct folding of the replication complex required for efficient enzymatic activity. In addition, cyclophilin B may also participate in the translocation of NS5B, as seen in the polymerase subunits of influenza virus, to facilitate binding to the viral RNA. In contrast to cyclosporine A, FK506 per se exhibits no inhibition of RNA replication in HCV replicon cells (65). FKBP8 is a member of the FKBP family but lacks several amino acid residues required for peptidyl-prolyl cis-trans isomerase and FK506 binding activities (29). Therefore, nonimmunosuppressive FK506 derivatives that are capable of binding to FKBP8 may exhibit anti-HCV activity. Recently, geldanamycin, an inhibitor of Hsp90, was shown to drastically impair the replication of poliovirus without any escape mutant emerging (15). Therefore, elucidation of host proteins, including immunophilins, cochaperones, and chaperones, participating in the HCV replication complex may lead to the development of new therapeutics for chronic hepatitis C with a broad spectrum and a low possibility of emergence of breakthrough viruses against antiviral drugs.
In conclusion, in this study we demonstrated that hB-ind1 is involved in HCV replication through interactions with NS5A, FKBP8, and Hsp90. Further clarification of the relationship between viral and host proteins is needed in order to understand the precise mechanism of HCV replication.
We thank H. Murase for secretarial work. We also thank T. Wakita, F. Chisari, and R. Bartenschlager for providing the infectious clones of JFH1, Huh7.5.1, and replicon cell lines, respectively.
This work was supported in part by grants-in-aid from the Ministry of Health, Labor, and Welfare; the Ministry of Education, Culture, Sports, Science, and Technology; the 21st Century Center of Excellence Program; and the Foundation for Biomedical Research and Innovation.
Published ahead of print on 26 December 2007.