|Home | About | Journals | Submit | Contact Us | Français|
During replication, hepatitis C virus (HCV) NS4B protein rearranges intracellular membranes to form foci, or the web, the putative site for HCV replication. To understand the role of the C terminal domain (CTD) in NS4B function, mutations were introduced into NS4B alone or in the context of HCV polyprotein. First, we show that the CTD is required for NS4B-induced web structure, but it is not sufficient to form the web nor is it required for NS4B membrane association. Interestingly, all the mutations introduced into the CTD impeded HCV genome replication, but only two resulted in a disruption of NS4B foci. Further, we found that NS4B interacts with NS3 and NS5A, and that mutations causing NS4B mislocalization have a similar effect on these proteins. Finally, we show that the redistribution of Rab5 to NS4B foci requires an intact CTD, suggesting that Rab5 facilitates NS4B foci formation through interaction with the CTD.
Hepatitis C virus (HCV) is a positive strand RNA virus that belongs to the Flaviviridae (Kato et al., 1990; Miller and Purcell, 1990) family of viruses including flaviviruses like Dengue virus and West Nile virus, as well as pestiviruses such as bovine viral diarrhea virus (BVDV). HCV belongs to a separate genus known as hepacivirus. Many positive strand RNA viruses replicate their genome in association with intracellular membranes. In some cases, these viruses rearrange internal membranes to form novel structures that provide the platform for all the components of the virus replication machinery. For HCV, these membrane structures, termed the membranous web, or web (Egger et al., 2002), have been observed in virus-infected cells (Rouille et al., 2006), cells expressing the subgenomic replicon (Gosert et al., 2003), precursor NS4A-B (Konan et al., 2003), or mature NS4B protein (Egger et al., 2002; Konan et al., 2003). The web contains the HCV replication complex (RC), which includes the replicase proteins (NS3, NS4A, NS4B, NS5A and NS5B), viral RNA and host factors. The web is thought to concentrate the replication machinery in specific subcellular locations to optimize HCV RNA synthesis, prevent double stranded RNA-activated interferon response in infected cells, protect viral RNA from host nucleases and proteases, or perhaps hinder host acquired immune response to HCV infection by reducing MHC-I cell surface presentation (Konan et al., 2003).
HCV NS4B is a poorly characterized hydrophobic protein, approximately 27-kDa in size (Hugle et al., 2001). The only biochemical functions associated with NS4B are its GTPase and ATPase activities (Einav et al., 2004; Thompson et al., 2009), but there is no evidence that these activities are associated with the formation of NS4B foci, which contain the HCV RC. Recently, it has been suggested that NS4B GTPase activity might play a role in NS4B-induced cellular transformation and tumor formation (Einav et al., 2008b). Further, NS4B has been reported to bind to the 3’ end of negative strand viral RNA (Einav et al., 2008a), suggesting that NS4B might tether HCV RNA onto the membranous web and facilitate positive sense RNA synthesis. Other activities associated with NS4B expression include modulation of NS5A hyperphosphorylation (Jones et al., 2009; Koch and Bartenschlager, 1999), transactivation of interleukin 8 promoter (Kadoya et al., 2005), suppression of translation (Kato et al., 2002) and modulation of the ER stress response through interaction with activating transcription factor 6β (Tong et al., 2002; Zheng et al., 2005). When co-expressed in separate constructs, NS4B protein has been found to interact with other nonstructural (NS) proteins (NS3, NS4A, NS5A and NS5B), all of which are involved in HCV RNA synthesis (Ali, Tardif, and Siddiqui, 2002; Alter, 1997; Egger et al., 2002; El-Hage and Luo, 2003; Gosert et al., 2003; Ishido, Fujita, and Hotta, 1998). These findings suggest that NS4B might provide the scaffold for the formation of the HCV RC. Similarly, BVDV NS4B has been shown to bind to NS3 and NS5A, two proteins involved in BVDV genome replication (Qu, McMullan, and Rice, 2001). Altogether, these reports indicate that NS4B plays a central role in genome replication among the Flaviviridae family of viruses. Finally, we have recently reported that NS4B interacts with Rab5, an early endosome protein with a putative role in the formation of the HCV RC (Stone et al., 2007).
The precise membrane topology of NS4B remains controversial; it is predicted to have at least four transmembrane domains (TMDs) (Lundin et al., 2003) with the last 70 amino acids (residues 192 to 261), also called the NS4B C-terminal domain (NS4B CTD), on the cytosolic side of the endoplasmic reticulum (ER) membrane. Until recently, very little was known about the role of NS4B CTD in NS4B function. Some of the features associated with the CTD include genetic interaction with NS3 protein (Paredes and Blight, 2008), involvement in NS4B oligomerization (Yu et al., 2006), interaction with viral RNA (Einav et al., 2008a), and hyperphosphorylation of NS5A protein (Jones et al., 2009). A recent report has also shown that the NS4B CTD is partially responsible for NS4B association with host membranes (Liefhebber et al., 2009). By comparison, the N-terminal domain has been reported to have NS4B GTPase activity (Einav et al., 2004; Thompson et al., 2009) and contains two alpha helices involved in NS4B host membranes association (Elazar et al., 2004; Gouttenoire et al., 2009).
To further elucidate the role of the NS4B protein in HCV replication, we have focused on the CTD. Since the CTD is cytosolic, we hypothesized that this domain plays an important role in HCV genome replication, perhaps through interaction with viral and host factors. In this report, we have used genetic and biochemical approaches to examine the CTD. We show that most of the tested residues in NS4B CTD are required for HCV genome replication. However, very few of the residues examined contribute to the formation of NS4B foci, which may represent HCV replication complexes. The significance of these findings will be discussed.
NS4B protein induces the web (Egger et al., 2002; Konan et al., 2003), which has been postulated to be the site for HCV replication. The web often appears as foci in cells expressing NS4B alone or in the context of the HCV replicon (Elazar et al., 2004; Lundin et al., 2003; Stone et al., 2007). To identify the determinants of NS4B-induced web formation, we focused first on the NS4B CTD, which has been predicted to be on the cytosolic side of the ER membrane (Lundin et al., 2006), is highly conserved among HCV isolates (Fig. 1A), and likely to be engaged in protein-protein interactions required for NS4B function. To define the role of the CTD in the formation of the web, Wt and truncated NS4B derivatives were engineered, each with a C-terminal GFP fusion to NS4B (Fig. 1B). GFP was chosen because it was shown previously not to alter NS4B membrane association or function (Elazar et al., 2004; Lundin et al., 2003). The resulting constructs were transfected into Huh7.5 cells and at 48 h post-transfection, the lysates were examined for NS4B expression using GFP antibody. As shown in Fig. 1C, Wt and mutant NS4B proteins were expressed similarly, suggesting that the deletions did not have a significant effect on NS4B expression or stability. Interestingly, a band similar in size to that of GFP protein could be observed in most of the NS4B-GFP fusion constructs, suggesting perhaps a partial internal translation initiation or degradation of NS4B protein.
We first tested the effects of NS4B CTD truncations on Wt NS4B subcellular distribution via fluorescence microscopy. GFP-fused constructs were detected both by GFP fluorescence and GFP staining. As shown in Fig. 1D (i–iii), expression of GFP alone resulted in a diffuse subcellular distribution with some nuclear fluorescence, whereas Wt NS4B-GFP showed the foci and perinuclear distribution (Fig. 1D, iv–vi) typically associated with native NS4B protein (Elazar et al., 2004; Hugle et al., 2001; Lundin et al., 2003). Interestingly, NS4B CTD truncations fell into three groups: one group displayed NS4B foci [(Δ192–226; Fig. 1D, vii–ix) and (Δ227–261; Fig. 1D, xvi–xviii) which were smaller and more numerous than the Wt. The second group showed an intermediate phenotype (Δ227–250; Fig. 1D, xiii–xv) consisting of a mostly reticular distribution with few scattered smaller foci. Finally, the last group (Δ192–261; Fig. 1D, x–xii) exhibited a reticular NS4B distribution, which was different from GFP expressed alone. These results suggest that the NS4B CTD is required for Wt NS4B subcellular distribution.
To elucidate the role of the CTD in the formation of the web, Huh7.5 cells expressing GFP, GFP-fused Wt NS4B (NS4B-GFP), or CTD-truncated NS4B [NS4B Δ (192–261)-GFP] protein, were examined via transmission electron microscopy (TEM). If NS4B CTD is required for the formation of the web, we hypothesized that a deletion of this CTD would lead to a disruption of this membrane structure. To ensure that we could reproducibly observe the web under transient conditions, transfected cells were sorted to enrich the GFP- or NS4B-GFP expressing cell population, followed by immunoblotting to confirm NS4B expression (data not shown). Since sorting results indicated that 70–90% of the transfected cells were positive for GFP expression, these cells were processed for subsequent TEM analysis. As shown in Fig. 2Aii, cells expressing NS4B-GFP typically contained a cluster of vesicles characteristic of the web. This rearranged membrane was found in approximately 40% of the NS4B-GFP-expressing cells by 48 h post-transfection. Cells expressing GFP alone did not show such a vesicle cluster (Fig. 2Ai), suggesting that the web was specifically induced by NS4B-GFP expression. Interestingly, when the cells were transfected with NS4B Δ(192–261)-GFP construct, the vesicles were formed but were no longer clustered. Instead, these vesicles were scattered throughout the cytosol (Fig. 2A iii & iv), suggesting that NS4B CTD might be required for the formation of the web. However, these results could not be explained by a change in NS4B membrane association. Indeed, when subjected to membrane floatation assay (Elazar et al., 2004), the membrane association of Wt NS4B and its CTD truncation counterpart was similar to calnexin (Fig. 2B), an integral membrane protein. The profile for these proteins was different from GFP and GAPDH, both soluble proteins (Fig. 2B). Like Wt NS4B, these data indicate that NS4B Δ(192–261)-GFP gene product was still membrane-bound.
The finding that the CTD truncations alter Wt NS4B subcellular distribution led us to determine whether the CTD alone could form NS4B foci as well as the web. As shown in Fig. 3A (i &ii), expression of the GFP-fused CTD [(192–261)-GFP or GFP-(192–261)] resulted in few foci with hollow foci resembling rings. Further, the location of GFP with respect to the CTD did not seem to have a major impact on NS4B fluorescence. These results imply that the CTD may not be sufficient to form the Wt NS4B foci. Interestingly, cells expressing the CTD alone did not appear to induce the cluster of vesicles typically observed with Wt NS4B protein. As shown in Fig. 3B (ii), Huh7.5 cells with GFP-fused CTD [(192–261)-GFP] protein displayed vesicles that were at least 1 µm in diameter, whereas the web-associated vesicles were usually less than 200 nm (compare 3Bii to 3Bi). More importantly, when NS4B (192–261)-GFP was expressed for 48 h, it was found to be cytotoxic and the intracellular vesicles were as large as 2 µm in diameter (data not shown). Taken together, these data suggest that the CTD alone may not be sufficient to induce the web structure.
Since NS4B CTD is required for the formation of the web, we predicted that some missense mutations in this domain would impede HCV genome replication. To test this hypothesis, we used PSIPRED (http://www.predictprotein.org/meta.php), a protein secondary structure prediction program, to search for putative structures in NS4B CTD. As shown in Fig. 4B, two α helices referred to as helix 1 and helix 2 were identified in the CTD. We focused our investigation on helix 2 because its deletion resulted in an intermediate NS4B subcellular distribution [compare Fig. 1D (iv–vi to xiii–xv)], implying that it could be involved in some aspect of NS4B function. The mutated CTD residues were targeted for two reasons. First, we chose conserved residues on the basis that they are essential for NS4B function(s) among all the HCV genotypes. Secondly, roughly every third amino acid (between positions 227 and 250) was subjected to a non-conservative substitution because we postulated that such mutations could reveal whether the mutated residues have functions other than their role in the formation of the web. The following substitutions were engineered (D228A, V233R, L237E, S239W, T241A, L245D and H250E) (Fig. 1A & 4B) and the mutated gene was subcloned into the subgenomic replicon with a C-terminal GFP fusion to NS5A (Fig. 4A) (Moradpour et al., 2004).
To determine the effect of NS4B CTD mutations on HCV replication, in vitro transcribed RNA was generated from Wt and mutant replicon DNA constructs and electroporated into Huh7.5 cells. At 24 h post-electroporation, complete DMEM with 0.5 mg/ml G418 was added to the cells to select for neomycin-resistant clones. The replication potential of Wt and mutant replicon RNAs was measured in colony formation units per microgram of input RNA (CFU/µg RNA). As shown in Fig. 4C, the replicon construct expressing Wt NS4B protein replicated efficiently. As expected, no G418-resistant clones were found in cells electroporated with the polymerase-deficient (Pol-) vector. Interestingly, all the replicon constructs with NS4B CTD mutations showed a dramatic decrease (at least 28-fold) in HCV genome replication. The replicons containing NS4B D228A, T241A and L245D mutations resulted in few G418-resistant clones whereas the least viable mutations were V233R, L237E and S239W. These data suggest that the conserved amino acid sequence of helix 2 of the CTD is indispensable for HCV genome replication.
Several plausible scenarios could explain why the mutations in NS4B CTD impede HCV genome replication. First, these mutations could affect NS4B synthesis, stability or polyprotein processing. Alternatively, they could affect NS4B interaction with viral or host factors involved in the formation of the HCV RC.
To test whether the CTD mutations alter NS4B synthesis or stability, Huh7.5 cells were transfected with DNA constructs expressing HCV polyprotein containing Wt or mutant NS4B. First, we determined the transient transfection and (35S)-methionine labeling conditions under which, expression of the Wt polyprotein could lead to detectable levels of both the mature replicase proteins and their precursor forms. As shown in Fig. 5A, when transiently expressed in Huh7.5 cells, most of the replicase proteins and their precursors (NS4A-B, NS4B-5A and NS5A-5B) were detected as early as 10 min after pulse labeling followed by immunoprecipitation (IP) with replicase-specific antibodies. Notably, when NS4B antibody was used, mature NS4B protein as well as its precursors (NS4A-B and NS4B-5A) could be detected. Additionally, a substantial amount of NS3 and a small amount of NS5A proteins (shown as asterisks) could also be seen. Further, when NS3 antibody was used, NS3 protein was detected as well as was a small amount of NS4A-B. These results suggest biochemical interactions between NS4B and at least NS3 protein in the context of the HCV replicase proteins. A band whose molecular weight is slightly higher than NS4B protein (shown by asterisks) was observed in αNS3, αNS4B and αNS5A lanes. The identity of this protein is not clear although it could represent a posttranslational modification of NS4B. Finally, when pooled NS3-, NS4B-, NS5A- and NS5B-specific antibodies were added to the cell lysates, most of the precursors were processed following a 60 min chase. On the basis of these results, a similar pulse-chase experiment was performed on cells expressing Wt or mutant polyprotein followed by IP with HCV-specific antibodies. As shown in Fig. 5B, no aberrant processing products were observed in cells transiently expressing the mutant polyproteins. The expression levels of most of the replicase proteins were comparable. Taken together, these results imply that mutations in the CTD may not significantly affect NS4B translation rate, stability or polyprotein processing. We also found that NS4B stability was not altered when cells expressing Wt and mutant constructs were subjected to a 2 h chase (Fig. 5C) followed by IP with NS4B-specific antibody. However, since NS4B half-life is approximately 11 h (Pietschmann et al., 2001), we cannot rule out the possibility that translation rate and polyprotein processing were not significantly altered for the CTD mutations.
To elucidate the mechanism whereby NS4B CTD mutations impede HCV genome replication, we examined the effect of these mutations on NS4B subcellular distribution. First, Wt or mutant NS4B was tested alone with a C-terminal GFP fusion. When these proteins were transiently expressed in Huh7.5 cells, their subcellular distribution fell into three distinct groups. One group (D228A, S239W, T241A, L245D and H250E) consistently showed a subcellular distribution similar to the Wt NS4B construct (Fig. 6i, ii & v–viii). The second group (V233R) had an intermediate phenotype (elongated foci), whereas the last group (L237E) displayed a mostly reticular NS4B distribution with few foci (Fig. 6iii & 6iv, respectively). Taken together, these findings suggest that two NS4B CTD mutations (V233R and L237E) may result in the mislocalization of NS4B in the cell.
To test the effect of NS4B substitutions (V233R and L237E) on other replicase proteins, these mutations were examined in the context of the HCV polyprotein. Briefly, Huh7.5 were transiently transfected with DNA constructs expressing HCV polyprotein containing Wt or mutant NS4B. At 48 h post-transfection, the cells were stained with NS4B-specific antibody whereas N5A was detected with the C-terminally fused GFP. As shown in Fig. 7A (i–iii), NS4B foci were observed and these foci merged with NS5AGFP fluorescence in the context of the Wt polyprotein. However, when the V233R mutation was expressed in the polyprotein, NS4B displayed a more reticular pattern (Fig. 7Aiv). More importantly, this change in NS4B staining resulted in a similar alteration in NS5A subcellular distribution (compare Fig. 7Av to 7Aii). Like V233R, when L237E was expressed in the polyprotein, both NS4B and NS5A displayed a reticular distribution (compare Fig 7Avii & 7Aviii). These data indicate that the two NS4B CTD mutations (V233R and L237E) caused a similar change in the intracellular distribution of both NS4B and NS5A proteins. Finally, we also tested whether such NS4B mutations had any effect on the distribution pattern of NS3, another replicase protein. As shown in Fig. 7B (iv–ix) and the table below, a change in Wt NS5A intracellular distribution resulted in a concomitant change in NS3 subcellular pattern (NS5A fluorescence is used here because NS5A co-localizes with NS4B in the context of the polyprotein). Taken together, these data indicate that NS4B, NS3 and NS5A are mislocalized in cells expressing NS4B V233R and L237E mutant proteins.
We have reported previously that Rab5 (an early endosome protein) is associated with HCV replicase proteins, and that silencing of Rab5 results in a significant decrease in NS5A foci as well as HCV replication (Stone et al., 2007). Since Rab5 is associated with NS4B foci, we hypothesized that any alteration in NS4B subcellular distribution would result in a similar change in Rab5. Indeed, in cells expressing the Wt polyprotein, Rab5 displays approximately 47% co-localization with NS5A protein [Fig. 8 (A, i–iii and table)] and those Rab5 foci appear to be as large as NS5A foci. However, when mutant polyproteins (with NS4B V233R or L237E) were expressed, Rab5 showed little to no co-localization with NS5A protein [Fig. 8 (A, iv–ix and table)]. Rab5 was also not redistributed in these cells and the foci were similar in size to Rab5 in parental cells (compare Fig. 8A, iv–ix to xiii). These results suggest that NS4B residues V233 and L237 may be involved in the recruitment of Rab5 to NS4B foci. Finally, mutations that had no visible effect on Wt NS4B subcellular pattern (e.g. NS4B S239W; FIG. 8A, x–xii) also redistributed Rab5 cellular localization. These data are consistent with the findings that HCV NS4B-induced foci are associated with cellular Rab5 protein (Stone et al., 2007).
Expression of HCV NS4B protein results in the formation of the web, which contains the HCV replication complex. Yet, very little is known about NS4B determinants of web induction. In this study, we show first that the NS4B CTD is required for the formation of NS4B foci and the web. However, we found that the CTD alone may not be sufficient to induce these structures. These results are consistent with previous report indicating that the NS4B N-terminal domain (NTD) also plays a role in the formation of NS4B foci (Elazar et al., 2004). Secondly, helix 2 in the CTD has multiple functions. Indeed, although all the seven mutations introduced into this helix impede HCV RNA synthesis (summary results in Table 2), only two of them (V233R and L237E) result in a disruption of NS4B foci. These data suggest that formation of NS4B foci may not be sufficient for HCV genome replication. Third, we show that NS4B interacts weakly with NS5A and strongly with NS3 in the context of the other replicase proteins, and that the two mutations leading to a disruption of NS4B foci have a similar effect on NS3 and NS5A. Finally, our data imply that NS4B CTD residues, V233 and L237, play a role in early endosome protein Rab5 association with NS4B foci.
We chose to investigate the NS4B CTD contribution to NS4B function for several reasons. First, NS4B membrane topology predicts that the CTD is on the cytosolic side of the ER membrane (Elazar et al., 2004; Lundin et al., 2003; Qu, McMullan, and Rice, 2001) and is possibly engaged in interactions required for HCV replication (Einav et al., 2008a). Second, in comparison to the NTD, the CTD has considerable sequence identity (ca. 85% amino acid sequence similarity) among all the different HCV isolates (Fig. 1A), suggesting that it might play a major role in NS4B function. Interestingly, like the NTD, the NS4B CTD is predicted to contain two alpha helical structures termed helix 1 and helix 2 (Fig. 4B). For the NTD, specific mutations in the two helices have resulted in the loss of NS4B foci (Elazar et al., 2004; Gouttenoire et al., 2009). Likewise, we have found that mutations in the CTD can lead to some alteration in Wt NS4B subcellular distribution. However, for the CTD, the most severe alterations were observed following truncation of the entire CTD, deletion of most of helix 2 residues, or V233R and L237E substitution mutations. These findings suggest that both the NTD and helix 2 in the CTD are needed for the formation of NS4B foci.
Since NS4B CTD is required but not sufficient for the formation of Wt NS4B foci, we reasoned that the CTD alone could not facilitate web formation. Indeed, when observed under TEM, the presence of unusually large vesicles (at least 1 µm in diameter) in cells expressing the CTD implies that this domain alone may interfere with web-like vesicles formation or may cause more of such vesicles to fuse. In contrast, the cytoxicity associated with the CTD expression could be interpreted to mean that this domain alone might interfere with normal vesicular trafficking. Note that the CTD construct used in this study, and others (Elazar et al., 2004; Targett-Adams, Boulant, and McLauchlan, 2008), contains NS4B residues 192 through 261. Recently, another CTD construct was found to display mostly small NS4B foci in Huh7 cells (Liefhebber et al., 2009). However, this subcellular distribution could be explained in part by the fact that the CTD included residues 188 through 261, and therefore four additional residues from transmembrane domain 4.
We have identified V233 and L237 as two putative CTD determinants required for the formation of NS4B foci since mutations in these residues result in a significant disruption of Wt-like NS4B foci and a similar mislocalization of the replicase proteins NS3 and NS5A. One plausible scenario is that V233 and L237 may participate in NS4B-NS4B (Yu et al., 2006) or NS4B-host protein interactions involved in the formation of these foci. Conversely, it is possible that the non-conservative mutations introduced into these residues (V233R and L237E) resulted in some structural modification in the CTD that could be deleterious for NS4B function. Indeed, a helical wheel prediction program, http://cti.itc.Virginia.EDU/~cmg/Demo/wheel/wheelApp.html, suggests that helix 2 might be an amphipathic alpha helical structure. Studies are under way to test this prediction.
The findings that NS4B may interact with NS3 in the context of other replicase proteins are supported by a report from Paredes et al. (Paredes and Blight, 2008) indicating a genetic interaction between NS4B and NS3 protein. Our study suggests that the mutated CTD helix 2 residues may not be required for this interaction (data not shown), but we cannot completely rule out interaction with the other residues in helix 2 or the remainder of the CTD. In contrast, the current report suggests that some CTD mutations may have altered NS4B interaction with host factor(s). Specifically, we have previously shown that NS4B interacts with Rab5, an early endosome protein (Stone et al., 2007), and that Rab5 silencing results in a loss of NS4B foci and a significant decrease in HCV RNA synthesis (Stone et al., 2007). Additionally, a recent report shows that Rab5 is required for cell culture JFH1 genome replication (Berger et al., 2009) further confirming the importance of Rab5 in HCV replication. Here, we demonstrate that Rab5 co-localizes with Wt and mutant NS4B proteins that display punctate fluorescence. However, the two CTD mutations (V233R and L237E) that resulted in the loss of NS4B foci no longer displayed NS4B co-localization with Rab5. Studies are under way to examine possible alterations in the interaction between the various NS4B mutant proteins and Rab5 co-factor.
In summary, we have identified the CTD as a determinant in NS4B ability to form the membranous web, a site of HCV genome replication. The cytoplasmic location of the NS4B CTD suggests plausible interactions with viral and host factors required for NS4B function in the context of HCV replication.
Cells expressing the NS5AGFP replicon (Moradpour et al., 2004) were derived from Huh7.5 cells, a highly permissive cell line for HCV replication (Blight, McKeating, and Rice, 2002), and were obtained from Dr. Charles Rice (The Rockefeller University, New York, NY). Replicon and parental cells were grown as monolayers in Dulbecco’s Modified Eagle’s Medium (DMEM) supplemented with 10% fetal bovine serum (FBS), non-essential amino acids (NEAA, Invitrogen), 100 U/ml penicillin, and 100 µg/ml streptomycin at 37°C in a 5% CO2 incubator. Replicon cells were also cultured in the presence of 0.3–0.5 mg/ml G418 (Geneticin, Invitrogen, Carlsbad, CA).
To construct plasmids carrying NS4B (WT and deletion mutations) with a C-terminal GFP fusion, NS4B was amplified from genotype 1B, strain Con 1 or J4L6S (Lohmann et al., 1999; Yanagi et al., 1998). Specifically, truncations were introduced into J4L6S NS4B sequence, whereas missense mutations were engineered into Con 1 NS4B sequence. Primers (Table 1) were designed to introduce a SalI site at the 5’end, a BamHI site at the 3’ end, and an AUG start codon immediately upstream of the NS4B coding region. The PCR products were digested with SalI and BamHI, and the purified fragments were subcloned into SalI- and BamHI-cleaved pEGFP-N2 vector (Clontech, Mountain View, CA).
For site-directed mutagenesis, the QuikChange PCR method (Stratagene, La Jolla, CA) was used. Briefly, two PCR primers, one with NS4B mutation and another with the Wt sequence, were used to amplify recombinant pCR 2.1-TOPO vector (Invitrogen, Carlsbad, CA) containing a 1.2 kb fragment (including NS4A-NS4B-NS5A sequence) flanked by NsiI at the 5’end and MluI at the 3’ end. To subclone NS4B mutations into pEGFP-N2 vector, NS4B was amplified using two PCR primers, one flanked by SalI and the other by BamHI. The PCR product was digested with SalI (5’ end) and BamHI (3’ end) and cloned into SalI- and BamHI-cleaved pEGFP-N2 to generate the recombinant vector expressing Wt or mutant NS4B alone.
The NS4B mutant replicon constructs were engineered in two steps. First, pI/5A-GFP-6 (generously provided by Charles Rice, The Rockefeller University) (Moradpour et al., 2004) was digested with SspI (5’ end) and MluI (3’ end) to remove an 868 bp fragment containing NS4B flanked by part of NS4A and NS5A sequences. Then, a 1.6 kb fragment [containing EMCV IRES and some pEGFP-N2 vector (Clontech) sequence] was digested with SspI and MluI, followed by ligation into SspI- and MluI-cleaved pI/5A-GFP-6 to create an intermediate vector. Finally, Wt or mutant NS4B DNA fragment (in pCR 2.1-TOPO vector) was cleaved with SspI and MluI. The purified DNA was ligated into the SspI- and MluI-cleaved intermediate vector to generate recombinant pI/5A-GFP-6 vector expressing Wt or mutant NS4B protein. DNA sequences from Wt and NS4B mutant replicons were confirmed by DNA sequencing.
We also engineered a recombinant pIRES vector (Clontech) containing NS5AGFP subgenomic replicon (pIR/I5AGFP) DNA sequence to express Wt and mutant HCV polyproteins under the transcriptional control of both CMV and T7 promoters. To this end, pI5AGFP-6 (Moradpour et al., 2004)] vector was digested with XbaI (5’ end) and MluI (3’ end) and the resulting 5’ UTR-neo-NS3-NS4A-NS4B-NS5A fragment was ligated into NheI- and MluI-cleaved pIRES vector. Next, pI5AGFP-6 vector was digested with MluI and EcoRV and the resulting NS5A-NS5B-3’ UTR fragment was ligated into MluI- and SmaI-cleaved pIRES vector. Finally, the recombinant pIRES vector, with NS5A-NS5B-3UTR, was digested with MluI and NotI, and the purified MluI-NS5A-NS5B-3’ UTR-NotI fragment was ligated into the MluI- and NotI-cleaved pIRES vector, containing 5’ UTR-neo-NS3-NS4A-NS4B-NS5A fragment, to generate pIR/I5AGFP vector.
Rabbit polyclonal antibody to HCV NS4B was obtained from Covance (Denver, PA) whereas mouse monoclonal antibody to NS4B was from Abcam (Cambridge, MA). Rabbit polyclonal antibodies to HCV NS3, NS5A and NS5B were generous gifts from Craig Cameron’s laboratory (Penn State University, University Park). Rabbit polyclonal antibody to GFP and Rab5 were obtained from Santa Cruz Biotechnology (Santa Cruz, CA). Calnexin and GAPDH antibodies were from Stressgen (Ann Arbor, MI) and Fitzgerald industries International, Inc. (Concord, MA), respectively. Horseradish peroxidase-conjugated secondary antibodies were obtained from Vector Laboratories (Burlingame, CA). Alexa fluor-conjugated secondary antibodies were obtained from Invitrogen (Carlsbad, CA).
For each experiment, parental cells (Huh7.5) or NS5AGP subgenomic replicon-expressing cells were trypsinized and grown overnight in 10 cm dishes or 6-well plates to obtain 70–80% confluent monolayer cells. Prior to transfection, the cells were washed with phosphate-buffered saline (PBS) and fed with 10 ml of fresh complete medium (for 10 cm dishes) or 2 ml of complete medium per well for 6-well plates. Cells were transfected according to the TransIT-LT1 protocol from Mirus(Madison, WI). The DNA mixture was added to each dish and incubated at 37°C for 24 or 48 h. With this procedure, DNA transfection efficiency was usually 70–90%.
For membrane floatation assay, 2–3 × 100 mm dishes (7 × 105 cells/dish) of parental cells (Huh7.5) were grown overnight and transfected as described above. Transfected cells were resuspended in homogenization buffer (150 mM NaCl, 50 mM Tris pH 7.4, 2 mM EDTA) containing protease inhibitors (1mM PMSF and 1 tablet of Complete Mini; Roche, Nutley, NJ). The cells were then lysed with 6–8 passages in a ball-bearing homogenizer to ensure approximately 90% lysis. Cell lysates were spun at 2500 xg/10 min at 4°C to pellet cellular debris and nuclei. A discontinuous iodixanol gradient (5%, 25% and 30%) (Elazar et al., 2004) was layered on the top of the homogenate and the samples were spun at 120,000×g for 4h 25 min at 4°C in a Ti80 Rotor. A total of 8 fractions (867 µl each) were collected from top to bottom. Each fraction was precipitated as described above, separated on 10% SDS-PAGE and processed for western blotting as described above. Typically, membrane-bound proteins were associated with fractions 1 to 4 whereas soluble proteins were prominent in fractions 5 to 8.
Huh7.5 cells were seeded at 1.5 × 105 cells/10 cm dish approximately 24 h prior to transfection. Before transfection, the cells were washed once in PBS and fed with fresh DMEM/10%FBS. The cells were co-transfected with 10 µg of DNA encoding Wt or mutant NS5A-GFP replicon in pIRES vector (pIRES-I/5A-GFP) and 5 µg of plasmid DNA encoding T7 polymerase (pCAGGS-T7 kindly provided by Biao He, The Pennsylvania State University). Parental cells, co-transfected with pIRES and T7 polymerase-encoding vector, were used as negative controls. DNA transfection with TransIT-LT1 reagent was done as described above.
At 48 h post transfection, the cells were trypsinized and washed twice with PBS. Each sample was then resuspended in 1 ml of DMEM without cysteine and methionine (Invitrogen) and incubated at 37°C with gentle rotation for 1h. After starvation, the cells were labeled with 200 µl of 500 µCi/ml Express 35S protein labeling mix (Perkin-Elmer) for 10–15 min. The cells were evenly split into aliquots, and some were lysed immediately in 1 ml of ice-cold RIPA buffer [150 mM NaCl, 50 mM Tris, pH 8.0, 1 mM EDTA, 1% NP-40, 0.1% SDS, 1 mM phenylmethylsulfonyl fluoride and complete protease inhibitor cocktail (Roche)]. Alternatively, the remaining label was removed after brief centrifugation (500 xg/1 min), replaced with complete DMEM containing 10 mM cysteine/methionine (pH 7.4), followed by a chase and lysis with 1 ml of ice-cold RIPA buffer. Cell lysates were precleared by incubation for 1 h at 4 °C with Protein A/G Plus agarose beads (Santa Cruz Biotechnology, Santa Cruz, CA). The supernatants were then incubated with HCV-specific antibodies at a dilution of 1:500 (anti-NS3, NS5A, or NS5B) or 1:250 (anti-NS4B) for 3–12 h at 4°C, mixed with protein A/G Plus agarose and incubated for 2h at 4°C. Protein A/G agarose Plus-bound complexes were collected by centrifugation at 500 xg/5 min, washed once with RIPA buffer, once with RIPA buffer/500 mM NaCl and once more with RIPA buffer. To examine the immuno-precipitates by sodium dodecyl sulfate-polyacrylamide gels (SDS-PAGE), the samples were resuspended in 30 µl of loading buffer (Laemmli, 1970), heated at 95 °C/5–10 min and centrifuged at 14,000 xg/2 min. The supernatants were separated on 10% SDS-PAGE, fixed in 20% methanol/7% acetic acid for 10 min at room temperature, and dried for 1 h at 80°C. Labeled proteins were visualized and quantitated on a PhosphorImager (Typhoon 8600, Amersham Pharmacia Biotechnology Inc./Molecular Dynamics, Piscataway, NJ).
Plasmid DNA containing subgenomic NS5AGFP replicon (pI/5A-GFP-6)(Moradpour et al., 2004) constructs were linearized with ScaI and purified using the QIAquick PCR purification kit (Qiagen, Valencia, CA). RNA was synthesized using the T7 RiboMAX Express Large Scale RNA Production Systems Kit (Promega, Madison, WI) according to the manufacturer’s instructions. The RNA was then isolated using the RNeasy miniprep kit (QIAGEN, Valencia, CA). Prior to electroporation, subconfluent Huh7.5 cells were trypsinized and resuspended in complete DMEM. The cells were then washed three times and resuspended at a concentration of 6.25 × 106 cells/ml in OptiMEM. For each electroporation, 0.4 ml of the cell suspension (2.5 × 106 cells) was mixed with 5 µg of HCV replicon RNA and electroporated in a 0.2-mm gap cuvette in a BioRad Gene Pulser (1 pulse, 0.13 kV). The cell suspension was recovered in complete DMEM for 10 min on ice prior to plating in a 10 cm dish. Resuspended cells (1 × 104, 1 × 105, 2.5 × 105 and 5 × 105) were seeded in 10 cm dishes together with cells electroporated with polymerase-deficient replicon RNA to obtain 5 × 105 cells per dish. After 24 h incubation, the medium was changed and replaced with complete medium supplemented with 0.5 mg/ml G418. The medium was changed every three to four days for three weeks. G418-resistant colonies were fixed with 4% formaldehyde/PBS and stained with 0.1% crystal violet in 70% ethanol. G418-resistant clones were counted and used to calculate the colony forming efficiency per microgram of input RNA (CFU/µg).
Huh7.5 cells were seeded onto coverslips and transfected in 10 cm dishes or 6-well plates as described above. At 48 h posttransfection, the coverslips were washed with PBS and fixed for 10 min in 4% formaldehyde/PBS. The cells were then permeabilized for 5 min at room temperature in 0.05% Triton-X 100/PBS, followed by staining with NS3 rabbit polyclonal antibody (or NS4B staining with a mouse monoclonal antibody) and Alexa fluor 594-conjugated secondary antibody. After three washes in PBS, the cells were stained with DAPI/PBS for 10 min at room temperature, followed by three more washes in PBS. The cells were mounted on glass slides in Vectashield (Vector Laboratories, Inc., Burlingame, CA) and the coverslips sealed with nail polish. The samples were then examined by fluorescence microscopy (Zeiss Axiovert 200 M) with a 63x lens. Digital images were taken with an Axiocam MRm CCD camera. An image stack was deconvolved using the iterative mode of the Axiovision software to exclude out-of-focus information. Images were saved as TIFF files, imported and processed in Adobe Photoshop.
Transfected cells or replicon cells were lysed in RIPA buffer (150 mM NaCl, 50 mM Tris, pH 8.0, 1 mM EDTA, 1% NP-40, 0.1% SDS, 1 mM phenylmethylsulfonyl fluoride, and 2 µg/ml leupeptin) and protein concentrations determined by Bio-Rad protein assay. Fifty to one hundred micrograms of total protein were typically resuspended in 4x SDS loading buffer (240 mM Tris pH 6.8, 4% SDS, 40% glycerol, 4% β-mercaptoethanol, 0.01% bromophenol blue) and boiled for 5 min at 95°C. The proteins were separated on 10% sodium dodecyl sulfate-polyacrylamide gel (SDS-PAGE) followed by transfer onto an Immobilon-P membrane (PVDF; Millipore, Billerica, MA). Antibody-bound proteins were detected by enhanced chemiluminescence detection method (ECL, Pierce, Rockford, Il).
For ultrastructural analysis, cells expressing GFP, NS4BGFP or the various mutant constructs were trypsinized, resuspended in complete DMEM and processed as follows. Briefly, the cells were resuspended in 2% glutaraldehyde/0.1M sodium cacodylate buffer and incubated on ice for 30 min. After a brief spin, fresh 2% glutaraldehde/0.1M sodium cacodylate was added to the pellet and the pellet was incubated overnight at 4°C. The cell pellet was rinsed with 0.1 M sodium cacodylate prior to postfixation with 1% osmium tetroxide/0.1 M cacodylate for 1–2 h at 4°C. After rinsing and en bloc staining in aqueous uranyl acetate, samples were dehydrated with graded ethanol concentrations, infiltrated with eponate resin and embedded overnight in eponate at 65 °C. Ultrathin sections were cut on Leica Ultracut UCT microtome (Wetzlar, Germany), collected on copper grids and stained with 1% uranyl acetate-1% lead citrate. The sections were imaged at 80kV in a JEOL JEM 1200 EXII (Peabody, MA) electron microscope.
We are grateful to Charles Rice, Craig Cameron and Biao He for reagents, Craig Cameron, Biao He and Karla Kirkegaard for suggestions and critical reading of the manuscript.
This work was supported by K22 CA129241 from the National Institute of Health.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.