Understanding the molecular mechanisms by which IFN-based therapies decreases HCV viral load, reduces the number of viral quasispecies, improves liver function, and reduces liver fibrosis in 15–30% of patients, is a priority in HCV research. Consequently, both viral and host factors have been implicated during the effective clinical response or resistance phenomenon of patients to IFN treatment . Different in vitro model systems have been developed to study the role of HCV polyprotein on host cell responses [12–21]. The implication of IFN-induced genes and their action in the antiviral response of the host to HCV expression is not yet fully understood.
To further characterize the antiviral response of the host during expression of HCV polyprotein, we developed a novel virus-cell system based on a poxvirus vector, that inducibly expresses the structural and nonstructural (except part of NS5B) proteins of HCV ORF from genotype 1b. The generated recombinant VT7-HCV7.9 virus contains the HCV DNA coding region inserted within the VV HA locus, under the transcriptional control of a T7 promoter, and expresses the T7 RNA polymerase upon induction with IPTG (see Figure 1A). Current systems relying on viral delivery of T7 RNA polymerase are restricted by the efficiency with which HCV cDNAs can be transfected into cells, which in the case of hepatocyte and hepatocyte-derived cell lines, is often low [16–18]. The poxvirus-based system described here permits both the regulated production of the HCV transcripts into cells and the efficient delivery of the HCV genome into a wide variety of primary and continuous cell lines.
In this study, we demonstrate that upon induction with IPTG, HCV proteins are efficiently produced in VT7-HCV7.9 infected cells of various origins. This observation indicates that the DNA fragment of HCV ORF included in the VV genome, is efficiently transcribed and translated into a viral polyprotein precursor that is correctly processed into mature structural and nonstructural HCV proteins, as confirmed with specific antibodies to individual HCV proteins. Significantly, inducible expression of HCV polyprotein in VT7-HCV7.9 infected cells caused a considerable reduction in the production of infectious VV, as well as striking inhibition in total protein synthesis, both viral and cellular. The translational block was observed by 6 h.p.i when all of the HCV proteins were produced. The inhibition of protein synthesis by HCV was highly specific and could not be solely attributed to the induction of HCV RNA transcript since cells infected with VT7-VP3 that expressed the IBDV ORF VP3 mRNA, did not trigger translational inhibition. Furthermore, the HCV ORF included in the VT7-HCV7.9 recombinant virus lacks the 5' UTR, bearing the HCV IRES, and the 3' UTR, both implicated in HCV replication and liver injury . The inhibition of protein synthesis that we have observed in induced VT7-HCV7.9 infected HeLa and HepG2 cells was associated with a significant increase in the phospho-eIF-2α-S51 levels, suggesting that HCV expression might control the cellular translation through eIF-2α-S51 phosphorylation. This translational control was confirmed with a dominant negative non-phosphorylated (NP) mutant Ser51-Ala (eIF-2α-S51A). Expression of the eIF-2α-S51A mutant in cells co-infected with VV-eIF-2α-NP and VT7-HCV7.9 in the presence of IPTG, rescued the translational block induced by HCV (Figure 6). Moreover, we showed that phosphorylation of eIF-2α-S51 was carried out by the cellular kinase PKR, as revealed in knockout PKR-/- cells (Figure 9). The role of PKR and eIF-2α-S51 phosphorylation in HCV infection has been widely studied due to the relevance of this kinase in the cellular antiviral response. As has been previously reported [23, 39–41], PKR mediated phosphorylation of eIF-2α-S51 results in inhibition of translation and a blockade of viral protein synthesis, which in turn, inhibits virus replication. For this reason, viruses employ a variety of strategies to inhibit PKR activation and function. Several groups have described the role of certain HCV proteins in cellular translation. HCV NS4A and NS4B proteins mediate translational inhibition and, perhaps, increased degradation of certain cellular proteins [42, 43]. In contrast, NS5A and E2 proteins are reported to enhance translation by inhibiting PKR functions [44, 45]. Therefore, it seems that during the course of HCV infection, there is a balance between inhibition and enhancement of host cell translation depending on the degree of activation/inhibition of the PKR pathway. Most of these studies have relied on systems that express HCV proteins individually. Nontheless, since all HCV proteins are potentially produced in vivo during virus infection of hepatocytes, it is important to use a full-length genome rather than individual HCV proteins to study the molecular mechanisms involved in virus-host cell interactions and in HCV pathogenesis. In our viral delivery system, the overall expression of structural and nonstructural HCV proteins by recombinant VT7-HCV7.9 virus did not reverse the action of PKR, since host cell translation was inhibited through phosphorylation of eIF-2α-S51 by the kinase. An incapability to prevent PKR activation by HCV polyprotein expression was reported by François and co-workers when they analysed the response to IFN of the human cell line UHCV-11 engineered to inducibly express the entire HCV genotype 1a polyprotein . Although we could not exclude the possibility that a certain level of inhibition of PKR by NS5A or E2 occurs at a much localized level, the resistance to IFN exhibited by some HCV genotypes as a result of viral protein expression, cannot be explained solely by inhibition of the negative control of PKR translation. It is possible that during the course of HCV infection, NS5A plays a role in inhibiting PKR locally at the site of HCV protein synthesis. NS5A may, however, participate in the blockade of IFN's antiviral action through another mechanism, such as the reported interaction with the Ras-associated Grb-2 protein . These results confirm the necessity to re-evaluate all types of interactions between any particular HCV protein and its cellular partner(s) in the context of expression of all of the HCV proteins. Consequently, as shown here by confocal microscopy (Figure 2), the HCV proteins are localized within aggregates in the cell cytoplasm which might influence their interaction with PKR, a protein found surrounding the nucleus, in microsomes and in the nucleolus [24, 48].
Several in vitro studies reveal that synthesis of HCV structural proteins or the full-length genome have a direct cytotoxic effect or activate an apoptotic response in osteosarcoma, hepatoma and B cell lines [29, 30, 49–51]. Furthermore, the alteration of ER membranes  and the activation of signalling pathways characteristic of an ER-stress condition, have been found to be associated with the expression of HCV proteins [53–55]. Although these data suggest that HCV may alter intracellular events with possible consequences on liver pathogenesis, the complex mechanism and the role of the viral proteins implicated are currently unknown. As we have shown in this work, expression of most of the HCV genome from VV induces a cell death phenomenon by apoptosis that should contribute to liver pathogenesis. Apoptosis induced by HCV polyprotein expression was prevented by Bcl-2 and by a general caspase inhibitor (Z-VAD-FMK) indicating a caspase-dependent death process. Even though PKR is the main kinase responsible for eIF-2α phosphorylation and for translation inhibition induced by the expression of HCV in VT7-HCV7.9 infected cells, it does not appear to be involved in apoptosis within this system, as revealed from studies performed in PKR+/+ and PKR-/- knockout cells. The extent of apoptosis induction by HCV expression was the same in PKR+/+ and in PKR-/- cells (Figure 9B), suggesting that other pathways may be involved. PKR induces apoptosis in response to activation by different stimuli, such as the accumulation of dsRNA as a by-product during virus replication , or when PKR is overexpressed in cells . Several authors, however, have reported that PKR can also be activated through the binding of heparin and other polyanions [57, 58], or by the cellular activator protein PACT/RAX [59, 60]. The events that mediate induction of apoptosis by PKR have been widely studied and both PKR-induced translational block by phosphorylation of eIF-2α, and NF-kB activation, have been shown to be activated during apoptosis . Since PKR has a number of potential substrates and signalling targets, it is likely that the phosphorylation of eIF-2α by PKR in response to HCV expression is not sufficient to mediate the pro-apoptotic effects of this kinase.
In this study, we also demonstrate the activation of endogenous RNase L and its role in the apoptosis induced by HCV expression (Figure 9A,C). Although it is widely accepted that the IFN-induced proteins PKR and RNase L require the expression of dsRNA for their activation (either directly in the case of PKR or indirectly via 2'-5'-OAS in the case of RNase L), there are several reports that documented the effect of HCV proteins on PKR and 2'-5'-OAS activation. The NS5A and E2 proteins can suppress the PKR pathway [44, 45], whereas the Core protein can transcriptionally activate the 2'-5'-OAS gene through an IRES present within IFN-inducible gene promoter . Like PKR, the 2-5AS/RNase L system can control virus growth by inducing apoptosis in response to viral infection [35, 36]. Overexpression of RNase L or activation of the endogenous enzyme induces apoptosis by a mitochondrial-caspase dependent pathway that is suppressed by Bcl-2 [63–65]. Similarly, apoptosis induced by HCV polyprotein expression was inhibited by Bcl-2 (Figure 8). Although the apoptotic levels induced by HCV proteins remain invariable in PKR+/+, PKR-/-, and RL+/+ cells, the levels are significantly decreased in RL-/- cells, indicating that inducible expression of HCV proteins by VT7-HCV7.9 triggers apoptosis through RNase L in a PKR-independent pathway. Under physiologic conditions, RNase L activity is tightly regulated by 2'-phosphodiesterase and RNase L inhibitor [66, 67] such that only a limited activation of RNase L occurs. The mechanism of the regulation of RNase L inhibitor is unknown, but the reduction of its expression seems to be advantageous for host defence together with the enhanced 2-5 OAS activity. Yu and co-workers  described that hepatic overexpression of PKR mRNA, and reduced expression of an RNase L inhibitor mRNA, are parameters that seem to contribute to an anti-HCV response. In agreement with our results, it has been reported that the absence of RNase L has an anti-apoptotic effect in multiple cell types treated with a variety of different agents . The effects that have been observed in this study upon HCV polyprotein expression from VV are likely to have biological significance during HCV infection as there is ample evidence that VV recombinants can be used to study the function of multiple genes and that the assigned function mimics the effects described in non-viral systems .