We provide here the first demonstration that a functional NES is present in HCV core protein. We show that amino acids 109 to 133 are responsible for the active export of the HCV core protein out of the nucleus, via a CRM-1–mediated nuclear export pathway. This NES was functional in transfected cells and in an in vitro model of HCV replication (HCVcc). The trafficking of core protein into the nucleus early in infection may help to establish infection and facilitate the interaction of core with nuclear molecules, with potentially important pathological consequences.
HCV core protein is thought to be a major viral factor promoting liver disease during HCV infection and the malignant transformation of hepatocytes, leading to the development of HCC through interactions with several host cell factors involved in a wide range of cellular processes. Indeed, in various experimental systems, HCV core has been reported to affect transcription mediated by various gene promoters and apoptosis, thereby contributing to cell transformation. The oncogenic activity of core might be related to its nuclear localization. In liver biopsy samples from HCV-infected patients, HCV core has been found mostly in the cytoplasm, being only rarely detected in the nucleus of infected hepatocytes. Nevertheless, a nuclear location of truncated core proteins was detected in tumor tissues from patients with HCV-related hepatocarcinoma. Similarly, the nuclear accumulation of core has been observed in transgenic mice producing the HCV core protein and developing HCC.
Taking into account the role of HCV core as a viral factor of major pathological significance and understanding the mechanisms regulating its subcellular distribution and trafficking are of critical importance. Several studies on transfected cells have shown the HCV core protein to be located in the cytoplasm or nucleus, depending on its length. Consistent with this dual localization, many studies have reported interactions with molecules located in either the cytoplasm or the nucleus.
Our findings confirmed that, in an in vitro transfection system based on human Huh7 and HepG2 hepatoma cell lines or immortalized Fa2-N4 human hepatocytes, the aa(1–173) and aa(1–160) core proteins were found in both the nucleus and the cytoplasm. These proteins were found exclusively in the nuclei of non human cells (CHO, ARL-6) and in human cells of non hepatic origin (HEK-293). By contrast, the shorter core proteins aa(1–120) and aa(1–140) were found exclusively in the nucleus. Thus, the nuclear/cytoplasmic subcellular distribution of core proteins aa(1–173) and aa(1–160) was specific to human cells of hepatic origin. Our observations suggest that core protein may contain signals for specific transport mechanisms controlling its distribution between the nucleus and the cytoplasm that are functional in human hepatic cells.
The differences in the distribution of the protein between the nucleus and cytoplasm in the cell types tested may reflect the availability and/or functionality of the carrier proteins in these cells. Indeed, the subcellular distribution of a given protein may be controlled by the differential expression of carrier proteins in various tissues or host species, and may depend on the differentiation status of the cell.
The nucleocytoplasmic trafficking of various proteins and RNA is controlled by importins and exportins (also called karyopherins) from the importin-β superfamily of proteins. These proteins can therefore gain entry into the nucleus only if they possess the appropriate NLS recognized by nuclear importin receptors, or if they react directly with the nuclear pore complex. Consistent with the nuclear localization of core in several experimental systems, three NLS have been identified in the N-terminal domain of this protein. These signals consist of clusters of basic amino acids in the aa(5–13), aa(38–43), and aa(58–71) regions; they are functional and able to target core to the nucleus.
For reentry into the cytoplasm, proteins must contain the sequences required for interaction with export factors (exportins), enabling them to leave the nucleus via the nuclear pore. The nuclear export of proteins is mediated mostly by NES, leucine-rich aa sequences recognized by soluble export receptors, such as Exportin1 (CRM-1). No functional export signal, directing translocation of the core protein from the nucleus to the cytoplasm has been identified in the core sequence to date, although the presence of such a signal has been suggested.
We used NetNES software to search for a classical leucine-rich NES in HCV core, and found such a signal in the C-terminal part of the immature form of core at aa(179–187). This protein is always found in cytoplasmic compartment and, thus, the role of this signal remains unclear. In addition, this NES sequence is removed when the mature core protein is generated by cellular peptide peptidase processing.
In addition to this classical leucine-rich NES in the C-terminus of the immature core protein, we identified a second, “atypical” NES in the mature form of core, by sequence alignment, as previously reported by Rowland et al. for the identification of NES in the PPRSV nucleocapsid protein. Indeed, multiple sequence alignments of the HCV core with other viral NES, generated with CLUSTAL W, led to the identification of a candidate NES sequence at aa(109–133). Consistent with the putative biological role of this NES, this sequence contains a cluster of hydrophobic residues with a sequence similar to those of the NES of HSV ICP27, HIV Rev, equine infectious anemia virus Rev, feline immunodeficiency virus Rev and porcine reproductive and respiratory syndrome virus N protein. Moreover, this signal was well conserved in consensus sequences from various HCV genotypes.
We provide evidence that the NES identified at aa(109–133) is functional in transfected hepatoma cells and can export core protein fragments from the nucleus to the cytoplasm. First, a peptide corresponding to aa(109–133) of HCV core, like the NES of the HIV Rev protein, counteracted the nuclear translocation driven by the strong NLS of SV40 large T-antigen. The nuclear export of this NES, observed by fluorescence microscopy, was confirmed by quantitative analyses. LMB treatment inhibited nuclear export of the core NES, demonstrating that the aa(109–133) nuclear export signal functions in a CRM-1-dependent manner. Indeed, “atypical” NES sequences may also function in a CRM1-dependent manner.
Export signals essentially consist of several closely spaced leucine residues, but other hydrophobic amino acids, such as methionine, isoleucine, valine, phenylalanine and tryptophan, may replace leucine in the recognition motifs. Using comprehensive alanine-scanning mutagenesis we showed that the replacement of the Leu(119), Ile(123) and Leu(126) residues by alanine modified the export capacity of our NES. Indeed, whereas the wild-type chimeric protein encoded by the plasmid (EGFP-NLSSV40-NEScore) carrying the intact aa(109–133) sequence was exported into the cytoplasm, the protein encoded by the mutated construct was mostly found in the nucleus. These observations, validated by quantitative analyses, confirmed that the aa(109–133) core sequence acted as a functional NES, targeting the protein from the nucleus to the cytoplasm.
The functionality of the NES at aa(109–133) was further analyzed in the context of mature HCV core protein. Firstly, LMB treatment also modified the subcellular distribution of core aa(1–173), increasing its accumulation in the nucleus. Secondly, mutations of the hydrophobic residues Leu(119), Ile(123) and Leu(126), with the replacement of these residues by alanine residues, modified the export capacity of the protein. Our quantitative studies confirmed that the proportions of the protein located in the nucleus (as opposed to the cytoplasm) were significantly higher for the mutated protein than for the wild-type protein. These data provide experimental evidence for a role of the identified NES in the CRM-1-dependent nuclear export of HCV core.
CRM-1 has been identified as the export receptor, the principal mediator of nuclear export, allowing the nuclear-cytoplasmic shuttling of proteins and RNA between cellular compartments. CRM-1-dependent transport is well conserved throughout eukaryotes and LMB is a recognized inhibitor of the active export of most molecules from the nucleus. Indeed, the “steady-state” localization of proteins does not always reflect the biological importance of their site of action. The use of LMB to block CRM-1-mediated nuclear export results in the detection of the NES-containing protein in the nuclear compartment, although its “steady-state” location appears to be exclusively or predominantly cytoplasmic, with the equilibrium of bidirectional transport favoring nuclear export. Nevertheless, CRM-1-mediated transport is a highly regulated process and this regulation includes the masking of NES, phosphorylation and heterodimerisation of the protein and the formation of disulfide bonds by an oxidative process. The availability of specific cofactors may also influence this regulation. The presence of such cofactors may contribute to the observed differences in the subcellular distribution of core proteins in various cell types (see above).
In our study, both the NES aa(109–133) and the adjacent hydrophobic sequence in domain II were required for a cytoplasmic distribution of core. Indeed, we found that only a relatively long core protein fragment, aa(109–160), was located in the cytoplasm, with shorter core fragments lacking either NES or this adjacent hydrophobic sequence being found in the nucleus. We showed that no other NES capable of enhancing core export from the nucleus was present in the adjacent (133–160)aa sequence. However this core fragment, in addition to the NES, was necessary for a cytoplasmic distribution of the protein. As this hydrophobic core fragment contains specific sites interacting with LDs and membranes, these interactions are probably required, in addition to our NES aa(109–133), for the maintenance of core protein in the cytoplasm. Consistent with this notion, the aa(120–150) core protein (containing a hydrophobic fragment interacting with LDs/membranes but not the NES) was found in the nucleus rather than the ER. Similarly, C-terminally truncated core proteins with NES, but lacking the hydrophobic fragment of DII, accumulated in the nucleus, thereby probably contributing to the development of HCC. Collectively, our observations indicate that both a nuclear export signal aa(109–133) and the adjacent hydrophobic sequence in domain II are required to target core protein to the cytoplasm and to keep it in this compartment.
The identification of a NES that was functional in transfected Huh7 hepatoma cells raised questions about the role of this signal in HCV infection. In the context of the infectious cell culture model in Huh7.5 cells (HCVcc), HCV core protein was found to colocalize with lipid droplets and membranes, but was not detected in the nucleus in most studies. However, knockdown of the PA28γ proteasome activator gene, blocking ubiquitin-independent nuclear degradation, led to the detection of a very small quantity of HCV core in the nuclei of infected Huh7 cells on immunofluorescence analysis.
Our key findings for transfected cells are consistent with the data obtained in the HCVcc replication model. First, our immunoelectron microscopy studies based on staining with anti-core antibodies provided evidence for a nuclear location of core as early as 20 minutes after the start of infection. This suggests that the nuclear trafficking of core takes place very early in the viral cycle, shortly after internalization of the virus. Second, the use of LMB to block CRM-1-dependent export resulted in the detection of HCV core protein in the nuclei of a number of JFH1-infected cells on confocal microscopy. Nuclear staining for HCV core was observed as several “spots” when various nuclear images (virtual cross sections) were analyzed by confocal microscopy. In particular, core was detected within the nucleus only if LMB treatment was applied early in infection (2 h post infection), and no core was detected in the nucleus if LMB was applied late in infection (48 h post infection). Intranuclear core was also observed in Huh7.5 cells in which PA28γ expression was knocked down but only after LMB treatment.
Early LMB treatment also decreased HCV RNA production, suggesting that the early shuttling of core between the cytoplasm and the nucleus may be important for virus multiplication. LMB had no toxic effect on the cells tested, but we cannot rule out the possibility that LMB treatment also influences other cell processes, in addition to core protein shuttling.
We conclude that the NES identified in HCV core protein is functional in the HCVcc replication system. Our electron microscopy studies suggest that some HCV core, derived from invading virus particles, is transported into the nucleus at very early stages of the viral life cycle. In addition, a fraction of core protein can be detected by immunofluorescence, in the nucleus, a few hours after infection, subsequently being exported to the cytoplasm in a CRM-1-dependent manner, as this export is blocked by LMB, a drug widely used to dissect nuclear export pathways.
Examples of viral proteins known to shuttle through the nuclear pore complex and for which the CRM-1-dependent pathway is known to export the corresponding viral RNA include HIV Rev and T-cell leukemia virus type 1 Rex. Structural and nonstructural proteins of several members of the flavivirus family, such as Japanese encephalitis virus (JEV), Dengue virus (DENV), and Kunjin virus (KUN), have also been shown to be actively translocated to the nucleus or to the nucleoli of infected cells, even when these viruses multiply entirely in the cell cytoplasm. This phenomenon may affect virus infectivity or disease pathogenesis. Indeed, DENV NS5 RNA polymerase can be detected in the nucleus very shortly after infection, and this protein is exported from the nucleus in a CRM-1-dependent manner. Nuclear NS5 suppresses the production of IL-8, a cytokine playing an important role in the antiviral response.
DENV core protein also localizes to the nucleus (and nucleoli) at very early stages in the viral life cycle, due to its bipartite NLS. The mechanisms of DENV nuclear export remain unknown, as this process is insensitive to LMB, suggesting that it does not require a functional NES 
. Indeed, many other pathways exist for protein import and export, including the calreticulin pathway. Nevertheless, the nuclear localization of DENV core may regulate its replication cycle and apoptosis (see for review. Consequently, the construction of recombinant vaccines based on viral proteins deficient in nuclear trafficking signals could potentially lead to attenuation of the virus.
Our findings are consistent with the hypothesis that at least some HCV core protein is trafficked between the cytoplasmic and nuclear compartments early in HCV infection. Recombinant viruses mutated in the NES region investigated here, showed impaired virus production, producing less than 1% of the wild type virus in the HCVcc in vitro model. These observations provided evidence that this sequence is of importance for virus life cycle. The absence of the NES identified in this study and/or of the hydrophobic fragment of domain II (which act together to keep the protein in the cytoplasm) may account for the nuclear localization of the C-terminally truncated core proteins in patients with HCV-induced HCC and contribute to the cell transformation.