Cell culture

All cell lines were cultured at 37°C, in a humidified atmosphere containing 5% CO₂. The human hepatoma cell line Huh7, the rat hepatoma cell line ARL-6 and human embryonic kidney 293T cells were obtained from ATTC, immortalized human hepatocytes Fa2N-4 were purchased from XenoTech, Lenexa KS, USA, whereas the Huh 7.5 cell line was kindly provided by C. Rice. Cells were maintained in Dulbecco's modified Eagle's medium (DMEM; Gibco-Invitrogen) supplemented with 10% fetal calf serum, 1% non-essential aminoacids and antibiotics (penicillin 100 U/ml and streptomycin 100 µg/ml). The CHO-K1 cell line (ATCC#CCL-61) was obtained from LGC Prochem (Molsheim, France) and cultured in F12 medium (Invitrogen, Cergy-Pontoise, France) containing glutamine and supplemented with 10% fetal calf serum (FCS) antibiotics and anti-fungal compounds.

Plasmid construction

Plasmids were constructed by inserting core-encoding fragments of various lengths — aa(1–120), aa(1–140), aa(1–160), aa(1–173) — into pEGFPN1 (Clontech). These fragments were amplified by standard PCR, cleaved with EcoRI and BamHI and inserted into pEGFPN1.

We constructed several plasmids encoding enhanced green fluorescent protein (EGFP) or mCherry, to determine whether the NES and NLS sequences were functional. The reporter genes were fused with (i) the NLS of SV40 nuclear T antigen (EGFP-NLSSV40), (ii) the NLS of SV40 followed by the potential NES of the HCV core protein aa(109–133) (EGFP-NLSSV40-NEScore), or (iii) the NLS of SV40 followed by the functional NES of the HIV regulatory protein Rev (EGFP-NLSSV40-NESRev). Three more plasmids, encoding core fragments aa(109–160), aa(144–160) and aa(134–143), were constructed. These fragments were fused with the EGFP or mCherry sequence and the NLS of SV40 (EGFP-NLSSV40-core(109–160); EGFP-NLSSV40-core(144–160); EGFP-NLSSV40-core(134–143)).

All viral coding sequences were amplified by standard PCR and introduced into the pDONR207 plasmid (Invitrogen) via a recombinational cloning strategy (Gateway; Invitrogen). Viral sequences were then transferred from pDONR207 into Gateway-compatible versions of pEGFP-C1 (kindly provided by Y. Jacob) or mCherry vectors (Clontech), for expression in mammalian cells with an N-terminal EGFP or mCherry tag.

The PCR primers used to amplify and clone viral ORFs had 20 to 30 specific nucleotides matching the ends of the ORF, giving a Tm close to 60°C. For the recombinational cloning of PCR products, the 5′ ends of forward primers were fused to the attB1 5′-GGGGACAAGTTTGTACAAAAAAGCAGGCATG-3′ recombination sequence, whereas the reverse primers were fused to the attB2 5′-GGGGACCACTTTGTACAAGAAAGCTGGTTA-3′ recombination sequence. All the constructs used were amplified by transforming (by heat shock) Escherichia coli DH5α (Invitrogen).

Cell transfection

For overexpression, mammalian cells were transfected with pEGFP-C1 or mCherry vectors containing the viral ORFs, in the presence of the FuGene Transfection Reagent (Promega), according to the manufacturer's instructions. Unless otherwise specified, we dispensed 5x10⁴ cells per into each well of 24-well plates and, 24 to 48 h later, we transfected these cells with 500 ng of plasmid DNA per well.

Mutations in the NES sequence

Mutations (L119A, I123A, and L126A) were introduced into the NES sequence of HCV core protein with the QuikChange Lightning site-directed mutagenesis kit (Agilent-Stratagene). An EGFP expression plasmid (EGFP-NLSSV40-NEScore) encoding the putative HCV core export sequence aa(109–133) or EGFP-labeled core protein aa(1–173) was amplified with the 5′-phosphorylated mutagenic primers 5′-GCCGATACCGCTACATGCGGCTTCGCCGACCTCAT and 5′-GACCTTACCCGCATTACGCGACCTACGCCGGGGGT. Template DNA was then digested with Dpn1 and mutagenized plasmids were circularized by self-ligation with T4 DNA Ligase (New-England Biolabs). Mutant clones were selected and the sequences of their expression cassettes were checked.

Sequence analyses

Multiple sequence alignment was carried out with Clustal W. “Classical” NES were identified with NetNes software (www.expasy.org).

HCVcc (JFH1) cell culture

The plasmid corresponding to the genome of the JFH1 strain was kindly provided by T. Wakita and used to generate cell culture-produced virus (HCVcc). The virus was cultured as previously described. For cell infection, monolayers of Huh 7.5 cells were grown for 24 h in tissue culture plates. They were then inoculated by incubation with 25 µl of the virus preparation (containing approximately 10⁶ IU of HCV RNA) for 2 h at 37°C, to allow infection. Cells were analyzed at various time points after infection, as indicated, by immunofluorescence studies, quantitative RT-PCR or electron microscopy.

RNA interference

For RNA interference-based knockdown experiments, a 25-nucleotide siRNA (ccu cgu ugc uga agg ugg auc agg a) targeting PA28γ was purchased from Invitrogen (Stealth Select RNAi). Huh7.5 cells were tranfected with 20 nM siRNA, in the presence of JetPRIME (Ozyme). Control ON-TARGET plus (Dharmacon) non targeting siRNA was used to confirm silencing specificity. For analyses of the influence of proteasome activator silencing on the subcellular distribution of core, cells were transfected with siRNA targeting PA28γ or with control siRNA 18 h before infection with HCVcc.

LMB treatment

We investigated the role of CRM-1-dependent transport in the subcellular localization of core, by treating cells with leptomycin B (LMB, Sigma). Huh7 cells were cultured and transfected as described above. After 22 to 48 h, the cells were treated for 2 to 4 h, at 37°C, with LMB at a concentration of 2 ng/ml or 10 ng/ml in DMEM. For studies in the JFH1 infection model, 10 ng/ml LMB was applied at various time points during or after infection, and cells were incubated in a medium containing LMB for the indicated times.

Cell viability after LMB treatment was determined by counting live and dead cells, after trypan blue staining, in an automated cell counter (Countess; Invitrogen). We also quantified the ATP present in cultured cells, with the CellTiter-Glo® Assay (Promega), used according to the manufacturer's instructions. Untreated cells and cells treated with 10% DMSO (to induce cell death) were used as negative and positive controls, respectively.

Immunofluorescence and confocal microscopy

We investigated PA28γ production in untransfected Huh 7.5 cells and in Huh 7.5 cells transfected with an siRNA targeting PA28γ (or control siRNA), with a rabbit anti-PA28γ antibody and Alexa Fluor 488-conjugated anti-rabbit IgG. The distribution of HCV core in infected cells was analyzed with a monoclonal anti-core antibody, ACAP27, followed by an Alexa Fluor 568-tagged anti-mouse IgG. Staining with rabbit anti-lamin B antibody followed by Alexa Fluor 488-conjugated anti-rabbit IgG was used to outline the cell nuclei.

For fluorescence microscopy, infected or transfected cells were washed in PBS and fixed by incubation with 4% paraformaldehyde (PFA) in PBS for 20 min at 4°C. The cells were washed several times in PBS and permeabilized with 0.5% Triton X-100. They were then incubated in a blocking buffer containing 1% gelatin and 0.1% Tween 20 in PBS. Primary antibodies were incubated with the cells for 2 h (mostly at a concentration of 1 µg/ml in blocking buffer). After subsequent washes with 0.1% Tween 20 in PBS, secondary antibodies were added and the cells were incubated for 1 h. The cells were washed again and mounted in Vectashield medium containing DAPI (Vector Laboratories, Abcys, France). Fluorescent fields were captured with a Widefield ApoTome AxioCam upright (66RC10P) microscope.

Relative fluorescence analyses

We analyzed the relative fluorescence intensity of the proteins in the cytoplasm and nucleus, by converting bright-field immunofluorescence images to grayscale images, with Image J software. Boundaries were applied to demarcate the nuclear and cytoplasmic compartments and the fluorescence intensity of each compartment was measured with a script created with Acapella 2.0 image software (Perkin Elmer). All measurements were normalized with respect to background fluorescence. A nonparametric one-way ANOVA assay was carried out to compare fluorescence between the cytoplasm and the nucleus of 1138, 712 and 573 cells transfected with plasmids (EGFP-NLSSV40), (EGFP-NLSSV40-NESRev) and (EGFP-NLSSV40-NEScore), respectively. Graphs showing the means and variances of the ratio of nuclear to cytoplasmic fluorescence intensities were plotted for the three constructs considered.

For analyses of the mutated NES, 153 cells transfected with the wild-type (EGFP-NLSSV40-NEScore) plasmid and 130 cells transfected with the mutated construct were analyzed, and the results are presented as a graph showing the means and variances of the ratio of nuclear to cytoplasmic fluorescence intensities for wild-type and mutated constructs. Similarly, 139 and 101 cells transfected with the wild-type and mutated core protein aa (1–173) constructs, respectively, were also considered. Graphs showing the means and variances of the ratio of nuclear to cytoplasmic fluorescence intensities were plotted for the wild-type and mutant proteins and a nonparametric t-test was used to evaluate the results obtained.

RT-quantitative PCR (RT-qPCR)

The HCV RNA associated with cells was quantified by one-step real-time RT-qPCR, with the SuperScript III Platinum One-Step qRT-PCR Kit (Invitrogen). The 5′-AGYGTTGGGTYGCGAAAG-3′ and 5′-CACTCGCAAGCRCCCT-3′ primers were used to amplify HCV RNA from JFH clones, and 6-FAM-CCTTGTGGTACTGCCTGA-MGB (Applied Biosystems, Foster City, CA, USA) was used as an internal probe. Real-time detection of the PCR products was carried out with an AbiPrism 7000 machine. HCV RNA was quantified and standardized with an HCV RNA quantification panel from AcroMetrix, and the values obtained are expressed as HCV RNA IU.

Western Blot

Huh7 cells (1.5×10⁴) were transfected with plasmids encoding core proteins of various lengths — aa(1–140), aa(1–160), aa(1–173) — fused to EGFP. Transfection was performed with jetPRIME™ Transfection Reagent (Polyplus Transfection Company, France), according to the manufacturer's instructions. After 24 h, the transfected cells were washed with PBS and lysed by incubation for 5 min on ice with 900 µl/well 1X RIPA buffer (20 mM Tris-HCl pH7.5, 150 mM NaCl, 1 mM Na₂EDTA, 1 mM EGTA, 1% NP-40, 1% sodium deoxycholate, 2.5 mM sodium pyrophosphate, 1 mM Na₃VO₄, 1 µg/ml leupeptin). Cells were scraped of the plates and collected by centrifugation for 10 min. at 14,000 rpm at 4°C. Supernatants were collected and separated by SDS-PAGE in 10% polyacrylamide gels and the protein bands were transferred to nitrocellulose membranes (Pure Nitrocellulose membrane, Bio-Rad). Proteins were detected by incubation with polyclonal anti-GFP (Roche) and anti-tubulin (Abcam) antibodies, followed by peroxidase-conjugated anti-rabbit immunoglobulin antibodies (Santa Cruz Biotechnology). The proteins bands were detected by enhanced chemiluminescence (Sigma).

Electron microscopy

Supernatants collected from Huh7.5 cells producing JFH1 were concentrated by centrifugation through a 20% sucrose cushion for 4 h at 32,000 rpm, in an SW32Ti rotor. For electron microscopy analyses, Huh7.5 cells were grown in 3 cm dishes for 24 h, washed with cold serum-free-DMEM and maintained for 10 min at 4°C. The medium was then replaced with 1 ml cold serum-free DMEM supplemented with a concentrated virus preparation (approx. 5×10⁹ IU/ml HCV RNA) and the cells were incubated for 30 min at 4°C. At the end of this adsorption step, the cells were placed for 20 min at 37°C and examined by transmission electron microscopy (TEM). Briefly, cells were fixed by incubation with 4% (v/v) paraformaldehyde (Delta Microscopies, Ayguevive, France) and 0.05% (v/v) glutaraldehyde in phosphate-buffered saline (PBS) pH 7.4, for 30 min at 37°C. They were washed with PBS and incubated for 10 min in blocking solution containing 0.5% v/v cold fish skin gelatin (FSG), 0.1% v/v saponin and 0.02 M glycine (all from Sigma Aldrich) in PBS pH 7.4, at room temperature. The cells were washed with 0.5% FSG in PBS (PBS/FSG) and incubated for 60 min with ACAP-27 monoclonal antibody, kindly provided by JF. Delagneau, (0.6 µg/ml, diluted in PBS/FSG). The cells were again washed in PBS/FSG, incubated for 30 min with anti-mouse IgG conjugated with Nanogold®particles (Yaphank, NY, USA) in PBS/FSG, washed in PBS/FSG and fixed by incubation with 1% v/v glutaraldehyde in PBS for 1 h at room temperature. The cells were washed in water and Nanogold® staining was enhanced by incubation for three minutes with the HQ silver enhancement kit (Nanoprobe, Yaphank, NY, USA). Cells were then post-fixed by incubation with 1% v/v osmium tetroxide in PBS for 1 h and processed by dehydration in a series of ethanol solutions for embedding in epoxy resin. TEM examination was performed on a Philips CM120 microscope, and images were acquired with an ORIUS SC200D CCD camera (Gatan Inc.).

Subcellular localization of core proteins

Several studies have shown that, when produced separately, the immature core protein aa(1–191) remains in the cytoplasm and the processed, mature core protein is present in both the nuclear and cytoplasmic compartments, whereas the shorter core proteins aa(1–120) or aa(1–140) are targeted to the nucleus.

Using plasmids encoding EGFP-labeled core proteins composed of aa(1–173), aa(1–160) or aa(1–140), we investigated the subcellular distribution of the protein in various cell lines. The short form of core aa(1–140) was found in the nucleus in all cell lines tested. By contrast, core proteins aa(1–160) and aa (1–173) were found in both the cytoplasmic and nuclear compartments in transfected Huh 7 cells, but solely in the nucleus of cells of non human origin, such as Chinese hamster ovary cells (CHO) (Figure 2) and rat hepatoma cells (ARL-6) (not shown). Results identical to those for Huh7 cells were obtained for other hepatic cells (HepG2 hepatoma cell line and Fa2N4 immortalized human hepatocytes) in which HCV core proteins aa(1–160) and aa (1–173) were found in both the cytoplasmic and nuclear compartments. By contrast, in human non hepatic cells, such as embryonic kidney cells (HEK-293) these two core proteins were found exclusively in the nucleus (not shown), as for CHO and ARL-6 cells.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 2. Subcellular localization of HCV core proteins of different lengths in CHO and Huh 7 cells.

Plasmids encoding EGFP-labeled core proteins composed of aa(1–140) (a,d), aa(1–160) (b,c), and aa(1–173) (c,f) were used to investigate the subcellular distributions of the encoded proteins in the several human and non-human cell lines. Cells were transfected with plasmids encoding the corresponding EGFP-labeled proteins, grown for 24 h and analyzed by fluorescence microscopy, as described in the Materials and Methods section. (a–c) results obtained for CHO cell line (the same type of distribution was observed for ARL-6 and HEK-293 cells) and (d–f) distribution representative for Huh 7 cells (and other cells of hepatic origin as HepG2 or Fa2-N4). Bar represents 10 µm.

https://doi.org/10.1371/journal.pone.0025854.g002

We performed Western Blot analyses of the EGFP-labeled core proteins generated, to check their integrity and to exclude the possibility of degradation of the constructs used for these experiments, which might affect the subcellular distribution of the corresponding proteins. These experiments (shown in Figure 3) confirmed that only one EGFP-labeled protein was produced in transfected Huh7 cells for each construct: a 47 kDa EGFP-labeled core protein for the plasmid encoding core aa(1–173), a 45 kDa EGFP-labeled protein for the plasmid encoding aa(1–160) and a 43 kDa EGFP-labeled protein for the plasmid encoding aa(1–140). There was therefore no degradation of the constructs and the distribution of core proteins was specific to the cells analyzed.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 3. Western Blot analyses of constructs used for cell transfection.

The integrity of the constructs used to transfect cells (shown in Figure 2) was assessed by Western Blotting. After separation of the proteins in cell lysates by SDS-PAGE, the bands were transferred to nitrocellulose membranes and for detection with anti-EGFP and anti-tubulin antibodies followed by peroxidase-labeled anti-rabbit IgG. The protein bands were detected by chemiluminescence. 1. Lysed Huh 7 cells; 2-Huh 7 cells transfected with pEGFPN1, 3-Huh7 cells transfected with core sequence aa(1-173) inserted into pEGFPN1; 4-Huh 7 cells transfected with the core aa(1–160) sequence in pEGFPN1; Huh 7 cells transfected with core sequence aa(1–140) inserted into pEGFPN1.

https://doi.org/10.1371/journal.pone.0025854.g003

These results suggest that, in addition to the previously described NLS, there may be other signals in HCV core and interacting sequences in human hepatic cells, regulating the subcellular distribution of core protein. These mechanisms may be nonfunctional or differently regulated in non human and human cells of non hepatic origin.

Identification of a “classical” NES in the immature form of HCV core

The subcellular distribution of the HCV core protein, like that of other viral proteins, may be influenced by the presence of nuclear localization signals (NLS) and nuclear export signals (NES). Indeed, the tripartite NLS located in the N-terminal DI have been described and shown to be functional. These and other functional domains of the core protein are shown in Figure 1. However, no NES capable of directing the translocation of the protein from the cell nucleus to the cytoplasm has ever been reported in HCV core.

We therefore used NetNES software (www.expasy.org) to search for a “classical” NES in the core protein. Such “classical sequences” are relatively short linear oligopetides enriched in leucine residues, usually consisting of the peptide sequence X-R(2-4)X-R2-X-R-X, where X is leucine, isoleucine or valine and R is any amino acid. This analysis identified a “classical” leucine-rich nuclear export sequence (NES Classic) within the C-terminal part of core, between aa 179 and 187 (Figure 4). Nevertheless, it is well established that sequential cleavages by signal peptidase and signal peptide peptidase remove the C-terminal hydrophobic DIII of core, to yield the mature form of core composed of aa(1–173), more recently defined as aa(1–177). Thus, the immature form core aa(1–191), attached to membranes, contains the classical NES, whereas the mature form of core does not (Figure 1).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 4. Identification of a classical NES in the immature form of HCV core protein.

A leucine-rich NES with an LXL motif was identified in domain III of core, between aa179 and 187, with NetNES Software. This sequence is absent from the mature form of core, because it is cleaved by SPP during core processing. Graphical plot of the values (NES score) calculated by the prediction server from the Markov Model (HMM), and Artificial Neural Network (NN) scores. If the calculated NES score exceeds the threshold, then the residue concerned is predicted to be involved in a nuclear export signal.

https://doi.org/10.1371/journal.pone.0025854.g004

Identification of a “non classical” NES in HCV core

As the mature core (aa(1–173)) was found in both the cytoplasm and the nucleus, another “non classical” or “atypical” NES is probably present in the mature form of core. Such sequences have been reported for several viral proteins, such as equine infectious anemia virus (EIAV) and feline immunodeficiency virus (FIV) Rev.

We therefore searched for another, “non classical” NES in the mature form of core by generating multiple sequence alignments and comparing the core aa sequence with known viral sequences enriched in hydrophobic residues that can function as NES motifs, as described by Rowland et al. 2003 for the N–protein of porcine reproductive and respiratory syndrome virus (PPRSV). We found that the aa(109–133) region of core contained a sequence similar to that of other “atypical” viral NES. An analysis comparing the putative NES of core with other viral sequences is shown in Figure 5A.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 5. Identification of a functional, “non-classical” NES” in the core protein.

(A) CLUSTAL W (1.81) software was used for multiple sequence alignment analysis, leading to the identification of a “non classical” NES sequence in domain II of the core protein. The potential NES signal aa(109–133) in core was compared with known viral NES sequences. Underlined regions correspond to hydrophobic amino-acid residues, and letters in bold typeface identify the conserved LXL motifs. The frame delineates a region of the export sequence containing amino-acid residues L(119), I(123) and L(126), which were replaced by alanine residues (the corresponding immunofluorescence analyses are shown in Figure 6). (B) The amino-acid sequences of the fragment corresponding to the putative NES aa(109–133) in HCV core proteins are well conserved in different HCV genotypes. The consensus sequences are shown, for each virus genotype, and were obtained by the alignment of 1245 sequences corresponding to the putative NES for HCV type 1A, 2078 sequences for HCV type 1B, and 95, 264, 60, 12 and 121 sequences for HCV types 2, 3, 4, 5 and 6, respectively. Sequences were obtained from the Los Alamos Data Bank (National Institutes of Health). (C) Schematic diagram of the plasmids used to investigate the functionality of the putative export sequence of core, aa(109–133). The SV40 NLS was used as a nuclear reporter, and the NES of the HIV Rev protein was used as a control export signal. The sequences shown were fused to either EGFP or m-Cherry, to allow the visualization of proteins in transfected cells. (D) Subcellular distribution of the proteins encoded by the plasmids depicted above. Huh7 cells grown on coverslips were transfected with the appropriate plasmids; 40 h after transfection, the cells were fixed in 4% PFA and examined by fluorescence microscopy. Panels a-c represent proteins labeled with EGFP, d-f the equivalent proteins labeled with m-Cherry. Proteins containing only the SV40 NLS were present mostly in the cell nuclei (a, d); proteins containing the control SV40 NLS and Rev NES were found in both the nucleus and the cytoplasm (b, e). The core sequence containing a putative NES, aa(109–133), is functional, because it was exported from the nucleus to the cytoplasm (c, f), like HIV Rev NES (b,e). Staining of nuclei with DAPI. (E) Graphical representation of nonparametric one-way ANOVA of the ratios of fluorescence between the nucleus and cytoplasm for the three plasmids. Cytoplasmic fluorescence is significantly higher for both EGFP-NLSSV40-NESRev and EGFP-NLSSV40-core aa(109–133).

https://doi.org/10.1371/journal.pone.0025854.g005

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 6. Fluorescence is predominantly nuclear for the protein encoded by the mutated NES.

(A) The subcellular distribution of the proteins encoded by the construct with the mutated nuclear export sequence in HCV core was investigated by immunofluorescence. Mutations were introduced into the NES, as shown in Figure 5 A: amino-acid residues L(119), I(123) and L(126) were replaced by alanine residues (underlined within the frame). (A) Plasmids containing the EGFP, NLSSV40-core aa(109–133) and EGFP NLSSV40-mutated core aa(109–133) sequences were used to transfect Huh 7 cells, as described for Figure 5. After transfection (40 h), the cells were fixed with 4% PFA and examined by fluorescence microscopy. Upper panel represents immunofluorescence analyses of the subcellular distribution of the protein produced by the wild-type EGFP-NLSSV40-core aa(109–133) construct, lower panel shows the distribution of the corresponding construct with a mutated NES. The protein encoded by the wild-type construct (W.T) was found in both the nucleus and the cytoplasm; the protein encoded by the mutated construct (Mut) was found mostly in cell nuclei. Staining of the nuclei with DAPI. (B) EGFP-tagged core protein encoded by the wild type core aa(1–173) construct was found in both the nucleus and the cytoplasm (upper panel); the protein encoded by the mutated construct (Mut) was found mostly in cell nuclei (lower panel). Staining of the nuclei with DAPI. Graphical representation of subcellular distribution of the fluoresecence signal, based on a nonparametric t-test is shown on the right side of each panel. The graph shows the ratios of fluorescence between the nucleus and cytoplasm for the two plasmids encoding the wild-type and mutated proteins.

https://doi.org/10.1371/journal.pone.0025854.g006

Comparative analyses of consensus sequences available for the core proteins of different genotypes provided evidence that this region was well conserved in various HCV genotypes (Figure 5B).

The aa(109–133) sequence is a functional NES

We investigated whether the putative NES sequence (aa(109–133)) could direct the export of the protein from the nucleus to the cytoplasm, using an approach previously used to characterize the NES of PPRSV and HSV proteins. We constructed three plasmids (shown in Figure 5C). In the first, EGFP or mCherry was C-terminally tagged with the NLS of SV40 large T antigen, which targets proteins to the nucleus. The second plasmid encoded the SV40 NLS fused to the potential NES sequence of the HCV core protein aa(109–133). Finally, EGFP or mCherry fused to the SV40 NLS and the functional NES of the HIV Rev regulatory protein was used as a control for nuclear export. The Rev protein of HIV is a shuttling protein that promotes the nuclear export of mRNAs encoding viral structural proteins in a CRM-1-dependent manner.

We transiently transfected Huh7 cells with these plasmids and, after 40 h, we determined the subcellular distribution of the encoded proteins by fluorescence and confocal microscopy (Figure 5D). Fluorescent proteins containing the SV40 NLS were found almost exclusively in the nucleus, whereas proteins containing both the SV40 NLS and the Rev NES were found in both the nucleus and the cytoplasm. Similarly, the protein containing the SV40 NLS and the putative core NES was found in both the nucleus and the cytoplasm, providing evidence that aa(109–133) of core is a functional NES capable of directing the chimeric protein from the nucleus to the cytoplasm. Similar results were obtained with plasmids encoding proteins tagged either with EGFP or with mCherry, excluding the possibility that the subcellular distribution of the encoded proteins was dependent on the fluorescent tag used.

We confirmed the subcellular distribution of the putative NES of core (aa(109–133)) and of the proteins encoded by control plasmids, in quantitative analyses (Figure 5E). These analyses provided evidence that the aa(109–133) core sequence acted as a functional NES, targeting the protein from the nucleus to the cytoplasm, like the HIV Rev protein NES.

Mutation of the NES sequence changes the subcellular distribution of HCV core

We hypothesized that the hydrophobic residues L(119), I(123) and L(126) in the core NES (Figure 5A, underlined within the frame) might be important for the nuclear export of this protein. We tested this hypothesis by carrying out comprehensive alanine-scanning mutagenesis and determining whether the replacement of these residues by alanine residues decreased the nuclear export of the protein, thereby resulting in accumulation of larger amounts of protein in the nucleus.

Cells were transfected with mutated constructs and the subcellular distributions of the resulting proteins were analyzed by immunofluorescence microscopy and compared with that of the non mutated control.

Transfection efficiency for the construct containing SV40 NLS fused to the mutated putative export sequence aa(109–133) of the core protein was similar to that for the wild-type construct, but the mutated protein had a different subcellular distribution. Indeed, whereas the wild-type chimeric protein with a putative export sequence aa(109–133) was found principally in the cytoplasm, the protein encoded by the mutated construct was mostly nuclear (Figure 6 A). Quantitative analyses confirmed that mutations affecting this core fragment significantly decreased the nuclear export of the protein produced and thus provided evidence that the aa(109–133) core sequence acted as a functional NES, targeting the protein from the nucleus to the cytoplasm.

The C-terminal aa(140-160) sequence contributes to the cytoplasmic localization of core

We have demonstrated above that aa(109–133) is a functional NES. Consistent with this finding, core protein aa(1–120), which does not contain this sequence but does contain a functional NLS in the N-terminus of core, is found exclusively in the nucleus of hepatoma cells. However, another core fragment, aa(1–140), which includes the NES identified above, is also found exclusively in the nucleus, whereas larger core fragments, consisting of aa(1–160) or aa(1–173), were found in both the nucleus and the cytoplasm of hepatoma cells (Figure 2), as previously reported. We therefore conclude that, in addition to our NES, HCV core region aa(140–160) influences the subcellular distribution of the core protein.

The core protein sequence has been analyzed for determination of its secondary structure. Both aa(109–133) and aa(179–187) form α-helices. These sequences correspond to the two NES motifs identified in HCV core: the classical motif in DIII (Figure 4) and the “atypical” motif in core DII (Figure 5A). The adjacent aa(134–143) sequence was identified as a loop region, whereas aa(144–160) was also found to be rich in α-helices. This fragment may therefore influence the subcellular distribution of HCV core protein. It is thus possible that either there is another NES in this region that enhances the effect of the NES in aa(109–133) identified in this study, or that this sequence affects the subcellular distribution of core due to interaction with membranes and/or lipid droplets.

We investigated the possibility of another NES being present in the aa(109–160) region of core, by constructing three new plasmids encoding protein fragments aa(109–160), aa(134–143), and aa(144–160) of core (Figure 7A). Huh7 cells were transfected with these plasmids and, 40 hours later, we determined the subcellular distribution of the encoded proteins by fluorescence microscopy.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 7. Core protein fragment aa(134–160) contains no other functional NES.

(A) Three plasmids encoding the core fragments aa(109–160), aa(134–143) and aa(144–160) fused to EGFP were constructed as described in the Materials and Methods section. (B) Analysis of the subcellular distribution of the corresponding proteins encoded by the plasmids shown in (A). Cells were transfected with plasmids encoding corresponding EGFP-labeled proteins, grown for 24 h, fixed in 4% PFA and analyzed by fluorescence microcopy. Core protein fragment aa(109–160) was located in the cytoplasm (a), whereas proteins corresponding to core fragments aa(134–143) (b) and aa(144–160) (c) remained in the nucleus. Staining of the nuclei with DAPI. Bar represents 10 µm.

https://doi.org/10.1371/journal.pone.0025854.g007

The aa(109–160) core protein fragment was found predominantly in the cytoplasm, whereas aa(134–143) and aa(144–160) were found solely in the nucleus (Figure 7B). These experiments showed clearly that aa sequences (144–160) and (134–143) did not function as additional NES. Indeed, in the absence of the aa(109–133) sequence identified in this study, these core protein fragments were not exported from the nucleus. We therefore conclude that the cytoplasmic distribution of core requires a functional NES within aa(109–133). However, the presence of this sequence is not sufficient and an additional fragment aa(140–160) is also required to maintain the protein in the cytoplasm.

The hydrophobic C-terminal DII of HCV core, containing this sequence, is required for the binding of core protein to lipid droplets and membranes, and for core protein folding and stability. Indeed, residues of two amphipathic helices, aa(119–136) and aa(148–164), in DII are critical for this association, as shown by extensive mutational analyses. Therefore, the core fragment aa(109–160) is found mostly in the cytoplasm, because, in addition to the NES, it mediates interactions of core with the cytoplasmic lipid droplets and/or membranes via specific sites in the aa(140–160) sequence.

The nuclear export mediated by the core NES is CRM-1-dependent

Proteins containing NES are recognized by the CRM-1 export receptor, a member of the karyopherin superfamily of importin-β nuclear transport receptors. CRM-1 exports cargo proteins containing an NES from the nucleus to the cytoplasm. Leptomycin B is a potent and specific nuclear export inhibitor that alkylates CRM-1, specifically inhibiting the CRM-1-dependent nuclear export of proteins such as HIV Rev.

We therefore investigated whether the nuclear export mediated by the NES in aa(109–133) was sensitive to LMB, and thus dependent on the CRM-1-dependent export pathway. Huh7 cells were transfected with a plasmid encoding this sequence and, 40 hours later, cells were treated with 2 or 10 ng/ml LMB. We found that, at both concentrations used, LMB inhibited the nuclear export mediated by the core aa(109–133) sequence (Figure 8). Nuclear export driven by the HIV Rev protein NES, used as a control, was inhibited by LMB to a similar extent, whereas aa sequences containing the SV40 NLS were always nuclear. The EGFP-tagged control protein was found in both the cytoplasm and nucleus, and this passive distribution was not sensitive to LMB treatment. The aa(109–133) region of the core protein may therefore be considered a functional NES, operating in a CRM-1-dependent manner.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 8. Nuclear export of core NES is mediated by CRM-1.

Huh7 cells were transfected with a plasmid encoding EGFP-tagged NLS SV40 (a, b), EGFP-tagged NLS SV40 and HIV Rev NES (c, d) or the plasmid encoding EGFP-tagged NLS SV40 and putative core NES, aa(109-133) (e, f). Cells transfected with the EGFP-tagged control plasmid pEGFPC1 are shown (g, h). Twenty-two h after transfection, the cells were treated for 4 h with 10 ng/ml LMB (b, d, f, h; +LMB). The cells were subsequently washed, fixed and analyzed by immunofluorescence microscopy. The bar represents 10 µm.

https://doi.org/10.1371/journal.pone.0025854.g008

Following the transfection of Huh7 cells with a plasmid encoding an EGFP-labeled core aa(1–173), the protein was found in both the nucleus and cytoplasm, unless translocation had been blocked by LMB. This treatment increased the proportion of core found in the nucleus although the protein was also present in the cytoplasm in several cells (data not shown). The subcellular distributions of core protein in the presence and absence of LMB were confirmed by confocal microscopy (Figure 9).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 9. Subcellular localization of core in LMB-treated cells.

(A) Production of the EGFP-labeled core protein aa(1–173) in Huh7 cells results in the presence of this protein in both the cytoplasmic (a) and nuclear (b) compartments. Huh7 cells were transfected with a plasmid encoding EGFP–labeled aa(1–173) protein and, 40 h later, cells were treated with 10 ng/ml LMB. The inhibition of nuclear export by LMB induced an increase in the accumulation of HCV core protein in the nucleus (c). (B) Subcellular localization of core protein, with and without LMB treatement, confirmed by confocal microscopy analyses. Huh7 cells were transfected with a plasmid encoding the EGFP-labeled core protein aa(1–173), as described in (A). Slides correspond to the panels shown in A: the cytoplasmic (a) and nuclear (b) distribution of the core protein in non treated cells and its nuclear localization after LMB treatment (c). The bar indicates 20 µm.

https://doi.org/10.1371/journal.pone.0025854.g009

In addition, mutation of the hydrophobic residues L(119), I(123) and L(126) in the NES region of core protein aa(1–173) significantly increased the nuclear localization of this protein, to levels greater than observed for the wild-type protein (Figure 6B).

Thus, the mature core protein produced in transfected hepatoma cells is located in both the cytoplasmic and nuclear compartments. The translocation of the protein from the nucleus follows the CRM-1-dependent pathway, because it is inhibited by LMB. Similarly, mutations of the three hydrophobic amino acids of the NES region spanning aa(109–133) significantly increase the nuclear localization of the core protein.

Detection of core in the nucleus by electron microscopy

In a recent study of HCV infection, we detected HCV particles by immunoelectron microscopy during the entry of the virus into Huh 7.5 cells (P.Maillard, M.Walic et al. PloS One in press). In addition, core and some non structural proteins of other Flaviviruses are trafficked to the nucleus at very early phase of infection. These findings prompted us to investigate whether HCV core protein could also be directed to the nucleus at such early stages of infection. For electron microscopy analysis, virus particles produced in Huh 7.5 cells were concentrated by pelleting through the sucrose cushion and cells were infected with concentrated virus preparations as described in the Materials and Methods section. Virus core protein was stained with the ACAP27 monoclonal antibody, followed by an anti-mouse IgG secondary antibody labeled with colloidal gold. HCV core protein was observed in the cell nucleus 20 minutes after infection, where it was located close to the nuclear membrane or the nuclear pores, as shown by immunostaining with anti-core antibodies (Figure 10 A–C). Taking into account the time after infection, the intranuclear core protein was almost certainly derived from internalized virus particles rather than represented HCV core protein produced de novo in infected cells.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 10. Nuclear localization of core protein in HCV infection, as demonstrated by electron microscopy.

For electron microscopy analysis, virus particles produced in Huh 7.5 cells were concentrated from the cell supernatant by centrifugation through a sucrose cushion for 4 h at 32,000 rpm in an SW 32 Ti rotor. Concentrated virus preparation was incubated with Huh 7.5 cells at 4°C. The cells were then transferred to 37°C and incubated for a further 20 min. Cells were washed, fixed with 4% PFA and stained with monoclonal anti-core antibody ACAP27, followed by secondary, colloidal gold-labeled anti-mouse IgG. Nanogold staining was enhanced by incubation with the HQ silver enhancement kit, and cells were post-fixed by incubation with 1% osmium tetroxide (for details, see the Materials and Methods section). Non infected cells were also incubated with anti-core antibodies. Panels A, B and C, show representative pictures of HCV core localization in HCVcc-infected cells. The presence of one silver-enhanced gold particle is indicated. LD, lipid droplet; M, mitochondrion; NP nuclear pore. (D) Quantitative evaluation of HCV core labeling (gold particles) in the perinuclear vs other areas of HCV-infected and non infected cells, from randomly selected intracellular zones of equivalent area. Thirty cells were considered for each analysis.

https://doi.org/10.1371/journal.pone.0025854.g010

Quantitative analyses performed on HCV-infected and non infected cells after staining with anti-core antibodies confirmed the specificity of our observations and showed that a larger proportion of the gold-labeled HCV core was present in the cell nucleus than elsewhere in the cell at this stage of infection (Figure 10 D). No such staining was observed when unrelated monoclonal antibodies were used. These observations suggest that, at very early stages of HCV infection, at least some of the HCV core protein is directed to the nucleus.

Detection of core in the nucleus by immunofluorescence analysis after LMB treatment

In most studies in the HCVcc in vitro replication system, HCV core protein has not been found in the nucleus. However, if PA28γ, an activator of the proteasome, which is responsible for ubiquitin-independent degradation in the nucleus, was inhibited or absent, very small amounts of core could be detected, by immunofluorescence microscopy, in the nuclei of HCV-infected Huh7 cells.

We investigated whether the NES identified was functional in the JFH1 HCVcc cell culture model. Immunofluorescence microscopy detected HCV core in the cytoplasm of infected cells, but no evidence of a nuclear localization of core was obtained at various times after infection, consistent with previous findings. We therefore used a small interfering RNA to silence the PA28γ gene, to determine whether this treatment increased the amount of core protein in the nucleus, as previously reported.

PA28γ expression was successfully transiently knocked down by a specific siRNA targeting PA28γ, but not by a control, unrelated siRNA (Figure 11A). Despite careful analysis, no core was detected, by immunofluorescence microscopy, in the nucleus of either siRNA-silenced cells or cells transfected with a control RNA (Figure 11B). Instead, HCV core protein was concentrated in the perinuclear area of the cell, in both silenced and untreated cells.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 11. Cytoplasmic localization of core in HCV-infected cells by immunofluorescence without LMB treatment.

(A) Silencing of the proteasome activator PA28γ. Huh 7.5 cells were transfected with an siRNA targeting PA28γ or a control non targeting siRNA 18 h before infection with JFH1. For analysis of the expression of PA28γ by immunofluorescence, cells were stained with rabbit anti-PA28γ antibody, followed by Alexa Fluor 488-conjugated anti-rabbit IgG. Staining for HCV core was carried out 48 h after infection, with the monoclonal anti-core antibody ACAP-27, followed by Alexa Fluor 568-tagged anti-mouse IgG (in red). (a) Non treated HCV (JFH1)-infected Huh 7.5 cells; (b) HCV-infected cells transfected with control, non-targeting siRNA before infection; (c) cells with PA28γ knockdown due to transfection with a specific PA28γ-targeting siRNA. (B) Expression of core in Huh7.5 cells after silencing of the PA28γ proteasome activator. JFH1-infected cells were stained with rabbit anti-lamin B antibody and Alexa Fluor 488-conjugated anti-rabbit IgG as a secondary antibody, to outline the cell nuclei, and with ACAP27 anti-core antibody followed by Alexa Fluor 568-conjugated anti-mouse IgG, for subcellular localization of HCV core. (d) JFH1-infected Huh 7.5 cells without PA28γ silencing, (corresponding to the image shown in (a) panel A); (e) HCV-infected cells transfected with control, non-targeting siRNA before infection (corresponding to the image shown in (b) panel A); and (f) cells with PA28γ knockdown with a PA28γ-specific siRNA before infection with HCV (corresponding to (c) in panel A). Staining of the nuclear membrane with anti-lamin B (green) and with anti-core antibody (red), as described above. No nuclear staining of core was detected, in either siRNA-silenced cells or in cells transfected with a control si-RNA.

https://doi.org/10.1371/journal.pone.0025854.g011

Nevertheless, the treatment of infected cells with LMB, which was maintained in the culture medium for several hours to block the nuclear export of core protein, resulted in the detection of core within the nucleus in several cells, distributed as single points at several sites in the nucleus (Figure 12). We detected HCV core in the nucleus only when LMB was added at early time points (2–4 h) after infection, but not when the drug was added 48 hours after infection.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 12. Intranuclear localization of core in HCV-infected cells treated with LMB.

Huh 7.5 cells were infected with the JFH1 strain of HCVcc and, 2 h after infection, 10 ng/ml LMB was added, with maintenance of this concentration of LMB in the cell culture until 48 h post infection. (A) Detection of intranuclear core in LMB-treated cells. Several cells are shown in which core is present in the nucleus (designated a-i). These cells were analyzed further: 18 cross sections of images were taken to visualize the intranuclear localization of core, at different levels within the nucleus. (B) Cut view of Z stacks from the cells shown in A with the same designation (a–i). Images were captured with an Axioplan 2 microscope with Apotome (Zeiss) and Axiovision 4.6. Orthogonal views were obtained from Z-stack images, with a resolution of 0.33 µm. The nuclear membrane is stained with rabbit anti-lamin B antibody, followed by anti-rabbit IgG tagged with Alexa Fluor 488 (in green), for better visualization of the intranuclear localization of core. HCV core protein is stained with anti-core ACAP27 antibody, followed by Alexa-Fluor 568 labeled anti-mouse IgG (in red). Bar represents 10 µm. Negative controls for these studies (without LMB treatment) are shown in Figure 11, in which no core was detected in the nucleus, whether or not PA28γ was silenced.

https://doi.org/10.1371/journal.pone.0025854.g012

These findings clearly demonstrate that at least some core protein is targeted to the nucleus in the JFH1 HCVcc infection model, at early stages of infection. Core seems to be exported from the nucleus via the CRM-1-dependent pathway, because a nuclear distribution of this protein was observed only in LMB-treated cells.

Leptomycin B affects HCV production

As core was detected in the nucleus of HCV-infected cells treated with LMB, we investigated the possible effects of LMB treatment on virus multiplication. We infected Huh 7.5 cells with HCVcc in the presence or absence of 10 ng/ml LMB or added the drug to the cell supernatant at various time points after infection. Cells were then grown in the medium containing LMB until 32 hours post infection, and intracellular HCV RNA was quantified by RT-qPCR. Incubation with the drug for eight hours was needed for efficient inhibition of infection (Figure 13 A). Infection levels were about 60% lower than control infection levels when the drug was added during the first six hours of infection. By contrast, the percentage inhibition was only 20% if LMB was added eight hours after infection (not shown) and 0% if added twenty-four hours after infection (Figure 11 C). Control experiments excluded a cytotoxic effect of LMB or its solvent on Huh 7.5 cells (Figure 13 B, D).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 13. LMB treatment influences virus production.

(A). Cells were treated with 10 ng/ml LMB for various time periods to determine the incubation time required to affect infection. Cells were infected with HCVcc (JFH1) and the drug was added immediately after infection. Cells were grown in the medium containing LMB, for the time indicated and HCV RNA was then extracted and quantified by RT-qPCR. The values are expressed as a percent of the amount of HCV RNA present in cells grown without LMB. (B). Control experiments carried out to demonstrate that the incubation of cells for 2–8 h with 10 ng/ml LMB (shown in (A) had no toxic effect on cell viability. Cell viability after LMB treatments was determined by counting live and dead cells after trypan blue staining, or by measuring cellular ATP present in culture wells as described in Materials and Methods. Untreated cells and cells treated with 10% DMSO (to induce cell death) were used as negative and positive controls, respectively. The results are expressed as a percent of the value obtained for an untreated control. (C). Treatment with LMB early in infection significantly decreases infection levels. Cells were infected with HCVcc 2 h at 37°C in the presence of 10 ng/ml LMB (T0) or infected with HCVcc and then treated with the drug at the indicated time points after infection (2 h, 4 h, or 6 h) and then incubated for a further 8 h. The treatment of cells with LMB 0–6 h post infection significantly decreases infection levels, whereas the same treatment (for 8 h) applied 24 h after infection has no effect on intracellular HCV RNA levels, as shown by comparison with the untreated control. (D) Control experiments for (C) showing that the application of LMB or its solvent (ethanol) at the same concentrations and for the same time period as used for (C) does not influence cell viability, as demonstrated by comparison with an untreated control. DMSO (at a concentration of 10%) was used as a positive control, to decrease cell viability. Values are expressed as a percent of untreated control.

https://doi.org/10.1371/journal.pone.0025854.g013

Overall, these data showed that early LMB treatment influenced HCV RNA production, suggesting that the early shuttling of core between the cytoplasm and the nucleus might be important for virus multiplication.