HSV-1 Cgal+ Infection Promotes Quaking RNA Binding Protein Production and Induces Nuclear-Cytoplasmic Shuttling of Quaking I-5 Isoform in Human Hepatoma Cells*

Herpesvirus type 1 (HSV-1) based oncolytic vectors arise as a promising therapeutic alternative for neoplastic diseases including hepatocellular carcinoma. However, the mechanisms mediating the host cell response to such treatments are not completely known. It is well established that HSV-1 infection induces functional and structural alterations in the nucleus of the host cell. In the present work, we have used gel-based and shotgun proteomic strategies to elucidate the signaling pathways impaired in the nucleus of human hepatoma cells (Huh7) upon HSV-1 Cgal+ infection. Both approaches allowed the identification of differential proteins suggesting impairment of cell functions involved in many aspects of host-virus interaction such as transcription regulation, mRNA processing, and mRNA splicing. Based on our proteomic data and additional functional studies, cellular protein quaking content (QKI) increases 4 hours postinfection (hpi), when viral immediate-early genes such as ICP4 and ICP27 could be also detected. Depletion of QKI expression by small interfering RNA results in reduction of viral immediate-early protein levels, subsequent decrease in early and late viral protein content, and a reduction in the viral yield indicating that QKI directly interferes with viral replication. In particular, HSV-1 Cgal+ induces a transient increase in quaking I-5 isoform (QKI-5) levels, in parallel with an enhancement of p27Kip1 protein content. Moreover, immunofluorescence microscopy showed an early nuclear redistribution of QKI-5, shuttling from the nucleus to the cytosol and colocalizing with nectin-1 in cell to cell contact regions at 16–24 hpi. This evidence sheds new light on mechanisms mediating hepatoma cell response to HSV-1 vectors highlighting QKI as a central molecular mediator.

Herpesvirus type 1 (HSV-1) based oncolytic vectors arise as a promising therapeutic alternative for neoplastic diseases including hepatocellular carcinoma. However, the mechanisms mediating the host cell response to such treatments are not completely known. It is well established that HSV-1 infection induces functional and structural alterations in the nucleus of the host cell. In the present work, we have used gel-based and shotgun proteomic strategies to elucidate the signaling pathways impaired in the nucleus of human hepatoma cells (Huh7) upon HSV-1 Cgal ؉ infection. Both approaches allowed the identification of differential proteins suggesting impairment of cell functions involved in many aspects of hostvirus interaction such as transcription regulation, mRNA processing, and mRNA splicing. Based on our proteomic data and additional functional studies, cellular protein quaking content (QKI) increases 4 hours postinfection (hpi), when viral immediate-early genes such as ICP4 and ICP27 could be also detected. Depletion of QKI expression by small interfering RNA results in reduction of viral immediate-early protein levels, subsequent decrease in early and late viral protein content, and a reduction in the viral yield indicating that QKI directly interferes with viral replication. In particular, HSV-1 Cgal ؉ induces a transient increase in quaking I-5 isoform (QKI-5) levels, in parallel with an enhancement of p27 Kip1  Herpesvirus type 1 (HSV-1) 1 is a large, double-stranded DNA virus with a genome of 153 kbp, encoding at least 89 proteins. HSV-1 replicates in the nucleus of the host cell and its gene expression follows a temporal pattern including three stages: immediate early (IE), early (E), and late (L) genes (1). The HSV genome is replicated via a rolling circle mechanism. It commences around 3-4 hours postinfection (hpi) reaching maximum efficiency between 8 -16 hpi (2), taking a single round of lytic replication from viral entry to release ϳ16 -20 h in permissive tissue culture cells (3). The process of infection begins when the virions bind heparan sulfate moieties present on host cell surfaces. Within the first 30 min of infection, the initial attachment triggers a cascade of molecular interactions involving multiple viral and host cell proteins and receptors, leading to penetration of the viral nucleocapsid and tegument proteins into the cytoplasm (4). After penetration, viral capsids and associated tegument proteins interact with dynein and use the microtubule network to transit the cytosol to the nuclear envelope, where they dock with nuclear pores and release their uncoated genomes in the nucleoplasm for viral transcription and replication. The temporal program of viral gene expression is highly regulated (5,6). The first genes transcribed during viral infection are the IE genes that serve as transactivators of E genes. E proteins include the enzymes that are required for replication of the viral genome. The temporal program of HSV-1 gene expression ends with the appearance of the L genes, which constitute the structural proteins of the virus. Functional IE proteins are required for the synthesis of all the virally encoded proteins (7). At the same time, the virus acts to inhibit host cell RNA metabolism via a mechanism called virion host shutoff (8), causing a destabilization of mRNAs and cellular polyribosomes (9). The virion host shutoff is supplemented soon after infection by a concomitant inhibition of host cellular protein synthesis and RNA splicing mediated by the immediate-early HSV-1 ICP27 protein (10). These alter cellular transcription and RNA processing factors such as polyadenylation factors and the phosphorylation state of RNA polymerase II (11) to transcribe the viral genome at the expense of its host cell. Although the synthesis of most cellular proteins is progressively inhibited during the course of infection, some specific cellular proteins continue to be efficiently synthesized, even during the late phase (12,13). Recent studies have used different protein separation methods and relative quantification strategies to study the cellular response to different viral infections (14). Comparative proteomics based on a combination of 2-DE with mass spectrometry (MS) has been used to describe protein profiles of HSV-1-infected cells (3,15,16). In targetedproteomic studies, it has been described that HSV-1 VP19C and VP26 proteins associate to ribosomes in HeLa cells (13), and HSV-1 ICP8 and ICP27 interact directly with members of large cellular complexes involved in cellular translation, replication, and chromatin remodelling suggesting new insights into viral replication mechanisms (17,18).
Hepatocellular carcinoma (HCC) is one of the most common malignancies worldwide with a global annual incidence of nearly 1 million cases (19) and an estimated 600,000 deaths per year (20). Although the identification of the main risk factors and the routine screening of the population at risk may lead to the early diagnosis of HCC, the prognosis is poor mainly because of the aggressiveness of the lesions at the time of diagnosis and also to the lack of effective therapies. The development of HSV-1-based oncolytic vectors (21) strengthened by induction of oncoapoptosis (22,23), is a promising research approach to target specifically and efficiently human cancer cells (23,24). Because cancer cells are especially sensitive to apoptosis induced by modified HSV-1 strains (23,25,26), there is an increasing interest in the identification of cellular intermediates orchestrating the host tumoral cell response to HSV-1 strains to promote the development of more efficient and specific vectors. By comparative cytosolic and microsomal proteome analysis we have previously identified erlin-2, Bif-1, RuvB-like 2, and PP2A as novel HSV-1 Cgal ϩ targets involved in the regulation of human hepatoma cell death. The deregulation of the Raf/MAP kinase and FAK/PI3K/Akt survival routes together with the impairment of the mitochondrial apoptotic pathway involving the activation and deactivation of Bcl2 family members and the activation of caspase 3 late in infection (24 hpi) were, at least in part, the effector processes responsible for Huh7 cell death (16). To gain insight on the mechanisms orchestrating the Huh7 cell response to HSV-1 Cgal ϩ infection before activation of apoptotic pathways, we have analyzed the Huh7 nuclear proteome alterations induced by HSV-1 Cgal ϩ at 8 hpi, employing two-dimensional-DIGE and tandem mass tag (TMT) isobaric labeling quantitative mass spectrometry using an LTQ-Orbitrap instrument. We provide 62 differential components of the nuclear proteome of Huh7 human hepatoma cells that may regulate host-virus interaction, cell cycle regulation, and RNA homeostasis upon infection highlighting QKI as a cellular factor necessary for an optimal HSV-1 Cgal ϩ replication.
Virus Production-Vero cells were used for propagation and titration of HSV-1 Cgal ϩ . HSV-1 Cgal ϩ is a replication-competent HSV-1 strain derived from Cgal⌬3, which derives from HSV-1 17synϩ with both copies of ICP4 deleted and the insertion of LacZ gene from the intergenic region IGR54. HSV-1 Cgal ϩ is obtained by repairing both copies of ICP4 (29). Cells were maintained in monolayer with Dulbecco's modified Eagles medium (DMEM) containing 5 to 10% fetal bovine serum and penicillin and streptomycin.
Culture and treatment of human liver cancer cells-Huh7 cells (JCRB Genebank, Japan) were cultured in DMEM supplemented with 10% fetal bovine serum, L-glutamine, and penicillin and streptomycin. Huh7 cells (1 ϫ 10 6 or 5 ϫ 10 6 cells/dish were used for analytical and proteomic experiments respectively) were infected with HSV-1 Cgal ϩ at multiplicity of infection (MOI) of 5 plaque forming units (pfu)/cell. After incubation for 1 h at 37°C, cells were washed and incubated with fresh culture medium under the same conditions during the indicated periods of time. There were no statistically significant differences between control Huh7 cells and Mock-infected Huh7 cells (incubated with a minimum quantity of noninfected Vero cells supernatant). Mock-infected Huh7 cells were used as the reference sample. For QKI silencing, specific small interfering RNAs (siRNA) were used. Transfections were carried out with the DharmaFECT reagent (Dhar-maFECT 4 Transfection Reagent T-2004 -03; Dharmacon Research) according to the instructions of the manufacturer. siRNAs targeting QKI (siQKI: 5Ј-GGCACCUACAGAGAUGCCAACAU-3Ј) and control (siGL: 5Ј-CGUACGCGGAAUACUUCGAUU-3Ј) were from Dharmacon Research. 36 h after transfection, cells were infected as mentioned above with HSV-1 Cgal ϩ at MOI of 5 pfu/cell for the indicated periods of time and then processed for appropriate analysis.
Two-dimensional-DIGE, Imaging, and Mass Spectrometry-The culture medium was removed after 8 hpi and cells were washed three times with ice cold phosphate-buffered saline (PBS). Subcellular fractionation was performed using the Qproteome Cell Compartment Kit from Qiagen. Nuclear fraction was isolated by differential centrifugation according to the manufacturer's recommendations. After ace-tone precipitation, protein samples were solubilized in two-dimensional DIGE sample buffer: 7 M urea, 2 M thiourea, 4% 3-[(3cholamidopropyl) dimethylammonio]-1-propanesulfonate (CHAPS), 30 mM Tris, buffered to pH 8. Protein concentration was determined using the Bradford's assay (Bio-Rad). Then 50 g protein were labeled with 400 pmol of CyDye DIGE Fluor minimal dyes (GE Healthcare) and incubated on ice in the dark for 30 min according to the manufacturer's instructions (Cy3, Cy5 for samples and Cy2 for internal control consisting of a mixture composed by equal amounts of protein from all samples). Paired samples were reverse-labeled to prevent potential dye labeling bias. The reaction was stopped by addition of 1 l of 10 mM lysine and incubated on ice for 10 min. Samples were up-loaded onto IPG strips, 24 cm, pH 3-11NL (GE Healthcare), and subjected to isoelectric focusing (IEF) in IPGphor™ IEF System (GE Healthcare) according to the manufacturer's recommendations. Upon IEF, strips were incubated in equilibration buffer (50 mM Tris-HCl, pH 8.8, 6 M urea, 30% glycerol, 2% SDS, a trace of bromphenol blue), containing 0.5% dithiotreitol for 15 min and thereafter in the same buffer with 4.5% iodoacetamide for 15 min. For the second dimension, strips were loaded on top of 12.5% polyacrylamide gels (21 ϫ 24 cm) and run (1W/gel) for 12-14 h until the bromphenol blue dye reached the gel bottom-end. Subsequently, two-dimensional gels were scanned using a Typhoon TM Trio Imager (GE Healthcare) at 100-m resolution with ex/em of 488/520, 532/580, and 633/670 nm for Cy2, Cy3, and Cy5 respectively. The photomultiplier tube was set to ensure that the maximum pixel intensity was between 90,000 and 99,000 pixels. Image analysis was performed using DeCyder 6.5 software (GE Healthcare) as described in the user's manual. Three independent experiments were performed. Briefly, the differential in-gel analysis module was used for spot detection, spot volume quantification and volume ratio normalization of different samples in the same gel. Then the biological variation analysis module was used to match protein spots among different gels and to identify protein spots that exhibit significant differences. Manual editing was performed in the biological variation analysis module to ensure that spots were correctly matched between different gels, and to get rid of streaks and speckles. Differential expressed spots were considered for MS analysis when the fold change was larger than 1.2 and the p value after t test was below 0.05. Preparative gels were run with 350 g of protein following the same procedure described above. Proteins were visualized by staining with SYPRO Ruby Protein Gel Stain (Bio-Rad) and images were acquired with a Typhoon TM Trio Imager using ex/em of 532/560 nm. Spots differentially represented were excised manually and gel specimens were processed with a MassPrep station (Waters) as described elsewhere (30). In-gel tryptic digestion was performed with 12.5 ng/l trypsin in 50 mM ammonium bicarbonate for 12 h at 37°C. The resulting peptides were extracted with 5% formic acid (FA), 50% acetonitrile (ACN). Samples were then concentrated in a speed-vac before MS analysis. NanoLC-ESI-MS/MS analysis was performed as described previously (16). Data processing was performed with MassLynx 4.0. Database searching was done with three independent search engines: ProteinLynx Global Server 2.3 (Waters), Phenyx 2.6 (GeneBio), and Mascot Server 2.2 (Matrix Science) against UniprotKB/Swiss-Prot Release 51.6 with 257964 entries (human taxonomy: 15720 entries). Parameters used in Mascot searches were: enzyme, trypsin; variable modifications, carbamidomethylation of cysteine, and oxidation of methionine; maximum missed cleavages, 1; peptide mass tolerance settings/windows was 50 ppm; product mass tolerance, 0.1 Da. Probability p of random matches was set to the default value of 0.05.
Reduction, Alkylation, Digestion, and Labeling with 6-Plex TMT-Thirty micrograms of nuclear protein extracts from mock and HSV-1 Cgal ϩ -infected Huh7 cells were dissolved in 300 l triethylammonium hydrogen carbonate buffer (TEAB) 0.1 M adjusted to pH 8 (Fluka). SDS (Fluka) 1% in water (w:w) was added to obtain a final 0.1% SDS in all samples. Two microliter of 50 mM tris-(2-carboxyethyl)-phosphine hydrochloride (TCEP-HCl) (Sigma) was added. The reaction was performed for 1 h at 60°C. One l of iodoacetamide (Sigma) at 400 mM was then added, and the mixtures were reacted for 30 min in the dark at room temperature. An amount of 10 l of freshly prepared trypsin at 0.2 g/l concentration in TEAB (0.1 M) was added. The digestion was performed overnight at 37°C. TMT labeling was achieved at room temperature, after addition of 6-plex TMT reagents (0.83 mg) in 40.3 l ACN (Sigma-Aldrich). Nuclear extracts from Mock-infected cells (triplicates) were labeled by TMT with reporters at m/z ϭ 126.1, 127.1, and 128.1 and nuclear extracts from HSV-1 Cgal ϩ -Huh7 infected cells by TMT with reporters at m/z ϭ 129.1, 130.1, and 131.1. After 1 h reaction, 8 l of 5% hydroxylamine (Sigma) (w/V) was added in each tube and mixed for 15 min. The six samples were pooled in a new tube and dried in a speed-vac for storage at Ϫ20°C. The pooled TMT-labeled sample was ressuspended in 1 ml water:ACN (95:5) 0.1% trifluoroacetic acid (Fluka, Buchs, Switzerland) and cleaned using a OASIS ® HLB 1 cc (30 mg) cartridges (Waters, Milford, MA) connected to a vacuum system according to the manufacturer's instructions with slight modifications.

Off-gel Separation of Pooled TMT-labeled Sample, LC-ESI LTQ-Orbitrap MS/MS Analysis, and Database
Search-After drying, the pooled TMT-labeled sample was dissolved in 1616.4 l H 2 O with 172.8 l glycerol 50% (Agilent Technologies, Santa Clara, CA) and 10.8 l of carrier ampholytes IPG buffer pH 3-10 (GE Healthcare). The peptide isoelectric focusing separation was carried out using the 3100 OFFGEL Fractionator (Agilent Technologies). The IPG strip (pH 3-10, 13 cm) (GE Healthcare) was assembled on the off-gel tray and rehydrated for 30 min with a solution of 89.8% H 2 O, 9.6% glycerol 50%, and 0.6% of carrier ampholytes. The sample was loaded in the 12 off-gel wells. The isoelectric focusing was achieved overnight with a limiting current of 50 A and a limit of 20 kV⅐h before holding the voltage to 500 V. The fractions were collected and their pH was measured (744 pH Meter and Biotrode from Metrohm). The fractions were dried, cleaned with C18 ultramicrospin columns (Harvard Apparatus, Holiston, MA) according to the manufacturer's instructions with slight modifications and dried again. NanoLC-ESI-MS/MS was performed on a LTQ Orbitrap XL (Thermo Electron) equipped with a NanoAcquity system (Waters). Peptides were trapped on a homemade 5 m 200 Å Magic C18 AQ (Michrom) 0.1 mm ϫ 20 mm precolumn and separated on a homemade 5 m 100 Å Magic C18 AQ (Michrom) 0.075 mm ϫ 150 mm column with a gravity-pulled emitter. The nanoLC was run for 85 min using a gradient of water/FA 99.9%/ 0.1% and ACN/FA 99.9%/0.1% as already described (31) at a flow rate of 220 nL⅐min Ϫ1 . For MS survey scans, the Orbitrap resolution was set to 60,000 and the ion population was 5 ϫ 10 5 with an m/z window from 400 to 2000. A maximum of three precursors was selected for both collision induced-dissociation (CID) in the LTQ and higher-energy C-trap dissociation (HCD) with analysis in the Orbitrap. For MS/MS in the LTQ, the ion population was 1 ϫ 10 4 (isolation width of 2 m/z), whereas for MS/MS detection in the Orbitrap, it was 2 ϫ 10 5 (isolation width of 4 m/z), with resolution of 7500, first mass at m/z ϭ 100, and maximum injection time of 750 ms. The normalized collision energies were 35% for CID and 50% for HCD. Dynamic time exclusion was 60 s. Peak lists were generated using an in-house-written Perl script. CID and HCD spectra were merged using a custom-made program (31). The combined mgf file from the nanoLC-ESI-MS/MS analysis of the 12 off-gel fractions was searched against UniProt-Swiss-Prot database (57.4 of 16-Jun-2009, 565634 entries) using Phenyx. Homo sapiens taxonomy (40335 entries) was specified. Variable amino acid modifications were oxidized methionine. 6-plex TMTlabeled peptide amino terminus and lysine (ϩ229.1629 Da) and car-bamidomethylation of cysteines were set as fixed modifications. Trypsin was selected as the enzyme, with one potential missed cleavage. Only one search round was used with selection of "turbo" scoring. The peptide p value was 1 ϫ 10 Ϫ3 . The protein and peptide scores were set at 5.0, providing a false peptide discovery rate of 2%. The parent ion tolerance was 20 ppm. For all analyses, only proteins matching two different peptide sequences were selected. Reporterion intensities were extracted from peak lists using the dedicated Phenyx export. The reporter-ion intensities were corrected according to the isotopic purities of the reporter-ions provided by the manufacturer.
Bioinformatic Analysis-TMT data processing was performed using Linear models for microarrays data (32) to find out the peptides that showed significant differential expression between both experimental conditions. For a given protein a score, p, was calculated with the p values of its corresponding set of identified peptides S: To discard inconsistencies, we define a coherence score, s, for the set of peptides quantified for a given protein as the average expression similarity of all the possible peptide pairs. The similarity between two peptides pi and pj is computed as the logarithm of the p value obtained using the Spearman rank correlation p cor : The two resulting scores (p and s) are combined in a unique metric to obtain the ranked list of interesting proteins. The selection of differentially expressed proteins is based on the distribution of this score. All the analysis were performed using R (www.r-project.org) and Bioconductor (www.bioconductor.org) (33). The biological knowledge extraction was carried out through the use of Ingenuity Pathway Analysis (Ingenuity Systems, www.ingenuity.com), which database includes manually curated and fully traceable data derived from literature sources.
Immunoblotting Analysis-Equal amounts of protein (15 g) were resolved in 12.5% SDS-PAGE gels. Proteins were electrophoretically transferred onto nitrocellulose membranes for 45 min at 120 V. Membranes were probed with primary antibodies at 1:1000 dilution in 5% bovine serum albumin (BSA) or 5% nonfat milk, depending on the primary antibody used. After incubation with the appropriate horseradish peroxidase-conjugated secondary antibody (1:5000), the immunoreactivity was visualized by enhanced chemiluminiscence (Perkin Elmer). Equal loading of the gels was assessed by Ponceau staining and hybridization with a ␤-actin specific antibody (Abcam).
Plaque Assays-To measure virus production after inhibition of QKI expression, Huh7 cells were treated with control (GL) or QKI siRNAs for 36 h. Cell monolayers were then infected with HSV-1 Cgal ϩ (MOI ϭ 1pfu/cell). Cells and cell culture supernatants were collected at 16 and 24 hpi. Huh7 cells in six-well plates were inoculated with serial 10-fold dilutions of progeny virus. After a 1 h adsorption period at 37°C, the inoculum was removed and then overlaid with DMEM supplemented with 1% methylcellulose. After two washes with PBS, plaques were counted and visualized at 72 hpi by fixing and staining the cells with 0.5% crystal violet in water.
Immunofluorescence Microscopy-HSV-1 Cgal ϩ -infected or Mock-infected Huh7 cells were grown in chamber slides for the indicated periods of time, washed twice with PBS, fixed for 10 min in 3.7-4% formaldehyde, and made permeable by Triton-X100 5% for 30 min. After blocking in 2% BSA in TBS-0.05% Tween 20 (TBS-T) for 1 h at room temperature, cells were incubated with primary antibodies (rabbit anti-QKI5, dilution 1:1000 or mouse anti-nectin-1, dilution 1:100) overnight at 4°C. After washing in TBS-T, secondary antibodies, Alexa Fluor 488 goat anti-rabbit (A11008, Molecular Probes, Eugene, OR; dilution 1:200) or Alexa Fluor 568 goat anti-mouse (A11004, Molecular Probes; dilution 1:200) were applied for 60 min at room temperature. Nuclei were counterstained with 4Ј,6-diamidino-2-phenylindole dihydrochloride (40009, Biotium). Pictures were taken on a Nikon Eclipse E800 microscope equipped with epifluorescence optics. Three independent experiments were performed for the indicated periods of time. The immunofluorescence figures are representative of the overall effects observed under each set of conditions.

Two-dimensional-DIGE Analysis of Virally Infected Nuclei-
The capacity of infection and replication of HSV-1 Cgal ϩ in Huh7 cells were previously tested both in vitro and HCC xenograft murine models (15,16). Nuclear proteome was examined at 8 hpi, when most that 90% of the cells were infected showing minimal morphologic changes (16). Detection of the endoplasmic reticulum marker (KDEL) preferentially in the cytosolic fraction together with detection of histone H4, hepatocyte nuclear factor 1 ␣ (HNF-1␣) in addition to major histone deacetylase 1 (HDAC-1) location in the nuclear extracts indicated the efficiency of the enrichment procedure ( Fig. 1 supporting information). Protein mixtures from nuclear fractions extracted from Mock and HSV-1 Cgal ϩ -Huh7 infected cells (8 hpi) were compared by DIGE analysis, alternating Cy3 and Cy5 labeling to compensate for any differential observation resulting from the chemistry of the fluorescent dyes. Decyder analysis allowed detection and quantification of 2539 spots in the nuclear fraction (Fig. 1). Differences were accepted with t test Ͻ0.05 and a fold-change Ͼ1.2 ( Fig. 2 Supporting information). Differential spots were localized on preparative gels and were identified by nanoLC-ESI-MS/MS after in gel trypsin digestion (see supplemental material for technical details). From the resulting tryptic digests, 32 protein species (16 were up-regulated and 16 down-regulated in nuclear fractions of HSV-1 Cgal ϩ -Huh7 infected cells) corresponding to 24 proteins were unambiguously identified ( Table  I). Some of the differentially expressed proteins are involved in pre-mRNA splicing. Protein species from Non-POU domaincontaining octamer-binding protein, TAR DNA-binding protein 43, hnRNPC, hnRNPA2/B1, hnRNPH3, peptidyl-prolyl cis-trans isomerase E, and Pre-mRNA-splicing factor SPF27 were down-regulated in nuclear fraction of HSV-1 Cgal ϩinfected Huh7 cells. In contrast, protein species from hnRNPQ and splicing factor arginine/serine-rich 1 were up-regulated in the nucleus of Huh7 infected cells. The expression levels of Non-POU domain-containing octamer-binding protein, TAR DNA-binding protein 43, Pre-mRNA-splicing factor SPF27, and splicing factor arginine/serine-rich 1 genes were measured by RT-PCR in HSV-1 Cgal ϩ -Huh7 infected cells. mRNA levels of these splicing factors were unchanged at 8 hpi (data not shown) indicating that changes detected at protein level may be because of post-translational events. As often appears on two-dimensional gel-based studies, several different spots were identified as products of the same gene. In particular Lamin A-C was identified from six different spots with different pI, where five of them were down-regulated in Huh7infected cells. These observations suggest that different isoforms or post-translational modifications of this protein might play different roles in the course of HSV-1 Cgal ϩ infection as has been previously demonstrated (34).
TMT-Analysis of Virus-Infected Cell Nuclei-A total of 394 proteins were identified and quantified with a minimum of two peptides by TMT analysis comparing virally infected host cell nuclei to uninfected nuclei (see supplemental material). Subcellular location analysis performed with Ingenuity Pathway Analysis software and Human Protein Reference Database (www.hprd.org) revealed that 72.7% of quantified proteins were nuclear proteins suggesting the efficiency of the enrichment procedure (Fig. 3 supplemental material). From the ranked list generated by combination of two scores (p and s), only the upper 20 percentile was considered for further analysis. Moreover, a biological criteria based on subcellular location was used to filter nonnuclear proteins. Finally, 38 pro- teins showed significant changes (32 were up-regulated and six down-regulated in nuclear fractions of HSV-1 Cgal ϩ -Huh7 infected cells) (Table II). The estimation of the fold change for each protein was obtained as follows: the mean value of each peptide quantification is calculated for Mock and HSV-1Cgal ϩ conditions; then the ratios for each reporter are calculated and the log2 of the median of these ratios is considered as the protein fold change. The peptide labeling reproducibility between the different set of biological replicates and the coherence among unique peptides corresponding to the differentially expressed proteins are shown in Fig. 2. One third of the up-regulated proteins correspond to components of 40S and 60S ribosomal subunits and other proteins involved in different steps of translation. Annexin A2 was the unique deregulated protein identified by DIGE and TMT analysis. However, the differential proteins provided by the two strategies showed functional complementarity as both revealed alterations in similar biological processes including host-virus interaction, transcription regulation, mRNA processing, and mRNA splicing allowing a more comprehensive analysis of HSV-1 Cgal ϩ -infected cell nuclei.  Increase of Quaking RNA Binding Protein During HSV-1 Cgal ϩ Infection-QKI was one of the most interesting nuclear alterations induced by HSV-1 Cgal ϩ infection as revealed by TMT analysis. QKI is involved in the regulation of mRNA stability, nuclear retention, RNA transport, and translational modulation through interaction with the Quaking Response Element (35). Although conventional data processing suggested QKI-7 as the modified isoform (Table II), the three corresponding peptides used for protein identification and quantification (peptides from residue 26 to 44; 85-102; 177- 192) belong to the STAR (Signal Transduction Activator of RNA metabolism) domain of QKI, which is conserved between the different alternatively splice variants of qkI gene. To validate the potential up-regulation of QKI protein content in infected Huh7 cells, we used an anti-QKI-Pan antibody that recognizes the three major QKI isoforms (QKI-5, QKI-6, and QKI-7). As shown in Fig. 3, Mock-infected Huh7 cells expresses low levels of QKI that peak at 4 hpi leading to a time dependent decrease during the infection. Interestingly, viral immediate-early genes such as ICP4 and ICP27 are translated when QKI protein peaks 4hpi (Fig. 3).

Inhibition of QKI Expression Results in Reduction of Viral Protein Content and Viral Yield in Huh7
Cells-Based on the correlation between the up-regulation of QKI and the expression of viral immediate-early proteins in Huh7-infected cells, we hypothesize that QKI protein could be one of the cellular proteins that plays a direct role in viral replication. To test this possibility, we carried out experiments using siRNA strategy to knockdown QKI expression. A reduction of 50% in QKI protein levels was observed 36 h post-transfection in siQKI cells with respect to control siGL cells (supplemental material; Fig. 4) without observing any alterations in Huh7 cell morphology and cell viability until 72 h post-transfection. To determine whether QKI depletion can affect the efficiency of viral replication, Huh7 cells were submitted to HSV-1 Cgal ϩ infection (MOI ϭ 5 pfu/cell) 36 h post-transfection with the siRNAs. First, we monitored the capacity of infection of HSV-1 Cgal ϩ in siGL and siQKI Huh7 cells (36 h post-transfection) by X-Gal staining at 4, and 8 hpi as described previously (16). No differences in ␤-galactosidase transgene expression were observed in siGL and siQKI infected-Huh7 cells (supplemental material; Fig. 5). These data point out that HSV-1 Cgal ϩ infectivity is not hampered in cells where QKI expression is inhibited also suggesting that QKI is not essential for viral entry. The effect on viral replication was determined 4, 8, and 16 hpi by western-blot using specific antibodies against immediate-early viral proteins. As shown in Fig. 4, QKI protein expression was transiently increased 4 -8hpi in siGL-infected cells indicating that transfection with nonspecific siRNA has no significant consequence on QKI expression pattern upon infection. We observed that the levels of ICP27 and ICP4 proteins were decreased in cells transfected with QKI siRNA at 4, and 8 hpi with respect to siGL Huh7-infected cells. Although ICP4 levels were still failing in siQKI Huh7-infected cells, ICP27 levels were not significantly modified between both conditions at 16 hpi (Fig. 4A). Moreover ICP0 protein was also down-regulated at 8, and 16 hpi in siQKI Huh7-infected cells (Fig. 4A). These results indicate that immediate-early viral functions are partially delayed in cells where QKI expression is down-regulated. To determine whether the observed QKI dependent alteration was specific of immediate-early viral factors, the amounts of some early and late viral proteins were tracked by Western blot analysis. The expression of polymerase accessory viral protein UL42 (early gene) and the tegu-ment protein US11 (late gene) was also hampered in siQKI Huh7-infected cells respect to control siGL Huh7-infected cells at 8 hpi (Fig. 4B). Subsequent experiments were performed to analyze the effect of repression of QKI expression on the amounts of progeny virus produced in Huh7 cells for 16 and 24 h. Although viral yield significantly increased from 16 hpi to 24 hpi in siGL-infected cells, there was a block in the production of viral particles in siQKI-infected cells during the infection (Fig. 4C). Although no differences were detected between siGL and siQKI cells at 16 hpi, a 40% reduction in virus yield was detected in siQKI cells infected with HSV-1 Cgal ϩ (MOI ϭ 1 pfu/ml) compared with siGLinfected cells at 24 hpi (Fig. 4C). All this data demonstrate the pivotal role of QKI in the regulation of the expression of essential proteins mediating the early phases of HSV-1 Cgal ϩ replication.
HSV-1 Cgal ϩ Increases QKI-5 Isoform Levels in Human Hepatoma Cells-As TMT-based analysis detected an increase in a QKI isoform in the nucleus of Huh7-infected cells and QKI-5 isoform (QKI-5) presents a nuclear localization signal (36), subsequent experiments were performed to check the steady-state protein levels of QKI-5 in Huh7-infected cells using a specific antibody against a C-terminal peptide from QKI-5. The specificity of the QKI-5 antibody was assessed by Western blot and immunofluorescence analysis (supplemental material; Fig. 6). QKI-5 protein content was increased until 4 hpi maintaining its levels until 8 hpi to decrease late in infection in infected-human hepatoma cells (Fig. 5). At a functional level, QKI isoforms bind and stabilizes the mRNA encoding the cyclin-dependent kinase (CDK)-inhibitor p27 Kip1 , leading to an increased protein accumulation (37,38). To evaluate whether QKI protein was maintaining its wellknown function during the infection, additional experiments were performed to investigate one of its targets, p27 Kip1 . As shown in Fig. 5, p27 Kip1 displayed a transient increase during viral infection, in parallel with QKI-5 expression, suggesting that its mRNA may be stabilized by QKI-5 during the infection. Another negative regulator of the G1-S phase transition, that is functionally and structurally related to p27 Kip1 is p57 Kip2 (39). As shown in Fig. 5, p57 Kip2 showed a transient increase upon infection maintaining high levels onwards 8 hpi. The transient increase in p27 Kip1 and p57 Kip2 levels (Fig. 5) points out a partial blockage of the cell-cycle in HSV-1 Cgal ϩ -infected Huh7 cells at 4 -8 hpi.
QKI-5 Isoform Localizes in Different Intracellular Compartments During HSV-1 Cgal ϩ Infection-As QKI-5 has been described as a nucleo-cytoplasmic shuttle protein (36,40), additional experiments were carried out to study the subcel-lular localization of QKI-5 during HSV-1 Cgal ϩ infection using the same antibody mentioned above. Immunofluorescent staining of Mock-infected Huh7 cells revealed that QKI-5 showed a diffuse staining throughout the nucleus (excluded from nucleoli) forming 2-3 aggregate structures at perinuclear level in Mock-infected Huh7 cells (Fig. 6, upper panel) likely suggesting association with nuclear bodies, which identification still remain elusive. Between 4 and 8hpi, while still nuclear, the aggregate structures were dispersed and redistributed in more punctuate dots throughout the nucleus (Fig. 6, middle panel). By 16 -24 hpi, QKI-5 is localized in the cytoplasm forming a granular-type pattern, with low level of nuclear fluorescence (Fig. 6, lower panel), indicating that QKI-5 shuttles from the nucleus to the cytosol in Huh7 cells in response to HSV-1 Cgal ϩ infection. Strong linear staining were also observed in several cell-to-cell contact regions at 16 -24hpi (Fig. 6, lower panel). As nectin-1 is a specific cellular receptor for HSV-1 localized at cell contact areas in infected cells (41), colocalization experiments were undertaken. Nectin-1 distribution revealed a dotted pattern at the cell surface of infected-Huh7 cells 16 and 24 hpi (Fig. 7). Interestingly, nectin-1 staining co-localized with QKI-5 in cell-to-cell contact regions of some infected cells (Fig. 7), suggesting that QKI-5 may participate in the cell fusion phenomena induced by HSV-1 Cgal ϩ infection.

DISCUSSION
HSV-1 is one of the most promising viral platforms for the development on oncolytic vectors that can target, multiply in, and eradicate hepatoma cells (42), but the intermediates mediating the tumoral cell response must be elucidated to promote the development of more efficient and selective HSV-1based vectors. By comparative cytosolic and microsomal proteome analysis we have previously identified deregulation of central intermediates targeted by HSV-1 Cgal ϩ resulting in the impairment of apoptosis and cell survival pathways in human hepatoma cells (16). In this report, we have applied two different labeling proteomic approaches, two-dimensional-DIGE and TMT to analyze changes in the nuclear proteome composition that arise in infected-human hepatoma cells before caspase activation (16). The application of two different proteomic strategies was particularly useful in obtaining complementary information; accordingly, only one protein was found to be common to both experimental approaches. The complementary data result from the differences in sample preparation and analytical methods used by the two proteomic workflows. two-dimensional-DIGE is a gel-based method that labels samples at the protein level (43) and TMT is a liquid-based method that labels samples at the peptide level (31). Although DIGE separates only soluble proteins included in a pH range of 3 to 11 and a determined molecular weight, TMT can identify proteins outside these ranges. Moreover, DIGE is able to detect differential expression of posttranslationally modified proteins as well as different isoforms of proteins by resolving spots at different pI and molecular weight. However, these isoforms may not be distinguished with TMT approach, because labeling is completed at the peptide level and most peptide sequences are identical among a group of isoforms. In addition, proteins with extreme isoelectric points, large molecular weights, low solubility (hydrophobic proteins) and low copy numbers are poorly represented in Two-dimensional-DIGE experiments. On the other hand, TMT technique could lead to peptide loss because of TCEP used for protein reduction (44,45), interaction between peptides and gel matrix during the off-gel electrophoresis, and peptide tagging (46). Differential proteins revealed by both methods appear to be complementary in mediating parallel biological functions in HSV-1 Cgal ϩ -infected human hepatoma cells (Fig. 8), such as host-virus interaction, cellcycle regulation, transcription regulation, mRNA processing, and mRNA splicing. Despite the well-known protein synthesis shutoff inherent to HSV-1 infection, most of the differentially expressed proteins reported in this study were up-regulated. As it has been postulated, those proteins that continue to be efficiently synthesized upon HSV-1 infection could play a major role in determining the outcome of infection (12).
QKI is involved in the regulation of mRNA stability, nuclear retention, RNA transport, and translational modulation through interaction with Quaking Response Element located in the UTR of target mRNAs (47). QKI isoforms regulate aspects of RNA metabolism in several cellular processes including myelination, cell fate determination, embryogenesis, apoptosis, and protein translation (48). The essential function of QKI in cell biology is further suggested by its widespread expression in different cell types (36,38,48) and by the finding that total ablation of QKI in mice results in early embryonic lethality (49). Furthermore, evidence of alterations of QKI expression in cancer tissues (38,48) together with the critical regulator of p27 Kip1 , suggests that abnormal levels of this protein may contribute to deregulation of cell growth. We have shown that QKI protein and its target cyclin-dependent kinase inhibitor p27 Kip1 (CDKN1B) (37,38)  their levels in a parallel way at early times of HSV-1 Cgal ϩ infection, likely blocking the G1-S phase transition of the cell-cycle. These observations are in agreement with the proliferation blockage and tumorigenesis ability inhibition that occur when QKI is overexpressed (38,50). However, the late fall in p27 Kip1 protein levels in infected-human hepatoma cells might result from the HSV-1 virion host shutoff (8). In contrast, cyclin-dependent kinase inhibitor p57 Kip2 (CDKN1C) maintains its levels in infected-human hepatoma cells late in infection. p57 Kip2 promotes cell death via the mitochondrial apoptotic pathway with caspase 3 activation in cancer cells (51) suggesting that p57 Kip2 may mediate the multi-factorial impairment of proapoptotic factors previously described in HSV-1 Cgal ϩ -infected human hepatoma cells (16). The QKI up-regulation occurs during the immediate-early stage of HSV-1 infection in human hepatoma cells. Depletion of QKI did not compromised the entry and replication of HSV-1 virions, as evidenced by the expression of the reporter gene ␤-galactosidase. However, analysis of steady-state levels of viral proteins representing each of the three major kinetic classes (IE, E, and L genes) indicates that QKI knock-down induces a clear delay in viral protein synthesis. Besides the existing cooperativity between IE proteins regarding intracellular localization (52) and the interdependence for assembly into viral particles (53), IE proteins are continuously required for the synthesis of all the virally encoded proteins. ICP0 acts as a transactivator of all classes of HSV-1 genes (54,55,56,57). ICP4 is absolutely required for progression beyond the immediate-early phase of gene expression because of its role as a transcriptional activator of early and late (58,59,60). The IE protein ICP27 is also required for early DNA replication genes such as UL42 (61,62) and also for transcription of viral late genes (61,63). In view of our results, QKI is part of the cellular machinery necessary for an optimal HSV-1 Cgal ϩ replication as is also the case for human immunodeficiency virus (HIV) (64).
The mammalian qkI gene undergoes a complex pattern of alternative splicing and generates at least three major transcripts of 5, 6, and 7 kb, corresponding to QKI-5, QKI-6, and QKI-7 isoforms (48). QKI-6 and QKI-7 are mainly cytoplasmic isoforms whereas the unique sequence located at the C terminus of the QKI-5 isoform possesses a noncanonical nuclear localization signal (48). Moreover, the RNA binding domain is important for the localization of QKI-5 in the nucleoplasm (36). By immunofluorescence microscopy analysis, QKI-5 showed a diffuse staining throughout the nucleus as has been previously reported in neuronal and epithelial HeLa cells (36,48). In contrast to HeLa cells, QKI-5 forms 2-3 aggregate structures at perinuclear level in human hepatoma cells. Furthermore, we showed that HSV-1 Cgal ϩ infection induces an early QKI-5 nuclear spreading forming dotted structures that may be promyelocytic leukemia protein bodies or Virus-Induced Chaperone-Enriched domains, previously related with HSV-1 replication (65,66). After 16 hpi, HSV-1 Cgal ϩ induces QKI-5 nuclear-cytoplasmic shuttling, an event previously shown during normal embryonic neuronal cell fate decision (40). This nuclear-cytoplasmic shuttling may be mediated by posttranslational modifications, already described for this protein, such as phosphorylation (67) or arginine methylation (68) as previously confirmed for Sam68, member of the STAR family that shares the heterogeneous ribonucleoprotein K homologue domain (KH domain) with QKI-5 (68). The mechanisms promoting QKI-5 shuttling and the biological implications of its cytoplasmic localization remain unclear. Concerning the first issue, active transcription mediated by RNA polymerase II is necessary for nuclear retention of QKI-5 (36,69,70). According to this mechanism, QKI-5 shuttling may be a consequence of changes in the activation state of RNA polymerase II that occur during HSV-1 infection (71). The function of cytosolic QKI-5 might be related with the transport and processing of specific cellular RNAs (35) in a process integrating other QKI isoforms (36). Late in infection, QKI-5 is channeled to the cell surface where it colocalizes with nectin-1 in cellto-cell contact areas in some infected cells. Nectin-1 displayed a dotted distribution at the cell surface at 16 and 24 hpi in human hepatoma cells, as previously reported in HSV-1infected melanoma cells (41). Nectin-1 belongs to the Ca 2ϩindependent cell adhesion proteins of the immunoglobulin superfamily (72) found at adherens junctions (73). Although nectin-1 is considered to be the most important receptor for HSV-1 entry (74), it also serves as cell-cell spread mediator of HSV-1 (75). Cell-cell spread occurs when cell-associated virus propagates to adjacent cells through areas of cell contact by molecular events that are not completely known (76). Our co-localization studies suggest that QKI-5 may be involved in the nectin-1-dependent events occurring at the plasma membrane during cell-cell spread.In summary, we provide a broad analysis of the nuclear proteome of human hepatoma Huh7 cells upon HSV-1 Cgal ϩ infection using two distinct proteomic strategies. Complementary data sets were generated and integrated on a single functional interpretation to uncover novel response mechanisms triggered by HSV-1 Cgal ϩ vector in human hepatoma cells. We have demonstrated that early QKI protein up-regulation is necessary for optimal viral protein translation in human hepatoma cells. Moreover, the time-dependent re-distribution of specific QKI-5 isoform in Huh7infected cells may take part in the cell fusion events observed late in infection.