Serine/Arginine-rich Splicing Factor 2 Modulates Herpes Simplex Virus Type 1 Replication via Regulating Viral Gene Transcriptional Activity and Pre-mRNA Splicing*

Once it enters the host cell, herpes simplex virus type 1 (HSV-1) recruits a series of host cell factors to facilitate its life cycle. Here, we demonstrate that serine/arginine-rich splicing factor 2 (SRSF2), which is an important component of the splicing speckle, mediates HSV-1 replication by regulating viral gene expression at the transcriptional and posttranscriptional levels. Our results indicate that SRSF2 functions as a transcriptional activator by directly binding to infected cell polypeptide 0 (ICP0), infected cell polypeptide 27 (ICP27), and thymidine kinase promoters. Moreover, SRSF2 participates in ICP0 pre-mRNA splicing by recognizing binding sites in ICP0 exon 3. These findings provide insight into the functions of SRSF2 in HSV-1 replication and gene expression.

Herpes simplex virus type 1 (HSV-1) is a human ␣ herpesvirus that is associated with orofacial and genital herpes infections and is related to herpes encephalitis (1). The HSV-1 genome encodes Ͼ80 genes that are transcribed by RNA polymerase II (RNAP II) 3 (2). Of these viral genes, infected cell polypeptide 0 (ICP0) plays vital roles in facilitating the HSV-1 life cycle. In HSV-1-infected cells, ICP0 prevents the antiviral response triggered by dsDNA by degrading IFI16 (3), PML, and SP100 (4). Additionally, ICP0 inhibits host IRF3 nuclear signaling to prevent the interferon production-mediated antiviral response of the infected cells (5). ICP0 enables efficient viral replication by redistributing host CCND3 to ND10 bodies, which function as precursors of replication compartments (6).
ICP0 also participates in HSV-1 reactivation. In the HSV-1 genome, ICP0 dissociates HDAC1 and HDAC2 from the HDAC-RCOR1-REST-KDM1A complex to make the viral DNA accessible for transcription factor binding (7). After infection, HSV-1 uses a series of host cell factors to facilitate its life cycle. Host cell factor 1, which is a cellular transcriptional coactivator, plays a central role in initiating the expression of viral immediate early (IE) genes by interacting with numerous host cell transcription factors, such as virion protein 16 (8) and early growth response protein 1, a protein that mediates IE gene expression by binding to the key regulatory elements near the HSV-1 IE genes (9). Moreover, the DNA methyltransferase DNMT3A has been reported to promote HSV-1 replication by associating with the viral capsid protein VP26 (10).
Serine/arginine-rich splicing factor 2 (SRSF2 or SC35), which is a specific well known serine/arginine-rich (SR) protein family member, is well known as a mediator of genome stability, pre-mRNA splicing, mRNA nuclear export, and translational control (11)(12)(13)(14)(15). SRSF2 contains an RNA recognition motif for RNA binding and a domain rich in arginine and serine residues (RS domain) that facilitates its interaction with other SR splicing factors (13).
To date several studies have investigated the roles of SRSF2 in viral infection. In HIV-1 infection, SRSF2 was reported to modulate viral replication by negatively regulating the levels of genomic RNA and HIV-1 structural proteins (16). Additionally, SRSF2 regulates the splicing of the viral factors Tat (17)(18)(19)(20) and Rev (21), which are necessary for efficient HIV-1 replication. In human papillomavirus 16 (HPV16), SRSF2 positively regulated the expression and stability of E6E7 RNAs (22). In this report we reveal that the host splicing factor SRSF2 facilitates HSV-1 replication by regulating HSV-1 ICP0, infected cell polypeptide 27 (ICP27), and thymidine kinase (TK) expression. At the transcriptional level, SRSF2 functions as a transcriptional activator by directly binding to the HSV-1 ICP0, ICP27, and TK promoters. At the post-transcriptional level, SRSF2 participates in ICP0 pre-mRNA splicing by co-localizing with ICP0 exon 3. These findings provide insight into the functions of SRSF2 in HSV-1 replication and gene expression and broaden our views concerning the mechanisms through which gene expression is regulated.

SRSF2
Modulates HSV-1 Replication-HSV-1 infection has been reported to alter the localization of the SR domain splicing factor SRSF2, resulting in the inhibition of host cell pre-mRNA splicing (23). However, the effects of SRSF2 on the HSV-1 infection process have not been elucidated. To investigate the role of SRSF2 in HSV-1 replication, we first knocked down SRSF2 using SRSF2 siRNA and overexpressed SRSF2 with a pCMV-SRSF2-FLAG construct (Fig. 1, A and B). Then, we examined HSV-1 glycoprotein expression in HeLa cells using fluorescence microscopy. The HSV-1 glycoprotein density and intensity in the SRSF2 knockdown HeLa cells was much lower than that in the cells transfected with negative control siRNA (Fig.  1C). In contrast, increased SRSF2 expression resulted in a much higher density and intensity of the HSV-1 glycoproteins (Fig.  1D), suggesting that SRSF2 affected both HSV-1 replication and spread. Next, cells with down-regulated or up-regulated SRSF2 expression were infected with HSV-1 at a multiplicity of infection (m.o.i.) of 1 for 12, 24, or 36 h. The viral supernatants were collected, and the viral titers were determined using plaque assays. The HSV-1 titers in the culture media from SRSF2knocked down HeLa cells were markedly lower than the titers in HeLa cells transfected with the negative control siRNA (Fig.  1E). The opposite results were observed in the SRSF2-overexpressing cells (Fig. 1F). To investigate the effect of silencing and overexpression of SC5 on cell cytotoxicity in vitro, a Cell Counting Kit-8 (CCK8) assay was performed in HeLa cells transfected with SRSF2 siRNA or SRSF2-FLAG. The results revealed no difference between the SRSF2 siRNA ( Fig. 1G) or SRSF2-FLAG plasmid (Fig. 1H) and control groups 24, 36, and 48 h after the transfections, indicating no significant effect on cell viability or proliferative capacity.
SRSF2 Regulates Viral Gene Expression-In HSV-1-infected host cells, the IE genes are first activated, promoting the expression of viral proteins to facilitate viral replication. To study the mechanism by which SRSF2 modulates HSV-1 replication, we measured the expression levels of genes required for the appropriate expression of viral early and late gene products, such as HSV-1 ICP0, ICP27, and TK, in HeLa cells transfected with the SRSF2 siRNA. In HSV-1-infected HeLa cells, SRSF2 depletion led to the inhibition of ICP0, ICP27, and TK expression at the mRNA ( Fig. 2A) and protein levels (Fig. 2B). Conversely, upregulation of SRSF2 by SRSF2-FLAG plasmid transfection enhanced ICP0, ICP27, and TK expression at the mRNA (Fig.  2C) and protein levels (Fig. 2D). Also, we measured the expression level of host cellular nuclear paraspeckle assembly transcript 1 (NEAT1), an unrelated non-coding gene, in HeLa cells transfected with the SRSF2 siRNA or SRSF2-FLAG plasmid. The qPCR results showed that SRSF2 had no significant effect on NEAT1 expression (Fig. 2, A and C).
SRSF2 Enhances Viral Gene Transcription by Binding to Gene Promoters-Although SRSF2 is well known as a nuclear speckle component that participates in the process of cellular RNA maturation by binding to RNA, recent reports have indicated that it seems to function as a chromatin regulator (24). Therefore, we performed a luciferase assay to examine whether SRSF2 directly regulated viral gene transcription. We generated two luciferase reporter constructs by inserting the promoter fragments of ICP0 and ICP27 into pGL3-enhancer vectors. To study the effects of SRSF2 on the transcriptional activity of the HSV-1 TK promoter, we directly utilized a pRL-TK reporter vector because this vector contained the same HSV-1 TK gene promoter sequence. The results showed inhibited transcriptional activity of the ICP0, ICP27, and TK promoters in the SRSF2-depleted HeLa cells (Fig. 3A) and increased transcriptional activity in the SRSF2-overexpressing HeLa cells (Fig. 3B). Additionally, we measured the transcription of NEAT1 in HeLa cells transfected with the SRSF2 siRNA or SRSF2-FLAG plasmid as a negative control. To identify the SRSF2 binding sites in the ICP0, ICP27, and TK promoters, we designed sets of primer pairs that recognized the upstream and downstream transcriptional start sites (TSSs) of these genes ( Table 1) and performed a chromatin immunoprecipitation (ChIP) assay with anti-SRSF2 antibody or anti-IgG antibody in HSV-1-infected HeLa cells. The NEAT1 promoter was used as a negative control. The results showed that SRSF2 was enriched near the TSSs of ICP0 gene, ICP27 gene, and TK gene (Fig. 3C). Next, we used DNA fluorescence in situ hybridization (FISH) and immunofluorescence microscopy experiments to validate the interaction between SRSF2 and the HSV-1 ICP0, ICP27, and TK promoters. First, we prepared probes by amplifying the P4 region of ICP0, the P2 region of ICP27, and the P3 region of TK; the probes were labeled with biotinylated cytidine bisphosphate at the 3Ј end of the DNA. The results confirmed that SRSF2 was associated with the promoters of these genes (Fig. 3D). Additionally, pixel intensity plots were generated for each merged channel (Fig. 3D, right panels). To determine the specificity of this co-localization, we performed DNA-FISH and immunofluorescence microscopy experiments in mock-infected HeLa cells. As shown in Fig. 3D, bottom panels, the probe recognizing the ICP0 promoter fragment detected no signal in mock-infected nuclei.
In addition, we tested the co-localization of SRSF2 with Ser-5(P)-RNAP II, ICP27, and infected cell protein 8 (ICP8) using the immunofluorescence assay and pixel intensity analysis (Fig.  4A). Of these proteins, RNAP II widely exists at gene promoters and catalyzes DNA transcription and RNA synthesis, and ICP27 has been shown to recruit RNAP II to viral transcription sites to initiate their transcription (25); ICP8 was the first protein to be identified in HSV-1 replication compartments, which function as sites of viral DNA replication and late gene transcription (26). We also conducted an immunoprecipitation assay to confirm this association. SRSF2 siRNA-or negativecontrol-transfected cells were infected with HSV-1 at an m.o.i. of 1 for 4 h. Cells lysates were harvested and subjected to the immunoprecipitation assay with the SRSF2 antibody followed by Western blotting using antibodies against RNAP II, ICP27, ICP8, and SRSF2. RNAP II, ICP27, and ICP8 were specifically pulled down in the negative-control-transfected cell lysates. SRSF2 depletion resulted in a significant reduction in the interaction of SRSF2 with RNAP II, ICP27, and ICP8 (Fig. 4B). To remove the influence of the RNA or DNA in these associations, the cell lysate was preincubated with RNase A and DNase I and then subjected to the immunoprecipitation assay with the SRSF2 antibody. The result demonstrated that DNase and RNase treatment significantly reduced the association of SRSF2 with RNAP II without alteration in the association of SRSF2 with ICP8 and ICP27 (Fig. 4C), as suggested earlier (24). Together, these results suggested that SRSF2 activation enhanced the transcriptional activity of the ICP0, ICP27, and TK promoters through the co-localization of SRSF2 with viral gene promoters, RNAP II, ICP27, and ICP8.
SRSF2 Regulates ICP0 Pre-mRNA Splicing-As a nuclear pre-mRNA splicing factor, SRSF2 completes exon inclusion and exclusion via binding to exon splicing enhancers located within pre-mRNA exons (12). In HSV-1, all viral gene transcripts are intronless with the exception of ICP0. Therefore, we analyzed the ICP0 pre-mRNA sequence to search for SRSF2 binding motifs and identified a general consensus SRSF2-binding motif (27) located in exon 3 between 4735 bp and 4742 bp (Fig. 5A). To test whether SRSF2 binds to this site, we performed an RNA immunoprecipitation assay. SRSF2 siRNA-or negative-control-transfected cells were infected with HSV-1 at an m.o.i. of 1 for 4 h. The cell lysates were harvested and subjected to an immunoprecipitation assay with the SRSF2 antibody or IgG antibody followed by qRT-PCR using primers that recognized the ICP0 exon 1, exon 2, and exon 3 regions, including the potential SRSF2-binding site (Fig. 5A). An unrelated RNA (NEAT1) was used as a negative control. To avoid DNA contamination in the qRT-PCR, we elimi-

SRSF2-mediated HSV-1 Genes Expression
nated the remaining genomic DNA and then reverse-transcribed cDNA with the kit. The results demonstrated that SRSF2 had the ability to pull down fragments of ICP0 pre-mRNA exon 1, exon 2, and exon 3, whereas SRSF2 knockdown markedly decreased the amount of these fragments retrieved by SRSF2 (Fig. 5B). DECEMBER 16, 2016 • VOLUME 291 • NUMBER 51

SRSF2-mediated HSV-1 Genes Expression
Next, we prepared probes via transcribing ICP0 exon 3 containing the SRSF2 binding motif (WT) and ICP0 exon 3 with a deletion in the SRSF2 binding motif (Mut). The probes were labeled with digoxigenin (DIG) at the 3Ј end of the RNA in vitro and used to perform RNA FISH and immunofluorescence assay experiments. As shown in Fig. 5C, SRSF2 largely colocalized with ICP0 exon 3 containing the SRSF2 binding motif. The intensity plot showed yellow fluorescence, indicating colocalization of SRSF2 and ICP0 exon 3 containing the SRSF2 binding motif (Fig. 5C, upper panels). Conversely, the ICP0 exon 3 with a deletion in the SRSF2 binding motif did not colocalize with SRSF2. The intensity plot showed red and green fluorescence, indicating no colocalization between SRSF2 and ICP0 exon 3 due to the deletion of the SRSF2-binding motif (Fig. 5C, lower panels). These data are the first to show that SRSF2 recognizes and colocalizes with HSV-1 transcripts. Finally, we tested whether the colocalization between SRSF2 and ICP0 exon 3 influenced ICP0 pre-mRNA splicing. Because SRSF2 regulates the transcriptional activity of the ICP0 promoter, the expression difference between the ICP0 intron and the exons regulated by SRSF2 cannot be used as a standard for its effects on ICP0 pre-mRNA splicing. Therefore, we utilized the ratio of the ICP0 mRNA expression level to the pre-mRNA expression level to characterize its splicing ability. As shown in Fig. 5D, the expression ratio of the ICP0 mRNA was 4-fold higher than that of the ICP0 pre-mRNA in HeLa cells transfected with the negative control. However, the ratio declined significantly when SRSF2 expression was decreased with the siRNA. Conversely, overexpression of SRSF2 increased the ratio (Fig. 5E). Taken together, these results prove that SRSF2 regulates ICP0 pre-mRNA splicing by recognizing ICP0 exon 3.

Discussion
SRSF2 is a well known SR protein that is involved in genome stability, alternative splicing, and mRNA export and translation control. In our present study we found that SRSF2 promoted HSV-1 replication by regulating viral gene expression at the transcriptional and post-transcriptional levels. At the transcriptional level, SRSF2 was recruited to the promoters of ICP0, ICP27, and TK with RNAP II, ICP27, and ICP8 to initialize the transcription of these genes. At the post-transcriptional level, SRSF2 recognized the SRSF2 binding sites located in ICP0 pre-mRNA exon 3 to modulate its splicing. As an immediate early protein, ICP0 is synthesized and transported to the nucleus after the earliest stage of HSV-1 infection and plays vital roles in the promotion of viral gene expression (28), the distribution of host cellular nuclear bodies (29), the alteration of host gene expression (30), the disruption of antiviral pathways (4,31), and the assembly of new virion particles (32). Recently, Workenhe et al. (33) revealed that SRSF2 did not  t test). B, after co-transfection with SRSF2-FLAG, FLAG, or BLANK and the pGL3 enhancer plasmid containing the ICP0 promoter, ICP27 promoter, or a pRL-TK reporter vector with the same sequence as the HSV-1 TK gene for 36 h, HeLa cells were infected with HSV-1 for 4 h. The NEAT1 promoter reporter was used as a negative control. The relative transcriptional activity of the ICP0, ICP27, TK, and NEAT1 promoters was analyzed with a luciferase assay in three independent experiments. The data are presented as the mean Ϯ S.D. (*, p Ͻ 0.01, Student's t test). C, HeLa cells were infected with HSV-1 for 4 h. ChIP assays were performed with an anti-SRSF2 antibody or an anti-IgG antibody (Ab). -Fold enrichment of the ICP0, ICP27, and TK promoters by anti-SRSF2 relative to the input level was examined with qRT-PCR. The NEAT1 promoter was used as a negative control. D, HeLa cells were infected with HSV-1 at a m.o.i. of 1 for 4 h, fixed, incubated with a biotin-ICP0-P4 fragment (green), a biotin-ICP27-P2 fragment (green), and a biotin-TK-P3 fragment (green) and SRSF2 (red) and subjected to confocal microscopy analysis. The mock-infected nuclei were used as a negative control. The intensity plots for the red and green channels were analyzed with ImageJ software. DAPI (blue) was used to stain the nuclei. Scale bars, 10 m.

TABLE 1 Sequences of primers used in the ChIP assay
Primer positions are indicated relative to transcription start site.

Name
Forward primer Reverse primer Primer positions

SRSF2-mediated HSV-1 Genes Expression
influence ICP0-null HSV-1 replication, suggesting that ICP0 plays key roles in SRSF2-mediated HSV-1 replication. In this study, SRSF2 was revealed to function as a transcriptional activator and splicing factor in ICP0 expression. As a multifunc-tional viral protein, the SRSF2-mediated induction of ICP0 expression may greatly influence the HSV-1 life cycle. In our work we demonstrated that SRSF2 could co-localize with the promoters of ICP0, ICP27, and TK to alter their tran- (green), RNA polymerase II (red), ICP27 (red), and ICP8 (red), and subjected to confocal microscopy analysis. The intensity plots for the red and green channels were analyzed with ImageJ software. DAPI (blue) was used to stain the nuclei. Scale bars, 10 m. B, after 36 h of transfection with the SRSF2 siRNA or negative control siRNA, HeLa cells were infected with HSV-1 at a m.o.i. of 1 for 4 h. Cell lysates were harvested and subjected to an immunoprecipitation (IP) assay with the anti-SRSF2 antibody or the anti-IgG antibody (Ab). The retrieval of RNA polymerase II, ICP27, SRSF2, and ICP8 by endogenous SRSF2 and IgG was measured by Western blotting. Protein ratios for the indicated proteins/␤-actin were analyzed with ImageJ software, and statistical analysis was conducted with data from three independent experiments (right panel). The data are presented as the mean Ϯ S.D. (*, p Ͻ 0.01, Student's t test). C, HeLa cells were infected with HSV-1 at a m.o.i. of 1 for 4 h. Cell lysates were harvested with and without DNase and RNase treatment and then subjected to an immunoprecipitation assay with the anti-SRSF2 antibody. The retrieval of RNA polymerase II, ICP27, SRSF2, and ICP8 by endogenous SRSF2 was measured by Western blotting. DECEMBER 16, 2016 • VOLUME 291 • NUMBER 51 scriptional activity. First, SRSF2 positively regulated ICP0, ICP27, and TK expression; this regulation was associated with altered transcription. SRSF2 was originally recognized as a pre-mRNA-binding protein. However, Ji X et al. (24) demonstrated that SRSF2 appeared to associate with gene promoters instead of the RNA. Therefore, we hypothesized that SRSF2 might regulate the transcription of viral genes via interacting with their promoters. Based on the ChIP assay results, we concluded that SRSF2 binding sites were located in the TSSs of the ICP0, ICP27, and TK promoters. These results demonstrate for the first time that speckles component SRSF2 co-localizes with HSV-1 DNA, suggesting that the HSV-1 genome is recruited to the nuclear sub-structural speckles to regulate its replication and gene expression. Because HSV-1 genes are transcribed by RNAP II, we tested the relationship between SRSF2 and RNAP II. The immunofluorescence assay and immunoprecipitation assay results confirmed that SRSF2 physically interacts with RNAP II. Additionally, SRSF2 was shown to associate with

SRSF2-mediated HSV-1 Genes Expression
ICP8, an HSV-1 single-stranded DNA-binding protein that functions as an activator of viral DNA replication and late gene transcription (34). These results revealed the potential ability of SRSF2 to regulate gene expression by affecting viral gene transcription in addition to its traditional roles in pre-mRNA splicing.
Generally, SRSF2 participates in cellular pre-mRNA and alternative splicing by binding to its exon splicing enhancers (12). To screen the ICP0 pre-mRNA sequence, we identified general consensus SRSF2 binding motifs located in exon 3. The RNA immunoprecipitation assay and immunofluorescence assay results confirmed that SRSF2 colocalized with ICP0 pre-mRNA exon 3. To better show its effect on pre-mRNA splicing, we utilized the ratio of the ICP0 mRNA expression level to the pre-mRNA expression level to characterize its splicing ability.
As a traditional RNA-binding protein, SRSF2 bound to the HSV-1 ICP0, ICP27, and TK promoters to initiate their transcription in association with RNAP II and ICP27. Moreover, SRSF2 participated in ICP0 pre-mRNA splicing by recognizing SRSF2 binding motifs located in ICP0 exon 3. Taken together, our study provides useful insight into the functions of SRSF2 in viral infection. bated with the SRSF2 antibody (Abcam, ab11826) and IgG antibody (Abcam, ab205719). Chromatin DNA was purified with the Dynabeads Protein G kit (Invitrogen, 10004D) and subjected to real-time PCR. The region-specific primers are listed in Table 1.
Immunoprecipitation Assay-The immunoprecipitation assay was performed using the immunoprecipitation protein G Dynabeads kit (Invitrogen, 10007D) according to the manufacturer's protocol. To digest the RNA or DNA, the cell lysate was preincubated with either 250 g/ml RNase A (Takara, 2158) or 50 units of DNase I (Takara, 2212) for 10 min.
RNA/DNA FISH and Immunofluorescence Microscopy-To evaluate the influence of SRSF2 on HSV-1 replication, HeLa cells were transfected with SRSF2 siRNA or SRSF2-FLAG plasmid for 36 h. After infection with HSV-1 for 12 h, the cells were fixed with 4% paraformaldehyde for 10 min. After blocking, the cells were incubated with an anti-HSV-1 antibody (Abcam, ab9533). The cells were washed, counterstained with DAPI, and observed with an Olympus FV1000 confocal laser microscope. The fluorescence signal value was measured with Image J software and normalized to a single cell. To validate the interaction between SRSF2 and the ICP0, ICP27, and TK promoter fragments, DNA FISH was performed as previously described in Zhang et al. (37). Briefly, the P4 region of the ICP0 promoter (Ϫ380-(Ϫ206)), the P2 region of the ICP27 promoter (138 -297), and P3 region of the TK gene (-227-(Ϫ65)) were amplified and labeled with biotin using the BioNick Labeling System (Invitrogen, 18247-015). HSV-1-infected HeLa cells were incubated with these biotin-labeled promoter regions (10 ng/l) at 42°C for 12-16 h. For the co-localization studies, the cells were fixed again and incubated with the anti-SRSF2 antibody (Abcam, ab11826) after DNA-FISH. The cells were washed, counterstained with DAPI, and observed with an Olympus FV1000 confocal laser microscope. The intensity plots for the red and green channels were analyzed with ImageJ software. DNA-FISH and immunofluorescence microscopy experiments in mock-infected nuclei were used as a negative control. To investigate the interaction between SRSF2 and RNAP II, ICP27, or ICP8, HSV-1-infected HeLa cells were fixed and incubated with the anti-SRSF2 antibody (Abcam, ab11826 or Santa Cruz Biotechnology, sc-10252), the anti-RNAP II antibody (Abcam, ab5131), the anti-ICP27 antibody (Santa Cruz Biotechnology, sc-17544), or the anti-ICP8 antibody (Abcam, ab20194). The cells were washed, counterstained with DAPI, and observed with an Olympus FV1000 confocal laser microscope. The intensity plots for the red and green channels were analyzed with ImageJ software. To verify the interaction between SRSF2 and ICP0 exon 3, the ICP0 exon 3 region with or without the potential SRSF2-binding site was transcribed in vitro by T7 RNA polymerase and labeled with DIG) using the DIG Northern Starter Kit (Roche Applied Science, 12039672910). RNA-FISH and immunofluorescence microscopy were performed as previously described in (38). The intensity plots for the red and green channels were analyzed with ImageJ software.