Regulation of human neurotropic JC virus replication by alternative splicing factor SF2/ASF in glial cells

Abstract

Background: The human neurotropic virus, JC virus (JCV), is the etiologic agent of the fatal demyelinating disease of the central nervous system, Progressive Multifocal Leukoencephalopathy (PML) that is seen primarily in immunodeficient individuals. Productive infection of JCV occurs only in glial cells, and this restriction is, to a great extent, due to the activation of the viral promoter that has cell type-specific characteristics. Earlier studies led to the hypothesis that glial-specific activation of the JCV promoter is mediated through positive and negative transcription factors that control reactivation of the JCV genome under normal physiological conditions and suppress its activation in non-glial cells.

Methodology/principal findings: Using a variety of virological and molecular biological approaches, we demonstrate that the alternative splicing factor SF2/ASF has the capacity to exert a negative effect on transcription of the JCV promoter in glial cells through direct association with a specific DNA sequence within the viral enhancer/promoter region. Our results show that down-regulation of SF2/ASF in fetal and adult glial cells increases the level of JCV gene expression and its replication indicating that negative regulation of the JCV promoter by SF2/ASF may control reactivation of JCV replication in brain.

Conclusions/significance: Our results establish a new regulatory role for SF2/ASF in controlling gene expression at the transcriptional level.

Figure 1. Overexpression of SF2 inhibits JCV propagation in PHFA cells.

A. Southern blot analyses of JCV-infected PHFA cells. In lane 1, 2 ng of lineralized Mad-1 genome was used as positive control. In lane 2, DNA samples from uninfected cells were loaded as negative control. B. Western blot analyzes of whole cell extracts prepared in parallel to DNA samples in panel A, using specific antibodies against VP1 and agnoprotein. In lane 1, whole cell extracts from uninfected cells was loaded as negative control. Western blot analyzes of same extracts with anti-Grb2 antibody was used as loading control. C. Q-PCR analyses of the viral particles in the JCV infected-cells culture medium. D. RT-PCR analyses of JCV early (T-Ag) and late (VP1) gene products in JCV infected PHFA cells. In lane 1, Kb ladder was loaded as molecular weight marker. In lane 2, JCV Mad-1 genome was used as a positive control. E. Western blot analyzes of endogenous (SF2/ASF) and overexpressed (T7-SF2/ASF) levels of SF2/ASF in PHFA cells. Grb2 was probed in the same membranes as loading control.

Figure 2. Effect of SF2/ASF on JCV early gene splicing in glial cells.

A. Schematic structure of JCV early region unspliced and spliced RNAs and, the size of the expected amplification products with a primer set (PF and PR), used for the amplification of JCV gene products in panels B, C, and D. B. SF2/ASF inhibits expression of JCV early region driven by its own promoter. pBlueScript KS JCV early reporting JCV early region under JCV promoter (pBlue-JCV.E), was transiently transfected into PHFA cells in the presence or absence of an SF2/ASF expression plasmid. Total RNA was extracted and used for cDNA synthesis by reverse transcriptase reaction. JCV-early region gene products (pre-mRNA, t-Ag and T-Ag) were amplified and separated on a 3% agarose gel and stained with ethidium bromide. Lane 2 was pBlue-JCV.E plasmid DNA amplified as positive control. Lane 3 was untransfected PHFA cell extracts used as a negative control. All the bands reflecting the amplification products were separately cut from the gels, purified and RNA identities were confirmed by sequencing. GAPDH was also amplified in the same cDNA samples as input control. C. SF2/ASF inhibits splicing of JCV early region driven by CMV promoter. PHFA cells were transiently-transfected with pCMV-JCV.E in the presence or absence of an SF2/ASF expression plasmid. Total RNA was extracted and processed for reverse transcriptase reaction as described in panel B. In lane 2, pCMV-JCV.Early plasmid DNA was amplified and loaded as a positive control of primary transcript. Lane 3 was un-transfected PHFA cell extracts and was used as negative control of amplification. D. BSB8 cells were transfected with either vector alone (lane 3) or with SF2/ASF (lane 4) expression plasmids. Expression of JCV-Early gene products were analyzed by RT-PCR as described in panels B and C. U-87 MG cells (lane 2) were also processed as lanes 3 and 4, and were used as negative control of amplification. E. Over-expression of SF2/ASF inhibits T-Ag expression in BSB8 cells. BSB8 cells were plated in 6-well dishes and transfected with SF2/ASF expression plasmid in increasing concentrations (1X and 3X). Whole cell extracts were prepared at 2nd and 5th days post-transfection and processed by Western blotting using specific antibodies against T-Ag, SF2/ASF, and T7-SF2/ASF. Grb2 was probed as a loading control after stripping of the same primary membranes used for Large T antigen and T7-tagged SF2/ASF protein detections. F. SF2/ASF shows no inhibitory effect on SV40 Large T antigen expression in SVG-A cells. SVG A cells were plated in 6-well plates and transfected either with SF2/ASF-sense or with SF2/ASF-antisense expression plasmids. T-Ag expression was detected by Western blotting. Expression of SF2/ASF and Grb2 was detected in the same blots by using SF2/ASF-specific, T7-specific and Grb2-specific antibodies.

Figure 3. SF2/ASF inhibits JCV-early (JCV_E) and –late (JCV_L) transcription in PHFA cells.

A. JCV_E (black bars) or JCV_L (grey bars) reporter plasmids were transiently transfected into PHFA cells either alone or in combination with an SF2/ASF expression plasmid. B. SF2/ASF-mediated inhibition of JCV-early transcription can be rescued by a specific siRNA against SF2/ASF. JCV_E reporter plasmid was transiently transfected into PHFA cells either alone or in combination with an SF2/ASF expression plasmid (lanes 2-6). At 12 h post-transfection, cells were transfected either with a non-targeting siRNA or with a siRNA specific to SF2/ASF. C. Western blot analyses of the same extracts used for reporter assays in panel B, using specific antibodies to T7-SF2/ASF and Grab2. D. Band intensities of SF2/ASF expression from panel E were quantified and showed as bar graph. The Student's t-test was performed to calculate “P” values in panels A, B, and D.

Figure 4. SF2/ASF directly targets JCV regulatory region in vivo and in vitro.

A. BSB8 cells were transfected with an T7-SF2/ASF expression vector, cross-linked and ChiP assay performed using antibodies to Large T antigen (lane 4), Normal mouse serum (lane 5), T7 tagged SF2/ASF (lane 6), and no antibody (lane 3). In lane 2, JCV Mad-1 genomic DNA was used as positive control. B. Gel shift analysis of JCV oligonucleotides spanning the 98-bp repeated region of the viral promoter with recombinant SF2/ASF. Four oligonucleotide probes (CR1, CR2, CR3, CR4), and JCV regulatory region with early and late orientation are schematized at the top of the panel. C. Antibody supershift gel electrophoresis. CR3 oligonucleotide and GST-SF2/ASF complexes were incubated with normal rabbit serum (NRS) or anti-GST antibodies for an additional 20 min prior to gel electrophoresis. The star depicts a nonspecific band and the arrowhead points shifted specific complexes and the arrow points antibody supershifted specific complexes. D. Coomassie blue staining of GST (lane 1) and GST-SF2/ASF (lane 2).

Figure 5. Characterization of the interaction between SF2/ASF and JCV regulatory region by gel shift assays.

A. Competition analyses of the CR3 oligonucleotide of JCV regulatory region. Nuclear extracts from PHFA cells incubated with end-labeled double stranded CR3 oligonucleotide probe in the presence of cold CR3 (lanes 3 and 4) and CR2 (lanes 5 and 6) oligonucleotides. B. Presence of RNA in nuclear extracts influences the binding pattern to CR3 oligonucleotide. Nuclear extracts from PHFA cells incubated with end-labeled double stranded CR3 oligonucleotide probe in the presence or absence of RNase A (lanes 2 and 3, respectively). C. Gel shift analysis of JCV CR3 oligonucleotide in the primary human fetal glial cell extracts. End-labeled (lanes 2 and 4) or cold (lane 3) CR3 oligonucleotide incubated with PHFA nuclear extracts. Radioactively-labeled CR3/NP complexes (lanes 2 and 4) were used as reference to label and to cut the gel pieces containing the CR3/NP complexes in cold oligonucleotide reaction (lane 3, GP2) and in nuclear extract control lane (lane 5, GP1). Dotted-lines indicate the vertical and horizontal lining of expected running pattern of CR3/NP complexes. Solid boxes indicate the gel pieces (GP1, GP2), that were cut from the native gel. D. Nucleoprotein complexes in native gel pieces (GP1 and GP2) from panel C were denatured, resolved by SDS-PAGE, and were analyzed by Western blotting, using a specific anti-SF2/ASF antibody. In lane 1, nuclear extract from PHFA cells was loaded as positive control.

Figure 6. “RRM1” domain of SF2/ASF is required for the inhibition of JCV-Early promoter.

A. Schematic representation of SF2/ASF-FL (1) and its functional domain mutations, Mut.RS (2), RRM1-alone (3), and Mut.RRM1 (4). B. Expression of T7-SF2/ASF-FL and truncated forms of SF2/ASF were detected by using anti-T7 (top panel) and anti-SF2/ASF (bottom panel) antibodies. C. BSB8 cells transfected with T7-SF2/ASF constructs, schematized in panel A, cross-linked and ChiP assay performed using antibodies to actin (lane 4), to T-Ag (lane 5), to T7 (lanes 6 to 9), and no antibody (lane 3). D. JCV_E reporter plasmid transiently transfected into PHFA cells either alone (lane 1) or in combination with T7-SF2/ASF-FL-antisense (lane 2), T7-SF2/ASF-FL-sense (lane 3), T7-SF2/ASF-Mut.RS (lane 4), T7-SF2/ASF-RRM1 (lane 5), and T7-SF2/ASF-Mut.RRM1 (lane 6) expression plasmids. E. Subcellular localization of SF2/ASF and its truncated forms in PHFA cells.

Figure 7. Downregulation of SF2/ASF induces JCV propagation in primary human fetal and adult astrocytes.

A. Western blot analyses of SF2/ASF expression in PHFA and PHAA cells. Grb2 was probed as a loading control. Band intensities were quantified and shown as a bar graph. B. PHFA and PHAA cells infected with a lentivirus encoding SF2/ASF-shRNAs at day 0, transfected/infected with JCV Mad-1 strain at day 1. Q-PCR analyses of JCV viral particles in culture supernatants were performed as described in Fig. 1C. C. Whole cells extracts were prepared from the same infections as described in panel B and processed for Western blotting by using specific antibodies against SF2/ASF and VP1. Grb2 was probed as loading control. D. Immunocytochemical analyses of VP1 expression in PHFA and PHAA cells infected with JCV. E. Quantification of VP1 expression in infected cells from panel D.