Upregulation of Cellular Bcl-2 by the KSHV Encoded RTA Promotes Virion Production

Apoptosis of virus infected cells can restrict or dampen full blown virus propagation and this can serve as a protective mechanism against virus infection. Consequently, viruses can also delay programmed cell death by enhancing the expression of anti-apoptotic proteins. Human Bcl-2 is expressed on the surface of the mitochondrial membrane and functions as the regulator of the delicate balance between cell survival and apoptosis. In this report, we showed that the replication and transcription activator (RTA) encoded by KSHV ORF 50, a key regulator for KSHV reactivation from latent to lytic infection, upregulates the mRNA and protein levels of Bcl-2 in 293 cells, and TPA-induced KSHV-infected cells. Further analysis revealed that upregulation of the cellular Bcl-2 promoter by RTA is dose-dependent and acts through targeting of the CCN9GG motifs within the Bcl-2 promoter. The Bcl-2 P2 but not the P1 promoter is primarily responsive to RTA. The results of ChIP confirmed the direct interaction of RTA protein with the CCN9GG motifs. Knockdown of cellular Bcl-2 by lentivirus-delivered small hairpin RNA (shRNA) resulted in increased cell apoptosis and decreased virion production in KSHV-infected cells. These findings provide an insight into another mechanism by which KSHV utilizes the intrinsic apoptosis signaling pathways for prolonging the survival of lytically infected host cells to allow for maximum production of virus progeny.


Introduction
Kaposi's Sarcoma-Associated Herpesvirus (KSHV), the etiological factor associated with Kaposi's Sarcoma is also known as human herpesvirus 8 (HHV 8). Several other malignancies such as primary effusion lymphoma, and multicentric Castleman's disease have also been known to associate with KSHV [1,2,3,4]. KSHV belongs to the c-herpesviruses family, the life cycle of which consists of two distinct phases, latent and lytic replication [5,6,7,8]. During latency, the virus establishes persistent infection and only a small subset of genes such as ORF73, K12, ORF72, ORF71 and K15 are typically expressed [9,10]. Under conditions of lytic reactivation such as hypoxia the virus spreads to new target cells and the viral genes are activated in cascade mode [11,12,13]. The switch to lytic reactivation is experimentally induced by a number of intracellular or extracellular motivators including the chemical agents 12-O-tetradecanoyl phorbol-13 acetate (TPA) and sodium butyrate [14,15].
Apoptosis is a major antiviral cellular response against viral infection. The B-cell leukemia/lymphoma 2 (Bcl-2) family of proteins controls the intrinsic mitochondrial pathway of cellular apoptosis [35,36]. The Bcl-2 group of proteins is classified as proand anti-apoptotic members, all of which contain at least one highly conserved Bcl-2 homology (BH) domain [37]. The antiapoptotic proteins, such as Bcl-2, Mcl-1, and Bcl-x L , share BH domains 1-4 [38]. The pro-apoptotic members include the Bax and the BH3-only families. The members of the bax family such as bax, bak share BH domains 1-3, and the BH3-only families have only the short BH3 motif respectively. This BH3 motif plays a central role in the killing action of molecules [37]. Moreover, Bcl-2 functions as one of the important regulators to maintain the delicate balance between cell survival and apoptosis. The human Bcl-2 gene is overexpressed in numerous human cancers, including B-and T-cell lymphomas, cervical, lung, breast, prostate and colorectal cancers [39,40,41,42,43,44]. Its overexpression in tumor cells not only functions as an apoptosis inhibitor, but also results in resistance to chemotherapy or radiotherapy-induced apoptosis [45,46]. Thus, because of its function in anti-apoptosis, the Bcl-2 gene has become a strong potential target in development of anticancer therapies [47,48,49].
Bcl-2 protein levels can be regulated transcriptionally and post transcriptionally. Furthermore, the human Bcl-2 gene contains both the P1 and P2 promoters [47,49]. P1, the predominant promoter, is located 1.3-to 1.5-kbp upstream of the translation start site. The P1 promoter is mainly GC rich, with a number of transcription initiation sites and includes seven consensus binding sites for the Sp1 transcription factor. However, the P2 promoter has canonical CAAT as well as TATA boxes [47,49]. Some transcriptional factors such as CREBP (cAMP responsive element binding protein), and NFkB are known to be positive regulators of Bcl-2 transcription, and a few negative regulatory sites have also been described, including WT1, p53 binding sites [47]. Studies have shown that G-quadruplex structure upstream of the P1 promoter plays a role in transcription regulation [48,49]. In addition, Bcl-2 may also be modified post-translationally by phosphorylation which occurs at mitogen-activated protein kinase (MAPK) sites which can trigger proteasome-dependent degradation and ubiquitination [50]. Caspase-dependent cleavage can also occur leading to loss of its anti-apoptotic activity [50].
We now show that RTA can transactivate the cellular Bcl-2 promoter in a dose-dependent manner. RTA upregulates Bcl-2 through targeting of CCN 9 GG-like RTA responsive elements (RREs) within the promoter and that the P2 promoter is important for RTA-mediated transcriptional regulation of cellular Bcl-2. These findings provide insights into new mechanisms by which RTA transactivates cellular Bcl-2 and so contributes to enhanced efficiency of KSHV lytic replication.

Results
Bcl-2 transcipts and protein levels are elevated in cells transiently expressing RTA Bcl-2 mRNA transcripts and protein levels were examined to specifically detect the RTA-mediated cellular Bcl-2 upregulation (Fig. 1). HEK 293 and DG75 cells were transiently transfected with increasing amounts of RTA expression plasmid. Total RNAs were isolated from cells for use in RT-PCR assay. The results of the RT-PCR analysis showed that RTA clearly enhanced the Bcl-2 transcript levels in a dose-dependent manner. Furthermore, results of Western blot analyses also supported increased upregulation of Bcl-2 protein in a dose-dependent manner (Fig. 1A, B). Therefore, expression of RTA directly affects the transcription of Bcl-2 and upregulates Bcl-2 protein levels.
Bcl-2 expression is upregulated in TPA-induced KSHVinfected BC3 and BCBL1 cells KSHV lytic replication can be initiated by chemical inducers, like TPA and sodium butyrate [51]. Being an immediate early protein, expression of RTA can be observed at the first hour after TPA treatment [55,56,57]. To determine the changes in Bcl-2 level in KSHV induced cells, KSHV was induced from BC3 and BCBL1 cells as mentioned above and cells were harvested for analysis. The results from our RT-PCR and Western blot analyses showed that Bcl-2 expression were strongly upregulated in TPA and butyrate induced KSHV-positive B cells ( Fig. 2A, B). In contrast, there was no background RTA expression in uninduced cells subjected to the same analysis, and as expected in the KSHVnegative BJAB cell there was no RTA expression after induction. Moreover, the expression of Bcl-2 mRNA and protein in BJAB cells when treated with TPA had little or no obvious effect when uninduced cells were compared to the induced cells suggesting that the upregulation of Bcl-2 is due to the presence of KSHV and the viral induced factor RTA (Fig. 2).
The levels of Bcl-2 mRNA is enhanced in RTA-expressing

HEK 293 cells and in KSHV-infected B cells induced for lytic replication
In addition to RT-PCR analysis, we wanted to corroborate our results and so performed Northern blot analysis to verify Bcl-2specific mRNA in RTA expressing HEK 293 cells and in induced KSHV-infected B cells by using a Bcl-2-specific DNA probe. Results showed that the levels of total Bcl-2 mRNA were strongly increased in a dose-dependent manner in HEK 293 cells transfected with the RTA expression plasmid. Additionally, total Bcl-2 mRNA expression in TPA induced BC3 and BCBL1 cells was also strongly enhanced compared to uninduced cells with no evidence of change in KSHV negative BJAB cells. The same blots were re-probed with a GAPDH specific probe as control. The density of Bcl-2 signal was normalized against that of the GAPDH control to quantify the levels of Bcl-2 mRNA expression and was presented as a histogram for relative density levels (Fig. 3).

RTA activates the Bcl-2 gene promoter in a dosedependent manner
Since TPA-induced KSHV-infected cells and RTA expressing cells showed higher levels of Bcl-2 mRNA and protein expression, we further explored the molecular mechanism of Bcl-2 promoter upregulation by RTA. To analyze the role of RTA on regulating the Bcl-2 promoter, a fixed amount of full length Bcl-2 promoter luciferase construct (pGL3-Bcl-2) [47,49], was cotransfected with increasing amounts of pcDNA-RTA into HEK 293 and DG75 cells. After 24 h, the cells were harvested for luciferase reporter assay. The reporter assays showed that RTA can indeed activate the Bcl-2 promoter by approximately 12 to 35-fold compared to that of the vector alone in HEK 293 cells as well as 5 to 25-fold in DG75 cells ( Fig. 4 A and B). These results suggest that RTA can up-regulate Bcl-2 expression through regulation of cis-acting elements in the promoter.
The P2 promoter and the CCN 9 GG domains both contribute to Bcl-2 promoter activation by RTA To identify the potential motifs indispensable for the upregulation of cellular Bcl-2 by RTA, a series of truncated promoter luciferase contructs were generated and tested by reporter analysis in both HEK 293 and DG75, as shown in Fig. 5 A to C.
The two promoters P1 and P2 [47,49] and 9 CCN 9 GG-like RTA-responsive elements (RREs) are located within the full length Bcl-2 promoter (Fig. 5A). The CCN 9 GG-like sequences within the Bcl-2 promoter are similar to one of the RREs identified in the KSHV ORF57 promoter [58]. To identify which of these motifs play a critical role in regulating transcription of the Bcl-2 promoter by RTA, three truncations of the promoter named pGL3-DRREs, pGL3-DP1, pGL3-DP2 were generated to delete the RREs, the P1 and the P2 promoters, respectively. These promoter constructs were transfected into HEK 293 and DG75 cell lines with different amounts of RTA plasmids. As shown in the Fig. 5B and C, the deletion of P1 promoter did not result in change in promoter activity in HEK 293 cells (Fig. 5B). In contrast, deletion of P2 promoter as well as the deletion of the 9 CCN 9 GG-like RREs resulted in a complete shutdown of the promoter activity with an almost negligible change in activity when compared to promoter alone without RTA expression. Our data strongly indicated that the P2 promoter region and the CCN 9 GG domains were indispensable for RTA-mediated upregulation of Bcl-2.
To confirm the role of the CCN 9 GG motifs in Bcl-2 regulation, we performed reporter assays by using the pGL3-DRRE1,2 luciferase construct in which the first and second CCN 9 GG motifs were deleted (Fig. 6A). We found that mutation of the first and second CCN 9 GG RRE motifs in the Bcl-2 promoter dramatically diminished promoter activity compared with that of the pGL3-DP1 alone (Fig. 6B). First of all without RTA there was little or no activity comparable to basic luciferase vector alone. However, with the RTA expression construct the results showed about a 10 fold increase with the DRRE1,2, but a decrease of approximately 3 times compared to the DP1 reporter which contained all the RREs (Fig. 6B). This suggests that the CCN 9 GG motifs within the Bcl-2 definitely play a critical role in RTA-mediated regulation of this promoter. To show direct interaction of RTA protein with the CCN 9 GG motifs, CHIP was carried out using induced and uninduced BCBL1 cells. The presence of the first and second CCN 9 GG motifs within the immunoprecipitated complex was confirmed by PCR amplification. A clear band was detected in RTA antibody precipitated induced nuclei DNA, whereas there is no signal in RTA antibody precipitated uninduced nuclei DNA, and the two IgG negative controls failed to recover any PCR product (Fig. 6C, upper panel). Real-time PCR was also used to quantify the precipitated DNA. The amount of RTA antibody precipitated induced nuclei DNA was about 8 to 32 times higher when compared with that of the control groups (Fig. 6C, bottom panel).

Bcl-2 expression is important for prolonged survival of lytically infected host cells and contributes to enhanced production of virus progeny
On the basis of the results we observed above, we hypothesized that RTA could take advantage of the Bcl-2 pathway to promote cell survival which may result in enhanced virus progeny production. Hence, we wanted to determine whether upregulation  of cellular Bcl-2 by RTA may in fact affect cellular apoptosis. We constructed stable Bcl-2 knockdown cells (BC3-shBcl2, BJAB-shBcl2) and their control cells (BC3-shC,B JAB-shC) by lentivirus infection and puromycin selection (Fig. 7A). The expression of Bcl-2 in the Bcl-2 knockdown and the control cell lines were monitored by Western blot. Bcl-2 protein level was considerably decreased in both BJAB and BC3 cells transduced with the shBcl-2 lentivirus (Fig. 7B). To quantitate the KSHV virus progeny production, the virus supernatant of induced BC3-shC or BC3-shBcl2 cells was prepared at 48 h postinduction, and used as a template for PCR analysis of the KSHV ORF9. Our data indicated that Bcl-2 knockdown reduced the production of virus particles as determined by detection of the quantified product of the K9 gene amplified by PCR (Fig. 7C). This was about 2-fold less than the progeny production in the control cell line. These results indicate that RTA-mediated enhanced expression of Bcl-2 contributes to virus progeny production during lytic replication. The two cell lines examined had a distinctly different pattern of apoptosis when treated with TPA and butyrate. Under induction, the BJAB cell line which is negative for KSHV resulted in similar levels of apoptosis which was about 62-64% when the BJAB-shC cells were compared to the BJAB-shBcl2 cells (Fig. 7D). Importantly, the results from the KSHV-infected BC3 cells showed a clear increase in the level of apoptosis when comparing the BC3-shC cells to the BC3-shBcl2 cells (p,0.05) (Fig. 7D). These findings provide convincing evidence supporting the hypothesis that RTA can upregulate Bcl-2 expression and have a direct role in enhancing cell survival and thus an increase in virus progeny production.

Discussion
Similar to other large DNA herpesviruses, KSHV has two major life cycles referred to as the latent and lytic replication cycles. Latency exists in most of the KSHV infected cells which harbor the viral episomes [59,60], but a minority of cells do undergo the full lytic program to sustain KS pathological changes [61,62]. KSHV viral genes are successively activated in the lytic replication phase which can be triggered by RTA [53]. RTA regulates the KSHV lytic phase to reactivate the virus from latency and so initiate replication [55,56,57]. Previous studies showed that RTA transactivates the transcription of its downstream gene in two ways, one is through binding RREs, and another is by interacting with cellular transcription factors, such as K-RBP [63,64], CEBP-a [33,65], Oct-1 [30], and RBP-Jk [24].
Apoptosis of infected cells can restrict or dampen full blown virus replication and this can serve as a protective defense mechanism against virus infection. Consequently, viruses can also delay programmed cell death by enhancing the expression of antiapoptotic proteins [66,67,68]. The ability to suppress apoptosis may be essential to drive the enhanced survival of KSHV-positive cells so that the virus may replicate and spread, thus resulting in persistence of the virus in the host. KSHV can employ a number of molecular strategies which are related to induction of cell transformation, regulation of cell proliferation, and resistance of apoptosis. Therefore, there is a high probability that RTA contributes to virus lytic replication through the activation of cellular Bcl-2 expression very early during infection.
Human Bcl-2 is expressed on the surface of the mitochondrial membrane and functions as a major regulator of the delicate balance between cell survival and apoptosis [47,49]. In this study, we now present evidence for the interplay between RTA and Bcl-2. Both Bcl-2 mRNA and protein levels were upregulated in RTAexpressing human cells and in TPA-induced KSHV-infected cells. Furthermore, we demonstrated that RTA may in fact directly activate the Bcl-2 gene promoter. As expected, results of the reporter assay further established that activation of Bcl-2 by RTA was dose-dependent, and that the effect was not cell type specific. Thus, these experiments have now convincingly established an association between RTA expression and Bcl-2 upregulation. To determine the mechanisms by which RTA activates Bcl-2, we identified potential responsive elements within the Bcl-2 promoter. Interestingly, this promoter contains nine CCN 9 GGlike motifs and these CCN 9 GG-like motifs were recently identified KSHV RRE present in five gene promoters of KSHV [58] . Both experiments in vitro and in vivo have shown that this RRE was able to react with RTA so as to regulate the activation of RTA [58]. This provided the likelihood that a similar strategy may be used in upregulation of Bcl-2 by RTA. Our data from the reporter assays with a series of truncated promoters demonstrated that these CCN 9 GG-like motifs in fact do have an effect on transactivation of Bcl-2 by RTA. Deletion of the nine CCN 9 GG motifs completely shut down the Bcl-2 promoter activity. Additionally, deletion of the first and second CCN 9 GG motifs also led to a significant decrease in RTA-mediated Bcl-2 promoter activity. Furthermore, the results of ChIP confirmed the direct interaction of RTA protein with RREs within the Bcl2 promoter. This confirms our hypothesis that the CCN 9 GG motifs are important and likely to be critical for transactivation of Bcl2 by RTA. Interestingly, a highly consistent sequence, GNCCN 9 GGNG, with an identification as EBV RRE has been reported [69]. Another previous study on the EBV RTA-responsive promoters pointed out that the differential responsiveness of the CCN 9 GG motifs to RTA transcription and KSHV reactivation may result from a variation in the N 9 sequences [70]. The KSHV five gene promoters containing the consensus sequence CCN 9 GG also contained variations in the N 9 sequences [58]. The PAN-RRE had stronger interactivity with RTA than the K2-and K12-RREs [23]. A few cellular factors, as mentioned above, can clearly interact with RTA [53]. Their cis-acting binding domains are possibly in the vicinity of the CCN 9 GG sequence [58]. Thus, RTA can either bind to the CCN 9 GG motif or these cellular proteins bound to their cis-acting elements which may result in optimal activation of the promoter. In addition, our data also showed that the P2 promoter is indispensable for RTA-mediated transactivation of Bcl-2. The pGL3-DP1 construct had no detectable reduction of luciferase activity compared with the full length Bcl-2 promoter, whereas mutation of P2 promoter region was able to completely shut down the promoter activity. The P2 promoter contains both a CCAAT and a TATA box and an octamer motif [71]. However, the P1 region contains a TATA-less GC-rich promoter with seven binding sites for the Sp1 transcription factor [71]. How is it that the P2 promoter has a preference for Bcl-2 transactivation by RTA? We cannot however exclude the possibility that the TATA box and/or the octamer motif within the P2 promoter are indispensable for RTA-mediated upregulation of Bcl-2. These studies are yet to be done and will hopefully provide a deeper understanding of the mechanism.
Depletion of Bcl-2 clearly increases the level of cellular apoptosis, and decreases virus progeny production in induced BC3-shBcl2 cells compared with that of the BC3-shC cells. This provides strong evidence that upregulation of cellular Bcl-2 by RTA can definitely contribute to KSHV lytic replication. The utilization of the Bcl-2 pathway by RTA would be a responsible strategy for prolonging the survival of lytically infected cells and thus allow enhanced production of virus progeny.
In summary, our study which investigated the interaction between KSHV RTA and cellular Bcl-2 showed that RTA can transactivate Bcl-2 transcription through targeting of the RRE CCN 9 GG motifs within the Bcl-2 promoter. We also demonstrated that the P2 promoter region instead of the P1 promoter region was highly responsive to RTA. The results of ChIP confirmed the direct interaction of RTA protein with the CCN 9 GG motifs. Knockdown of cellular Bcl-2 by lentivirus-delivered small hairpin RNA (shRNA) resulted in increased cell apoptosis and decreased virion production in KSHV-infected cells. RTA can upregulate Bcl-2-mediated block to cellular apoptosis, and so prolong cell survival leading a further increase in production of virus progeny (Fig. 8). These findings provide an insight into the molecular mechanism by which KSHV interferes with the intrinsic apoptosis signaling pathways to promote viral lytic replication. luciferase constructs were transfected or cotransfected into ten million HEK 293 cells with 10, 20 mg of pcDNA-RTA respectively. Cell lysates were prepared and the luciferase assay was carried out as above mentioned in Fig. 4 legend. Deletion of the first and second CCN 9 GG motifs led to a significant decrease in RTAmediated Bcl-2 promoter activity. (C) A ChIP assay was carried out using induced or uninduced BCBL1 cells. RTA-DNA complexes were specifically precipitated with or without RTA antibody. The precipitated DNA was recovered by PCR with primers covering RRE 1 and 2 (upper panel). Real-time PCR was performed to quantify the immunoprecipitated promoter DNA. doi:10.1371/journal.pone.0023892.g006

Antibodies and cell cultures
RTA antibody was generated from a hybridoma (Lan et al. Institute Pasteur, Shanghai, China). The Bcl-2 mouse monoclonal antibody is a product of Santa Cruz Biotechnology, Inc (Santa Cruz, CA). The GAPDH mouse antibody was obtained from Novus Biologicals, LLC (Littleton, CO).
Human embryonic kidney fibroblast 293 was obtained from Jon Aster (Brigham and Women's Hospital, Boston, MA). BJAB is a KSHV and EBV negative B cell line and was provided by Elliott Kieff (Harvard Medical School, Boston, MA). DG75 is a KSHV negative B cell line. BC3 and BCBL1 are KSHV-positive body cavity-based lymphoma-derived cell lines obtained from the NIH AIDS Research and Reference Reagent Program [51]. DG75 were obtained from the American Type Culture Collection. Cells were cultured as described before [47,52].

Transfection and Induction of KSHV
HEK 293 and DG75 cells were transfected by using a Bio-Rad Gene Pulser II electroporator. Experimental method for transfection by electroporation was previously described [53]. BC3, BCBL1 and BC3-shBcl-2 cells were induced with 20 ng of TPA/ ml and 1.5 mM butyrate [51]. BJAB, BJAB-shC, BC3-shC cells were used as control. All cells were harvested for experimental use at 48 h postinduction.

Western blot
After treatment, the cells were harvested and lysed in a RIPA buffer. The protein concentration of the lysates was measured by Bradford assay. Samples were subjected to sodium dodecyl sulfatepolyacrylamide gel electrophoresis (SDS-PAGE) and then transferred to polyvinylidene difluoride (PVDF) membranes. Western boltting was performed using anti-Bcl2, anti-RTA, or anti-GAPDH antibody and goat anti-mouse IgG secondary antibody conjugated to IR dye 800 (Rockland Immunochemicals, Inc., Gilbertsville, PA). After the membranes were washed, the signals were scanned with Odyssey infrared image (LI-COR, Inc., Lincoln, NE).

Quantitative real-time PCR analysis
Total RNA extraction from cells (HEK 293, RTA-transfected HEK 293, DG75, RTA-transfected DG75, induced and uninduced BC3, BCBL1, BJAB cells), generation of cDNA, preparation of reaction mixture and RT-PCR amplification condition were described previously with minor modifications [47,52]. The primer sequences of Bcl-2 and GAPDH genes used for the real-time PCR are shown in Table 1 [47,52]. Data analyses were carried out by using a StepOnePlus RT-PCR system (Applied Biosystems, INC., Foster City, CA) with each sample tested in triplicates.

Northern blot analysis
A human Bcl-2 gene DNA probe of 539 bp and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) probe of 628 bp were generated by PCR using HEK 293 cells cDNA as template. Primer 3 software was used to design the primers (listed in Table 1). Both DNA probes were confirmed by sequencing (sequencing core, Upenn). The DNA probes were radiolabeled by random priming method with NEB #1500 L kit (New England Biolabs, Ipswich, MA). Ten microgram of total RNA from each cell (HEK 293, RTA-transfected HEK 293, induced or uninduced BC3, BCBL1, BJAB cells) was resolved by formaldehyde-agarose gel electrophoresis and transferred to a 0.22 mm pore nitrocellulose membrane. After prehybridization, the membrane was hybridized successively with 32 P-radiolabeled DNA probes, followed by washes twice with SSC 26 buffer (0.3 M NaCl, 30 mM sodium citrate, pH 7.0, 0.1% SDS) for 5 min at room temperature, then with SSC 0.56 buffer, 0.1% SDS for 10 min at 60uC. The autoradiographic images were analyzed by using a PhosphorImager with supplementary ImageQuant software (Molecular Dynamics Inc., Sunnyvale, CA).

Luciferase assay
Ten million HEK 293 and DG75 cells were cotransfected with 10 mg of the reporter plasmids and increasing amounts of RTA plasmids respectively. Cell lysates were prepared at 24 h posttransfection. Twenty five microliters of luciferase assay substrate was mixed with 40 ml cell lysate, and a LMaxII384 luminometer (Molecular Devices, Inc., Sunnyvale, CA) was used to measure luminescence level. The results showed the mean and standard deviation (SD) of the data of three independent experiments. Western blot analysis was used to confirm the expression of the proteins.

Chromatin immunoprecipitation (ChIP)
At 48 h post-induction, BCBL1 cells were fixed by adding 37% formaldehyde to a final concentration of 1% for 10 min. The crosslinking reaction was stopped by adding glycine to a final concentration of 0.125 M for 5 min. Cells were washed three times with 16 ice-cold PBS and resuspended in cell lysis buffer [5 Mm PIPES (pH 8.0), 85 mM KCl, 0.5% NP-40] containing 1 mM PMSF and 1 mg/ml protease inhibitors for 10 min at 4uC. After centrifugation, the cell pellets were resuspended in nuclear lysis buffer [50 mM Tris/HCl (pH 8.1), 10 mM EDTA, 1% SDS] and incubated for 10 min at 4uC. The resulting solution was diluted 5fold with dilution buffer [16.7 mM Tris/HCl (pH 8.1), 167 mM NaCl, 1.2 mM EDTA, 0.01% SDS, 1.1% Triton X-100, 1 mM PMSF, 1 mg/ml protease inhibitors] and sonicated. After centrifugation, the supernatant was pre-cleared with mouse IgG and immobilized Protein A. Immunoprecipitation was performed by incubation at 4uC with anti-RTA antibody or mouse IgG antibody. Immune complexes were collected by a further incubation with immobilized Protein A for 1-2 h at 4uC. The beads were washed four times with IP buffer [25 mM Tris/HCl (pH 7.2), 150 mM NaCl] and resuspended in elution buffer (1% SDS/0.1 M NaHCO 3 ). After centrifugation, the supernatant was neutralized with 1 M Tris/HCl (pH 7.2). RNase A and NaCl were added, followed by incubation at 65uC overnight to reverse the cross-linked DNA-protein complex. Proteinase K was added, followed by incubation at 55uC for 1-2 h. DNA extraction was carried out using phenol: chloroform and analysed by PCR using specific primers covering RRE1 and RRE 2 in the full length reporter vector. The DNA was quantified by SYBR green real-time PCR. For calculation of the relative DNA amount from quantitative real-time PCR, the Ct (threshold cycle) value of antibody-precipitated sample was normalized by the Ct value of the input, and the normalized Ct value of induced and uninduced groups were compared.

PCR amplification of viral progeny DNA
To investigate whether Bcl-2 upregulation can increase the production of viral progeny, the KSHV encoded ORF9 was amplified by PCR using viral DNA in TPA and butyrate induced BC3-shC and BC3-shBcl2 cells. Cells were induced and further incubated for 2 days at 37uC under 5% CO 2 . The supernatant was then collected and passed through a 0.45-mm-pore-size filter, and viral particles were spun down at 23,500 rpm for 20 min. Intact cells were discarded to prevent possible lysis and contamination from cellular viral DNA. The pellet was resuspended in 50 ml of 0.26 PBS, heated to 95uC for 15 min, and switched to 56uC for 1 h with proteinase K treatment (10 mg/ml). The enzyme was then destroyed by treatment at 95uC for 30 min. A 10-ml portion of virus lysate was used for PCR amplification of the KSHVspecific region of ORF9. The primers for amplification of ORF9 are listed in Table 1. KSHV genomic DNA isolated from BC3 cells served as control. The net density of the PCR product bands was quantitated using Kodak 1D 3.6 software (Eastman Kodak Company, Rochester, NY).

Apoptosis assay
Cellular apoptosis patterns were examined in induced and uninduced BC3-shBcl2, BJAB-shBcl2 and their control cell lines, BC3-shC and BJAB-shC using propidium iodide (PI) staining and flow ctyometry analysis using previously published protocols from our lab [47,52]. Briefly, 10 6 cells were collected, washed, fixed, then treated with PI staining and RNase A digestion, followed by examination of cellular apoptotic levels. Each sample was analyzed six times. Statistical analysis was carried out using one-way ANOVA (Analysis of Variance).