A nuclear role for Kaposi’s sarcoma-associated herpesvirus-encoded K13 protein in gene regulation
Kaposi’s sarcoma-associated herpesvirus (KSHV)-en- coded viral FLICE inhibitory protein K13 interacts with a cytosolic IjB kinase (IKK) complex to activate nuclear factor-jB (NF-jB). We recently reported that K13 antagonizes KSHV lytic regulator RTA (replication and transcription activator) and blocks lytic replication, but spares RTA-induced viral interleukin-6 (vIL6). Here we report that K13 is also present in the nuclear compart- ment, a property not shared by its structural homologs. K13interacts with and activates the nuclear IKK complex, and binds to the IjBa promoter. K13 mutants that are retained in the cytosol lack NF-jB activity. However, neither the IKKs nor NF-jB activation is required for nuclear localization of K13. Instead, this ability is dependent on a nuclear localization signal located in its N-terminal 40 amino acids. Finally, K13, along with p65/ RelA, binds to the promoters of a number of KSHV lytic genes, including RTA, ORF57 and vGPCR, but not to the promoter of the vIL6 gene. Thus, K13 has an unexpected nuclear role in viral and cellular gene regulation and its differential binding to the promoters of lytic genes may not only contribute to the inhibition of KSHV lytic replication, but may also account for the escape of vIL6 from K13-induced transcriptional suppression.
Keywords: K13; vFLIP; KSHV; RTA; IKK; NF-kB
Introduction
Kaposi’s sarcoma-associated herpesvirus (KSHV) has been linked to the development of Kaposi’s sarcoma, primary effusion lymphoma (PEL) and multicentric Castleman’s disease (Schulz, 2006). KSHV-encoded K13 protein contains two tandem death effector domains. Although originally classified as a viral FLICE inhibi- tory protein (vFLIP; Thome et al., 1997), K13 was subsequently shown to be a potent activator of the nuclear factor-kB (NF-kB) pathway (Chaudhary et al., 1999; Matta and Chaudhary, 2004). K13 activates the NF-kB pathway by activating a cytosolic IkB kinase (IKK) complex (Chaudhary et al., 1999; Liu et al., 2002; Field et al., 2003), which consists of two catalytic subunits, IKK1/IKKa and IKK2/IKKb, and a regula- tory subunit, NF-kB essential modulator (NEMO)/ IKKg. The activated IKK complex phosphorylates IkBa targeting it for ubiquitination and proteasomal degradation, which allows the released NF-kB subunits to enter the nucleus and activate expression of their target genes. The NF-kB pathway has been shown to involve in the various biological activities of K13, such as promotion of cellular survival, proliferation, trans- formation and cytokine secretion (An et al., 2003; Sun et al., 2003a, b, 2006; Guasparri et al., 2004; Chugh et al., 2005; Grossmann et al., 2006; Matta et al., 2007b).
We have recently demonstrated that K13 is a key regulator of KSHV lytic reactivation (Zhao et al., 2007). Ectopic expression of K13 can block lytic reactivation induced by treatment with 12-O-tetradecanoyl-phorbol- 13-acetate (TPA) or by expression of KSHV lytic regulator ORF50/RTA. This inhibitory effect of K13 on KSHV lytic reactivation is associated with a block in the expression of replication and transcription activator (RTA) and several downstream lytic genes, with the notable exception of viral interleukin-6 (vIL6; Zhao et al., 2007). However, it is not clear at the present how K13 blocks the expression of the lytic genes while sparing vIL6 expression.
In this report, we demonstrate that a significant fraction of K13 is present in the nuclear compartment. Nuclear accumulation of K13 is independent of its interaction with the IKK complex or its ability to activate the NF-kB pathway. Nuclear K13 gets differ- entially recruited to the promoters of KSHV lytic genes, which may provide a possible explanation for the suppression of KSHV lytic genes expression by K13 and escape of vIL6 from K13-induced transcriptional suppression.
Results
Nuclear localization of K13
To investigate the mechanism by which K13 blocks KSHV lytic replication, we used immunoblotting and examined the cellular distribution of K13 in three KSHV-infected PEL cell lines: BC-1, BC-3 and BCBL-1. The BC-1 cell line is dually infected with KSHV and the Epstein–Barr virus and expresses relatively high levels of endogenous K13 (Guasparri et al., 2004; Zhao et al., 2007). As shown in Figure 1a, we discovered that a significant fraction of K13 in the BC-1 cells was present in the nuclear compartment. Unlike BC-1 cells, the level of endogenous K13 in the BC-3 and BCBL-1 cell lines is relatively low and cannot be easily detected by immunoblotting. Therefore, we took advantage of polyclonal populations of these cells that had been engineered to stably express Flag-tagged K13 by retro- viral gene transfer (Matta et al., 2007a). Similar to the results with the BC-1 cells, a significant fraction of ectopically expressed Flag-tagged K13 was detected in the nuclear fraction in both BC-3 and BCBL-1 cell lines (Figure 1b). Finally, significant nuclear accumulation of ectopically expressed K13 was also observed in the Namalwa (human B-cell lymphoma) cell line that is not infected with KSHV, thereby demonstrating that this property is not limited to KSHV-infected cells and is independent of the activity of KSHV-encoded proteins.
K13 is unique among the vFLIPs in localizing into the nucleus
We next examined whether the ability to localize in the nucleus is unique to K13 or is shared by other vFLIPs. We constructed enhanced green fluorescent protein (EGFP) fusion constructs of K13 and the vFLIPs MC159 and MC160 of the Molluscum contagiosum virus and transfected them into 293FT cells. Although the wild-type EGFP, EGFP-MC159 and EGFP-MC160 demonstrated diffuse cellular fluorescence, a large percentage of EGFP-K13-expressing cells demonstrated strong nuclear fluorescence (Figure 2a). A luciferase reporter assay conducted in parallel demonstrated that K13 was also the only vFLIP that could activate the NF-kB pathway (Figure 2b). Collectively, these results demonstrate that K13 is unique among the vFLIPs in possessing the ability to accumulate in the nucleus and this property correlates with its ability to activate the NF-kB pathway.
PS-1145, an IKK inhibitor blocks K13-induced NF-kB but fails to block its nuclear accumulation K13 activates the NF-kB pathway by binding to and activating the IKK complex (Chaudhary et al., 1999; Liu et al., 2002; Field et al., 2003). We used PS-1145, a small molecule-specific inhibitor of the IKK activity (Hideshima et al., 2002), to delineate the relationship between the nuclear accumulation of K13 and its ability to activate the NF-kB pathway. 293FT cells were transfected with expression constructs encoding EGFP and EGFP-K13, along with reporter plasmids encoding NF-kB-Luc and pRSV/LacZ, and were subsequently treated with PS-1145. Treatment with PS-1145 resulted in a marked inhibition of K13-induced NF-kB activity but had no significant effect on its nuclear accumulation (Figures 2c and d). Thus, although the ability of K13 to activate the NF-kB pathway is closely linked to its ability to accumulate in the nucleus, this nuclear accumulation is not dependent on NF-kB activation.
Lack of requirement for IKK1, IKK2 and NEMO for nuclear localization of K13
Recent studies suggest that the IKK complex subunits shuttle between cytoplasm and nucleus (Anest et al., 2003; Yamamoto et al., 2003; Verma et al., 2004; Ear et al., 2005). Although our results with PS-1145 suggest that the kinase activity of the IKK complex is not required for nuclear accumulation of K13, they do not rule out the possibility that the interaction of K13 with the IKK complex subunits facilitates its nuclear accumulation. To examine this possibility, we stably expressed Flag-K13 in wild-type and NEMO-deficient Jurkat cells (Harhaj et al., 2000). Consistent with our published results (Liu et al., 2002), although K13 induced NF-kB activity in the wild-type Jurkat cells, it failed to do so in the NEMO-deficient cells (Figure 3a). However, more importantly, absence of NEMO had no significant effect on the nuclear accumulation of K13 (Figure 3b). The lack of requirement for the IKK subunits in the nuclear accumulation of K13 was further studied by using immortalized mouse embryonic fibro- blast cells (MEFs) that lack the expression of different IKK subunits and had been engineered to express K13 (Matta et al., 2003). Using both immunoblotting and immunofluorescence analysis, we observed nuclear accumulation of K13 in not only the wild-type MEFs but also in those deficient in IKK1, IKK2 or NEMO (Figures 3c and d). Although there was some difference in the absolute amount of nuclear K13 among different MEFs, this was probably due to a difference in the level of expression of the transduced protein in the different cells (Figure 3c). Together with the studies using NEMO-deficient Jurkat cells, the above results argue against the possibility that K13 accumulates in the nucleus by binding to the IKK complex subunits. Furthermore, as K13-induced NF-kB activation is defective in IKK1-, IKK2- and NEMO-deficient cells (Matta et al., 2003; Figure 3a), these results confirm that nuclear accumulation of K13 is not dependent on NF-kB activation.
Deletion mutagenesis analysis of K13 nuclear localization An analysis of K13 protein sequence did not reveal any classical nuclear localization signal (NLS; Lange et al., 2007). To determine the region of K13 that is essential for its nuclear localization, deletion mutants of K13 were constructed in the expression vector pEGFP-C2 and transfected into 293FT cells. A K13 mutant (K13(7– 188)), lacking the N-terminal six amino acids, accumu- lated in the nucleus and retained significant NF-kB reporter activity, whereas mutants K13(14–188), K13(22–188) that lacked the N-terminal 13 and 21 amino acids, respectively, lost NF-kB reporter activity and were retained in the cytosol (Figures 4a and b). A deletion mutant (K13D45–48) lacking the amino-acid residues 45–48 accumulated in the nucleus, albeit slowly, but lacked significant NF-kB reporter activity. More importantly, a short fragment of K13 (K13(1–40)), containing its N-terminal 40 amino acids, which lacked NF-kB activity, successfully targeted EGFP to the nuclear compartment (Figures 4a and b), thereby demonstrating the presence of a functional NLS in this region.
We also constructed C-terminal deletion mutants of K13. A K13 mutant (K13(1–181)) lacking the C-terminal seven amino acids retained significant NF- kB reporter activity and accumulated in the nucleus, whereas mutants lacking the C-terminal 18 (K13(1– 170)), 25 (K13(1–163)) and 33 (K13(1–155)) residues, respectively, showed progressive loss of NF-kB reporter activity and nuclear accumulation (Figures 4a and b). Thus, the mutant K13(1–155), which showed a complete loss of NF-kB activity, was entirely retained in the cytosol (Figures 4a and b). However, an EGFP fusion construct containing the C-terminal 33 residues (K13(156–188)) failed to accumulate in the nucleus (Figure 4a), thereby suggesting that this region, in itself, does not contain a functional NLS.
Finally, we generated EGFP fusion constructs of two previously described NF-kB-defective point mutants of K13, 67AAA (D67A/L68A/L70A) and 58AAA (E58A/C59A/L60A) (Sun et al., 2003b). As shown in Figures 4a and b, the 58AAA mutant, which demonstrated a complete loss of NF-kB activity, was completely excluded from the nucleus, whereas the 67AAA mutant, which retained some residual NF-kB activity, showed partial nuclear accumulation. Collectively, the above results demonstrate a good correlation between the cytosolic retention of K13 mutants and their loss of NF- kB activity as none of the K13 mutants that were retained in the cytosol could activate the NF-kB pathway (Figure 4c).
K13 interacts with and activates IKK complex in the nucleus and gets recruited to the IkBa promoter Although it is generallybelieved that the major signaling events during NF-kB activation occur in the cytosol, recent studies suggest that IkBa and IKK subunits also enter the nucleus where they regulate key steps of NF- kB activation (Gloire et al., 2006). To understand the mechanism by which nuclear K13 could contribute to NF-kB activation, we used K562 cells that ectopically express Flag-tagged K13 and show increased NF-kB DNA-binding activity (Figure 5a). Expression of K13 was readilydetected in the nuclear compartment in these cells and was associated with increased nuclear accu- mulation of NEMO, IKK1 and IKK2 (Figure 5b). K13 is known to activate the cytosolic IKK complex by binding to NEMO (Chaudhary et al., 1999; Liu et al., 2002; Field et al., 2003; Matta et al., 2003, 2007a). Interestingly, we discovered that K13 interacts with NEMO not only in the cytosol but also in the nucleus (Figure 5c), and results in the phosphorylation of both the cytosolic and the nuclear IkBa (Figure 5b). To confirm and extend the above studies, we took advantage of BCBL-1 cells that stably express a K13- ERTAM fusion protein (Matta et al., 2007a), which allows conditional activation of K13 activity upon treatment with 4-hydroxytamoxifen (4OHT). Consistent with our published results, 4OHT strongly induced NF-kB activity in K13-ERTAM cells (Figure 5d), which was associated with increased phosphorylation of the cytosolic IKK1/2 and IkBa (Figure 5e). More importantly, 4OHT treatment also resulted in the phosphorylation of the nuclear IKK1/2 and the nuclear IkBa, which were accompanied by increased nuclear accumulation of p65/RelA (Figure 5e).
It was recently demonstrated that IKK1 could modulate the transcriptional activity of NF-kB-respon- sive genes, such as IkBa and IL6, by binding to their promoters (Anest et al., 2003; Yamamoto et al., 2003). To test the possibility that K13 might similarly bind to the promoters of NF-kB-responsive genes, we per- formed chromatin immunoprecipitation assays (ChIP) and detected specific binding of K13 to the IkBa promoter in the K562 and BCBL-1 cells ectopically expressing Flag-tagged K13 and in BC-1 cells with endogenous K13 expression (Figures 5f–h). Collectively, these results suggest that in addition to its known role in the activation of the IKK complex in the cytosol, K13 also has nuclear roles in NF-kB activation, which could potentially explain the loss of NF-kB activity in K13 mutants that are retained in the cytosol.
K13 gets differentially recruited to the promoters of KSHV lytic genes
We previously reported that stimulation of K13 activity by 4OHT treatment in BCBL-1-TREx-RTA cells expressing K13-ERTAM could block TPA- and RTA- induced KSHV lytic replication (Zhao et al., 2007). Our results showing that K13 is also present in the nuclear compartment supported the intriguing possibility that nuclear localization of K13 might be involved in blocking the expression of KSHV lytic genes. To test this possibility, we first confirmed and extended our previous studies and demonstrated that stimulation of K13 activity by 4OHT in the K13-ERTAM-expressing BCBL-1-TREx-RTA cells could effectively block ex- pression of the ORF50/RTA, ORF57 and ORF74/ vGPCR transcripts following treatment with TPA, but had no significant impact on the expression of vIL6 (Figures 6a–d). We next performed ChIP assays with control and Flag antibodies on the vector- and Flag- tagged K13-expressing BCBL-1 cells and examined the recruitment of K13 to the promoters of these lytic genes using real-time PCR analysis. Using ChIP with the Flag antibody, K13 was readily detected on the ORF50/RTA, ORF57 and ORF74/vGPCR promoters in K13-expres- sing BCBL-1 cells (Figures 6e–g). However, binding of K13 was not detected on the vIL6 promoter (Figure 6h). Essentially identical results were obtained when the ChIP assay was repeated using a monoclonal antibody (8F6) against the full-length K13 protein (Figures 6e–h). No recruitment of K13 was detected on the viral ORF11 gene (Figure 6i), the cellular G3PDH gene (Figure 6j), or when the immunoprecipitation was repeated with control mouse immunoglobulin G (IgG; Figures 6e–j). Recruitment of K13 to the lytic promoters, however, was not limited to cells with ectopic K13 expression. A ChIP assay with the 8F6 antibody revealed significant recruitment of endogenous K13 to the ORF50/RTA and ORF57 promoters in the BC-1 cells (Figures 6k and l).
The results of the ChIP assays were confirmed by gel electrophoresis of the real-time PCR products (Supplementary Figures 1a and b). Collectively, the above results demonstrate differential recruitment of K13 to the promoters of KSHV lytic genes, which likely accounts for their differential suppression by K13.
Differential recruitment of p65 to lytic gene promoters in K13-expressing cells
Overexpression of p65/RelA subunit of NF-kB has been previouslyshown to block the stimulatoryeffect of RTA on the promoters of a number of KHSV lytic genes, including the RTA/ORF50 and ORF57 promoters (Brown et al., 2003). Therefore, we next examined whether differential recruitment of K13 to the promo- ters of lytic genes is also associated with differential recruitment of the p65/RelA to these promoters. As shown in Figures 7a and b, a ChIP assay using a p65 antibody revealed specific recruitment of p65/RelA to the promoters of ORF57, and vGPCR genes. A modest recruitment of p65/RelA to the ORF50/RTA promoter was observed as well (Figure 7c). However, similar to K13, no recruitment of p65 to the vIL6 promoter was observed (Figure 7d). The results of the ChIP assays were confirmed by gel electrophoresis of the real-time PCR products (Supplementary Figure 1c). Collectively, the above results suggest that differential recruitment of K13 and p65/RelA to the promoters of different lytic genes likely accounts for their differential inhibition by K13.
Discussion
On the basis of its homology to the prodomain of procaspase 8/FLICE, K13 was originally described as a vFLIP that protected virally infected cells against death receptor-induced apoptosis (Thome et al., 1997). Sub- sequent studies identified K13 as a potent activator of the NF-kB pathway through its ability to interact with and activate the cytosolic IKK complex (Chaudhary et al., 1999; Liu et al., 2002; Field et al., 2003; Matta et al., 2003; Chugh et al., 2005). In this report, we present evidence that K13 is also present in the nuclear compartment and its nuclear accumulation likely has an important role in viral and cellular gene regulation.
We have demonstrated the presence of a functional NLS in the N-terminal 40 amino acids of K13. Our results with the K13(7–188) and the K13(22–188) mutants further suggest that this NLS is most likely located in the region spanning amino acids 7–21. This region does not bear strong resemblance to the classical basic NLS described in the literature (Gorlich and Kutay, 1999). However, it is important to point out that the characterization of NLSs is still in its infancy and there is increasing appreciation that import signals unrelated to the classical NLS do exist (Gorlich and Kutay, 1999). Additional deletion and point mutagenesis studies are currently in progress to identify
critical amino acids within this region that are respon- sible for K13 nuclear transport. Finally, although the deletion mutagenesis studies suggested the presence of an NLS in the C-terminal 33 amino acids of K13, we were unable to delineate a functional NLS in this region either by sequence homology analysis or by fusing this region to EGFP in the mutant K13(156–188). There are several possible explanations for these results, including the possibility that the putative C-terminal NLS requires intra- or intermolecular interactions to be functional and such interactions are absent in the EGFP-K13(156–188) construct used in our study. It is also conceivable that the C-terminal region of K13 serves to unmask the N-terminal NLS but lacks a functional NLS of its own.
According to the classical view of NF-kB activation, IkBa is the cytoplasmic inhibitor of NF-kB, which binds to and retains the NF-kB complexes in an inactive state in the cytoplasm (Hayden and Ghosh, 2004). Signal- induced phosphorylation of IkBa by the cytosolic IKK complex is believed to lead to its ubiquitination and subsequent proteasomal degradation, allowing the nuclear import of NF-kB. However, as discussed by Ghosh and Karin (2002), several lines of evidence question the accuracy and generality of this simple model and suggest that IkBa also participates in the inhibition of NF-kB-dependent transcription in the cell nucleus. First, one of the first genes induced following NF-kB activation is IkBa itself, and newly synthesized IkBa is believed to rapidly enter the nucleus and dissociate promoter-bound NF-kB complexes (Baeuerle and Baltimore, 1988; Tran et al., 1997). Second, b-TrCP, the receptor that targets the phosphorylated IkBa for ubiquitination, is exclusively localized in the nucleus (Davis et al., 2002), which argues against cytosol as the primary site of IkBa degradation. Finally, the IKK subunits are either constitutively expressed in the nucleus or enter the nucleus following NF-kB-activating stimuli and participate in different steps of NF-kB activation, including transcriptional activation of NF- kB-responsive genes (Anest et al., 2003; Yamamoto et al., 2003; Ear et al., 2005).
Our results showing loss of NF-kB activity in the K13 mutants that are retained in the cytosol are consistent with these newly appreciated nuclear roles of IKK complex and IkBa in NF-kB activation and support the following speculative model of K13-induced activation of this pathway. Interaction of K13 with the cytosolic IKK complex not only results in the activation of the latter but also induces its nuclear accumulation. K13 may induce nuclear accumulation of the IKK complex by binding to it in the cytosol and carrying it into the nucleus. Alternatively, nuclear K13 may bind to the IKK complex subunits that are constantly shuttling between the nucleus and the cytosol and blocks their cytosolic export. In either case, interaction of K13 with the nuclear IKK complex keeps the latter in an activated state in the nucleus. In turn, the activated nuclear IKK complex leads to the phosphorylation of nuclear IkBa, resulting in its ubiquitination by nuclear b-TrCP and subsequent degradation by the nuclear proteasome, as has been suggested previously (Renard et al., 2000). Degradation of nuclear IkBa releases the nuclear NF-kB complexes from its inhibitory influences, allowing the latter to bind to their target promoters, while also preventing their export to the cytoplasm. The transcrip- tional activity of the promoter-bound NF-kB complexes is perhaps further modulated by the recruitment of K13 to these promoters. Thus, activation of the cytosolic IKK complex maynot be sufficient for sustained NF-kB activation by K13. Instead, a number of key regulatory steps in this process might take place in the nucleus and require nuclear accumulation of K13. As the NF-kB pathway is essential for keeping KSHV in a latent state, and for the survival of KSHV-infected cells, nuclear accumulation of K13 may be a clever strategy adopted by the virus to ensure the continuous presence of an active IKK complex in the nucleus that can guard against inadvertent interruption of this essential path- wayby nuclear accumulation of newlysynthesized IkBa. Although the above model explains a number of observations of the present study, it must be clarified, however, that alternative explanations do exist for the loss of NF-kB activity by K13 mutants that are retained in the cytosol. For example, a property, such as posttranslational modification or protein–protein inter- action, may be critical for both the nuclear accumula- tion of K13 and its ability to activate NF-kB, and the loss of this property may account for the close correlation between the two activities observed among the different K13 mutants.
We have recently shown that ectopic overexpression of K13 in BCBL-1 and JSC-1 cells blocks TPA-induced lytic replication of KSHV, which is associated with a block in the expression of RTA (Zhao et al., 2007). However, although RTA is the master regulator of KSHV lytic replication, and its ectopic expression is sufficient to induce a complete lytic replication (Nakamura et al., 2003), inhibition of RTA expression, in itself, is perhaps insufficient to account for the inhibitory effect of K13 on lytic replication. This notion is supported by our previous results showing that K13 could effectively block KSHV lytic replication in BCBL1-RTA-TREx cells even when the lytic replication was induced by doxycycline-induced expression of the RTA gene from a tetracycline-responsive promoter, and has led to the suggestion that suppression of transcrip- tional activation of downstream targets of RTA probably contributes to this effect (Zhao et al., 2007). Consistent with this notion, in this study, we present evidence that K13 effectively suppresses the expression of the ORF57 gene, which is not only a downstream target of RTA, but also a key posttranscriptional regulator of lytic gene expression (Kirshner et al., 2000; Han and Swaminathan, 2006; Nekorchuk et al., 2007). Furthermore, we have observed that K13 gets recruited to the promoters of the ORF50/RTA, ORF57 and ORF74/vGPCR genes, but is not recruited to the vIL6 promoter. In addition to K13, we have observed recruitment of p65/RelA to the promoters of the ORF57, ORF74/vGPCR and ORF50/RTA genes, and its lack of recruitment to the vIL6 promoter, in K13- expressing cells. Relevant to this discussion, Brown et al. (2003) have previously shown that p65/RelA can block KSHV lytic replication by blocking RTA-induced transcriptional activation of several lytic gene promo- ters, including the promoters of the ORF50/RTA and ORF57 genes. Therefore, taken together with the study by Brown et al., our results support the model that K13 blocks the expression of the lytic genes, such as ORF50/ RTA, ORF57 and ORF74/vGPCR, by helping to recruit p65/RelA to their promoters or by stabilizing the promoter-bound p65/RelA. In turn, the K13-p65/RelA complex may block the recruitment of RTA to these promoters or block the transcriptional activity of promoter-bound RTA or both. Along the same lines, lack of K13 and p65/RelA recruitment to the vIL6 promoter mayaccount for the escape of vIL6 from K13- induced transcriptional suppression.
It is important to emphasize that our model does not exclude other roles of K13 and NF-kB activation in promoting KSHV latency. In fact, a number of cellular proteins, such as RBP-Jk, AP1, SP1, EBPa and Oct1, are known to regulate lytic reactivation of KSHV and it is conceivable that interference with one or more of these proteins may also contribute to the inhibitory effect of K13 on KSHV lytic reactivation (West and Wood, 2003; Wang et al., 2004). A related question that remains to be answered is whether nuclear K13 can directly block lytic replication independent of NF-kB activation. We have tried genetic (that is, super- repressor form of IkBa) and pharmacological inhibitors of the NF-kB pathway to separate the direct vs indirect (that is, by NF-kB activation) effects of K13 on lytic replication. Unfortunately, the toxicity and lack of complete efficiency of these approaches has so far prevented us from obtaining definitive results (data not shown). To distinguish between these possibilities, we also attempted to isolate K13 mutants that are retained in the cytoplasm but still retain full NF-kB activity, but without success. We have also observed that the K13D45–48 mutant regains some NF-kB activity when stably expressed in the PEL cells (our unpublished observation), which makes it unsuitable for these studies. Thus, based on the available evidence, we are unable to delineate the individual contributions of K13 and NF-kB to the suppression of lytic genes at this time and settlement of this issue awaits the isolation of additional K13 mutants with selective loss of nuclear localization or NF-kB activities. The possi- bility does exist, however, that both the direct and indirect activities of K13 contribute to inhibition of KSHV lytic replication.
Materials and methods
Cell lines and reagents
293FT cells were obtained from Invitrogen (Carlsbad, CA, USA). PEL cell lines, Jurkat and Namalwa cells were obtained from the American Type Culture Collection. Polyclonal populations of these cells expressing Flag-K13 and K13- ERTAM-Flag have been described previously (Zhao et al., 2007; Matta et al., 2007a) or were generated by lentiviral-mediated gene transfer. K13-expressing MEFs have been described previously (Matta et al., 2003). NEMO-deficient Jurkat cells (Harhaj et al., 2000) were obtained from Dr Shao-Cong Sun (Penn State University). Polyclonal populations of these cells expressing an empty vector and Flag-tagged K13 were generated by retroviral-mediated gene transfer. Rabbit poly- clonal antibodies against control mouse IgG, control rabbit IgG, p65 and Sp1 were obtained from Santa Cruz Biotechnol- ogy (Santa Cruz, CA, USA). Antibodies against Flag and tubulin were from Sigma (St Louis, MO, USA). Phospho- specific antibodies were from Cell Signaling Technology (Danvers, MA, USA). A rat K13 monoclonal antibody was provided by Dr Mary Collins and a mouse monoclonal antibody against the full-length K13 protein (8F6) was generated in our laboratory. PS-1145 was purchased from Calbiochem (San Diego, CA, USA).
Plasmids and transfection
EGFP fusion constructs were generated by fusing the cDNAs of vFLIPs and any deletion and point mutants in frame to the C terminus of EGFP. 293FT and 293T cells were transfected with different EGFP constructs using calcium phosphate and examined under a fluorescence microscope 24 h posttransfection.
Luciferase reporter assays
The NF-kB reporter assay was performed essentially as described previously (Kumar et al., 2001). For details, please see Supplementary Information.
Immunofluorescence assay
MEFs were attached to poly-D-lysine (Sigma)-coated glass coverslips. The cells were fixed in 4% paraformaldehyde in phosphate-buffered saline, permeabilized with 0.2% Triton X-100 and incubated with a Flag monoclonal antibody (mAb; M2, Sigma) antibody (diluted to 1:1000) followed by Alexa Fluor 488 conjugated anti-mouse IgG (diluted to 1:200) (Molecular Probes, Eugene, OR, USA). Nuclei were stained with propidium iodide (PI) in mounting oil (Vectashield with PI, Vector Laboratories, Burlingame, CA, USA).
Cells were examined under a fluorescence microscope and photographed.
Chromatin immunoprecipitation assay and real-time PCR ChIP assays were performed as per the manufacturer’s instructions (Upstate Biotechnology, Lake Placid, NY, USA). For details, see Supplementary Information. Condi- tions for real-time PCR analysis have been described previously (Zhao et al., 2007).