Ubiquitylation of Cdk9 by Skp2 Facilitates Optimal
http://www.100md.com
病菌学杂志 2005年第17期
Rosalind Russell Medical Research Center, Departments of Medicine, Microbiology and Immunology, University of California at San Francisco, San Francisco, California 94143-0703
Institute of Biochemistry, Medical Faculty of the University of Ljubljana, Vrazov trg 2, 1000 Ljubljana, Slovenia
Department of Biology, Biotechnical Faculty, University of Ljubljana, Vecna pot 111, 1000 Ljubljana, Slovenia
ABSTRACT
By recruiting the positive transcriptional elongation factor b (P-TEFb) to paused RNA polymerase II, the transactivator Tat stimulates transcriptional elongation of the human immunodeficiency virus type 1 (HIV-1) genome. We found that cyclin-dependent kinase 9 (Cdk9), the catalytic subunit of P-TEFb, is ubiquitylated in vivo. This ubiquitylation depended on the Skp1/Cul1/F-box protein E3 ubiquitin ligase Skp2. Likewise, Tat required Skp2 since its transactivation of the HIV-1 long terminal repeat decreased in primary mouse embryonic fibroblasts, which lacked Skp2. The ubiquitylation of Cdk9 by Skp2 facilitated the formation of the ternary complex between P-TEFb, Tat, and transactivation response element. Thus, our findings underscore the requirement of ubiquitylation for the coactivator function in regulating HIV-1 transcriptional elongation.
INTRODUCTION
The transcription of eukaryotic genes by the multisubunit enzyme RNA polymerase II (RNAPII) is a remarkably complex process that requires the participation of a diverse collection of proteins, e.g., numerous DNA-binding transcriptional regulators, the Mediator complex, chromatin modifying enzymes, and a plethora of general initiation and elongation factors (19). Correspondingly, RNAPII transcription is regulated at multiple levels, including those of transcriptional elongation, of which human immunodeficiency virus type 1 (HIV-1) transcription is the paradigm (24). Soon after RNAPII initiates transcription on the HIV-1 long terminal repeat (LTR), it becomes arrested, yielding predominantly short viral transcripts. The C-terminal domain (CTD) of the largest subunit of RNAPII (Rpb1), which is composed of 52 repeats of the consensus heptapeptide Tyr-Ser-Pro-Thr-Ser-Pro-Ser, is hypophosphorylated, resulting in its association with a diverse collection of negative transcription elongation factors (N-TEFs), of which DRB sensitivity-inducing factor (DSIF) and negative elongation factor (NELF) are critical components (26, 29, 21). The transition from initiation to elongation of transcription is accompanied by massive phosphorylation of the CTD. Positive transcriptional elongation factor b (P-TEFb), which consists of cyclin T1 (CycT1) and the cyclin-dependent kinase 9 (Cdk9) (21), binds with Tat the transactivation response element (TAR) RNA that forms on a nascent viral transcripts, resulting in hyperphosphorylated CTD. In addition, P-TEFb phosphorylates Spt5 (10) and RD, which are part of DSIF and NELF, resulting in their dissociation from TAR (6). As a consequence, productive elongation of HIV-1 transcription ensues.
Considering the diversity of physiological signals that regulate gene expression, it is not surprising that transcriptional activators and coactivators themselves are subject to multiple modes of regulation. Common themes in their regulation include numerous posttranslational modifications, such as protein phosphorylation, acetylation, methylation, and ubiquitylation. Indeed, a number of recent reports indicate that the ubiquitin/proteasome system regulates RNAPII transcription (3, 17). Whereas in the classical ubiquitin-dependent proteolysis pathway, where ubiquitin molecules are repetitively conjugated to lysine residues in target proteins through a multienzyme cascade, thus licensing them for degradation by the 26S proteasome, the conjugation of ubiquitin also leads to many unanticipated biological consequences, including transcriptional activation. The latter function is carried out by a growing cast of E3 ubiquitin ligases, which together with E2 ubiquitin-conjugating enzymes and E1 ubiquitin-activating enzymes constitute the multienzyme ubiquitin-dependent proteolysis cascade (20). For example, whereas the yeast E3 enzyme Met30 mediates transcriptional activation by VP16 (22), c-Myc activity requires the mammalian E3 ubiquitin ligase Skp2 (14, 25). While the F-box proteins Met30 and Skp2 bind to their specific proteins, a degenerate, approximately 40-amino-acid sequence motif called the F-box binds the adaptor Skp1, which in turn recruits target proteins to a heterodimer between the Cullin protein family member Cul1 and the RING finger protein Rbx1 (Cul/Rbx1), resulting in their ubiquitylation. Since they retain the ability to bind their corresponding substrates but are no longer part of the Skp1/Cul1/Rbx1 protein complex, the F-box deleted mutant F-box proteins (F) act as dominant-negative proteins in vivo (25).
Recently, the ubiquitin/proteasome system has been implicated in the regulation of HIV-1 transcription. On one hand, the E3 ubiquitin ligase Hdm2 binds Tat, mediates its ubiquitylation, and acts as a positive cofactor for Tat transactivation of the HIV-1 promoter (1). On the other hand, Skp2 binds the CycT1 subunit of P-TEFb to promote the ubiquitylation of Cdk9 (13). In the present study, we investigated whether the ubiquitylation of P-TEFb itself would regulate positively HIV-1 transcription analogously to the ubiquitylation of Tat. We found that Cdk9 is ubiquitylated in vivo by the action of Skp2. Critically, this ubiquitylation increased transcriptional elongation of the HIV-1 genome due to the formation of a stronger ternary RNA-protein complex between P-TEFb, Tat, and TAR.
MATERIALS AND METHODS
Cell culture and cell lines. Monkey embryonic kidney COS and human embryonic kidney 293 cells were maintained in Dulbecco modified Eagle medium containing 10% fetal calf serum, 100 mM L-glutamine, and 50 μg each of penicillin and streptomycin per ml. B4 cells were a gift from Q. Zhou and were cultured as described previously (18). Primary mouse embryonic fibroblast (MEF) cells were a gift from K. Nakayama. All cells were grown at 37°C with 5% CO2.
Plasmid DNAs. Plasmid reporters pHIVSCAT, pSLIIBCAT, and pG6TAR were described previously (9, 16). Plasmids coding for HA.Cdk9, Tat, Rev.Tat, and Gal4.CycT1 were described previously (5, 16). The plasmids pcDNA3.1-Skp2, pcDNA3.1-Skp2F, and pcDNA3.1--TrCP were gift from C. Carrano and were described previously (2). The plasmid pcDNA3.1-Myc.Ub was a gift from S. C. Sun and was described previously (28). The plasmid pGEM7WT was described previously (5).
Chemicals and immunoreagents. MG-132 was obtained from Calbiochem (St. Louis, MO). Mouse monoclonal anti-Cdk9, anti-ubiquitin, rabbit polyclonal anti-HA, anti-Myc, anti-Cdk9, and goat polyclonal anti-actin antibodies were obtained from Santa Cruz Biotechnology (Santa Cruz, CA). The mouse monoclonal anti-FLAG M2 antibody was purchased from Sigma.
Transient transfection and CAT reporter gene assay. COS cells were seeded into 50- or 100-mm-diameter petri dishes approximately 12 h prior to transfection and transiently transfected with Lipofectamine according to the manufacturer's instructions (Gibco-BRL, Rockville, Md.). MEF cell lines were seeded into six-well plates approximately 24 h prior to transfection and transiently transfected with FuGENE6 reagent according to the manufacturer's instructions (Gibco-BRL). Chloramphenicol acetyltransferase (CAT) enzymatic assays were performed as described previously (5). In all transfections, the amount of DNA was equilibrated with a corresponding empty vector. A cytomegalovirus--galactosidase plasmid reporter (Gibco-BRL) was used to monitor transfection efficiency. The activity of the reporter plasmid alone is given as 1. The data are representative of three independent transfections, which were performed in duplicates. Error bars give the standard errors of the mean.
in vivo ubiquitylation assay, immunoprecipitation assay, and Western blot analysis. 293 and B4 cells were mock treated or treated for 4 h with MG-132 (50 μM) and then lysed in 0.4 ml of lysis buffer A (1% [vol/vol] NP-40, 10 mM Tris-HCl [pH 7.4], 150 mM NaCl, 2 mM EDTA, 0.1% protease inhibitor) for 1 h at 4°C. Total cell lysates were examined for the presence of hemagglutinin (HA) epitope-tagged Cdk9 species by Western blotting with the rabbit polyclonal anti-HA antibody. COS cells were transfected with indicated plasmid vectors. About 36 h posttransfection, the cells were mock treated or treated for 4 h with MG-132 (50 μM) and then lysed in 0.8 ml of lysis buffer A for 1 h at 4°C. Total cell lysates were immunoprecipitated with the rabbit polyclonal anti-HA antibody and examined for the presence of ubiquitin conjugates by Western blotting with the rabbit polyclonal anti-Myc antibody. MEFs were treated for 2 h with MG-132 (50 μM) and then lysed in 0.4 ml of lysis buffer A for 1 h at 4°C. Total cell lysates were immunoprecipitated with the rabbit polyclonal anti-Cdk9 antibody and examined for the presence of ubiquitin conjugates [(Ub)n Conj.] by Western blotting with the mouse monoclonal anti-ubiquitin antibody. An immunoprecipitation assay was performed as described previously (12). Western blotting was performed according to standard protocols.
Preparation of TAR RNA and RNA capture assay. -32P-labeled TAR RNA was prepared from the pGEM7WT plasmid by transcribing the linearized plasmid template (HindIII) with T7 RNA polymerase in the presence of [-32P]UTP by using the MEGAshortscript T7 kit (Ambion, Austin, Tex.). TAR RNA was excised from a 5% denaturing polyacrylamide gel, eluted (1.95 M ammonium acetate, 1% sodium dodecyl sulfate, 60 μg of tRNA/ml) for 6 h at 60°C, purified by using a phenol-chloroform extraction, and precipitated with 2.5 volumes of ethanol. Prior to use, the RNA pellet was dissolved in RNA capture buffer (50 mM HEPES-KOH [pH 7.6], 50 mM KCl, 0.1 mM EDTA, 5% glycerol, 7 mM MgCl2, 10 mM dithiothreitol). Total cell lysates were prepared from MEF cells, which expressed transiently the FLAG epitope-tagged human CycT1 protein, by using the lysis buffer A. P-TEFb complexes were immunoprecipitated onto protein A-Sepharose beads (Pharmacia, Piscataway, N.J.) with the mouse monoclonal anti-FLAG M2 antibody, washed three times in the wash buffer (50 mM Tris-HCl [pH 7.4], 0.25 M NaCl, 0.1% Triton X-100, 5 mM EDTA, 10 mM dithiothreitol, 1 mM sodium metabisulfite, 0.2 mM phenylmethylsulfonyl fluoride) and two times in RNA capture buffer before resuspension in 300 μl of RNA capture buffer. RNA capture assays were performed in the presence or absence of recombinant GST-Tat protein (500 ng), -32P-labeled TAR RNA and tRNA (10 ng/ml). Binding reaction mixtures were incubated at room temperature for 1 h with gentle rotation and then washed extensively with RNA capture buffer. Capture TAR RNA from the resulting RNA-protein complexes was extracted with TRIzol reagent (Gibco-BRL, Rockville, Md.), concentrated by precipitation (0.3 M sodium acetate, 20 μg of tRNA, 2.5 volumes of ethanol), resolved on 10% denaturing polyacrylamide gel, and analyzed by autoradiography.
RESULTS
Cdk9 is ubiquitylated in vivo. To determine whether Cdk9 is ubiquitylated, we performed an in vivo ubiquitylation assay (Fig. 1). Briefly, we coexpressed the hybrid HA.Cdk9 protein (Cdk9) with a Myc epitope-tagged ubiquitin or an empty plasmid effector pcDNA3.1 in COS cells and incubated these cells in the absence or presence of the proteasomal inhibitor MG-132. After isolation of Cdk9 by immunoprecipitation, the polyubiquitylated Cdk9 protein was detected by Western blotting with the anti-Myc antibody. In the absence of MG-132, we detected a light smear of high-molecular-weight ubiquitin conjugates in Cdk9 immunoprecipitations (Fig. 1, lane 1). Importantly, whereas in the presence of MG-132, this smear increased significantly, it was absent in cells where Cdk9 was not expressed (Fig. 1, compare lanes 2 and 3). In contrast, we found that the hybrid CycT1.HA protein was not ubiquitylated in vivo (data not presented). Overall, in agreement with the findings of Kiernan et al. (13), these data demonstrate that Cdk9 but not CycT1 is ubiquitylated in vivo.
Skp2 promotes ubiquitylation of Cdk9. To identify the E3 ubiquitin ligase that targets Cdk9, we developed a ubiquitylation assay where we examined a smear of high-molecular-weight ubiquitin conjugates of Cdk9 (Fig. 2). Briefly, we used B4 cells, which are human embryonal kidney 293 cells that express the HA epitope-tagged Cdk9 protein (Cdk9) and the parental 293 cells as controls. Cells were untreated or treated with MG-132, and cell lysates were examined for the presence of Cdk9 by Western blotting with the anti-HA antibody. We detected Cdk9 in B4 but not in parental 293 cells (Fig. 2a, compare lanes 1 and 2 with lanes 3 to 6). Since Cdk9 is ubiquitylated in vivo (Fig. 1), we also observed a smear of high-molecular-weight ubiquitin.Cdk9 conjugates in the presence of MG-132 only in cell lysates prepared from B4 but not in lysates prepared from parental 293 cells (Fig. 2a, compare lanes 2 and 4). Therefore, we expressed a number of wild-type and mutant F proteins in B4 cells and examined their effects on the ubiquitylation of Cdk9. The expression of Skp2 increased amounts of high-molecular-weight ubiquitin.Cdk9 conjugates significantly in MG-132-treated B4 cells. In contrast, the expression of the mutant dominant-negative Skp2F protein resulted in the loss of this ubiquitylation (Fig. 2a, compare lane 4 with lanes 5 and 6). Since the expression of another F-box protein, -TrCP, had no effect on Cdk9 ubiquitylation (data not presented), these effects were specific. Thus, our analysis identified Skp2 as the E3 ubiquitin ligase for Cdk9.
Next, we wanted to demonstrate the requirement of Skp2 for Cdk9 ubiquitylation in an endogenous system. We took advantage of primary MEFs which do or do not express Skp2 (Skp2+/+ and Skp2–/– MEFs, respectively) and performed in vivo ubiquitylation assays in these cells (Fig. 2b). Briefly, we treated cells with MG-132, prepared total cell lysates, immunoprecipitated the endogenous Cdk9 protein, and detected the polyubiquitylated Cdk9 protein by Western blotting with the anti-ubiquitin antibody. In the presence of Skp2, we detected the polyubiquitylated Cdk9 protein. On the contrary, this ubiquitin-Cdk9 (Ub.Cdk9) conjugate was not detected in the absence of Skp2 (Fig. 2b, upper panel, compare lanes 1 and 2). Moreover, while we observed slightly lower levels of Cdk9 in total cell lysates prepared from Skp2+/+ MEF cells compared to Skp2–/– MEF cells, those of actin remained unaffected (Fig. 2b, lower panel, compare lanes 1 and 2). Taken together, these studies confirmed that Skp2 is required for the ubiquitylation of Cdk9.
Skp2 facilitates Tat transactivation. To determine the importance and specificity of Skp2-mediated Cdk9 ubiquitylation for transcriptional activation, we tested its effects on Tat transactivation (Fig. 3). First, we expressed the plasmid reporter pHIVSCAT, together with Tat and a plasmid effector for Skp2 or -TrCP in COS cells, and measured levels of Tat transactivation with CAT reporter assays. Tat activated the pHIVSCAT 26-fold. The expression of Skp2 increased this activation to 54-fold. In contrast, when we expressed -TrCP instead of Skp2, the activity of pHIVSCAT remained unaffected (Fig. 3a). We conclude that Skp2 increases Tat transactivation specifically. Moreover, since the expression of Skp2 enhanced this transcription, Skp2 may play a positive role in Tat transactivation.
To address this stimulatory role of Skp2 for Tat transactivation, we next coexpressed the mutant dominant-negative Skp2F protein, Tat, and pHIVSCAT (Fig. 3b). Increasing amounts of the mutant Skp2F protein resulted in a dose-dependent repression of Tat transactivation (Fig. 3b, compare bar 2 with bars 3 and 4). Importantly, when we coexpressed the highest amount of the mutant Skp2F protein with increasing amounts of the wild-type Skp2 protein, we restored Tat transactivation to normal levels (Fig. 3b, compare bars 3 and 4 with bars 5 and 6). These data confirmed that Skp2 plays a positive role in Tat transactivation.
To explore the specificity of the inhibitory effect of the mutant Skp2F protein on Tat transactivation, we mimicked the natural recruitment of P-TEFb to the HIV-1 LTR by using two distinct artificial tethering systems of P-TEFb to DNA or RNA elements in the HIV-1 LTR and tested the effects of the mutant Skp2F protein on transcriptional activation by P-TEFb in these systems (Fig. 3c and d). The DNA-tethering system consists of the hybrid Gal4.CycT1 protein and the pG6TAR reporter plasmid, containing six high-affinity Gal4 DNA-binding sites in the HIV-1 LTR, whereas the RNA-tethering system takes advantage of the binding between the hybrid Rev.Tat protein and the Rev response element (RRE), which was fused to the stem of TAR, thus creating the pSLIIBCAT reporter plasmid (12, 24). In sharp contrast to Tat transactivation, neither the activation of the pG6TAR reporter plasmid by the DNA-tethered hybrid Gal4.CycT1 protein, nor the activation of pSLIIBCAT reporter plasmid by the RNA-tethered hybrid Rev.Tat protein was affected by the mutant Skp2F protein (Fig. 3c and d, compare bars 2 with 3). Thus, effects observed in Tat transactivation were specific for TAR.
Next, we wanted to confirm the above findings genetically in an endogenous system. For these experiments, we used wild-type and Skp2-deficient MEF cell lines (Skp2+/+ and Skp2–/– MEFs, respectively; Fig. 4). We coexpressed pHIVSCAT and Tat alone or in the presence of a human FLAG epitope-tagged CycT1 protein in these cells and measured CAT activity. Whereas Tat activated the pHIVSCAT 7-fold in the Skp2+/+ MEF cells, this activation decreased 2.5-fold in the Skp2–/– MEF cells (Fig. 4a, compare bars 3 and 4). Furthermore, we used the artificial tethering systems of P-TEFb to DNA or RNA elements and tested the requirement of Skp2 for transcriptional activation by P-TEFb in these systems (Fig. 4b). Consistent with the findings presented in Fig. 3c and d, the loss of Skp2 did not result in a decreased activation of the pG6TAR and pSLIIBCAT reporter plasmids by the hybrid Gal4.CycT1 and Rev.Tat proteins, respectively, underscoring again the specificity of effects observed in Tat transactivation (Fig. 4b, compare bars 3 with 4 and 7 with 8). Thus, Skp2 facilitates optimal Tat transactivation. Since the transcriptional activation by the Rev.Tat chimera was unaffected by the absence of Skp2 and the ubiquitylation of Cdk9 depended on Skp2 (Fig. 2b), the ubiquitylation of Cdk9 could contribute to the formation of the ternary complex between P-TEFb, Tat, and TAR.
Skp2 facilitates the formation of the RNA-protein complex between P-TEFb, Tat, and TAR. To examine the prediction that Skp2-mediated ubiquitylation of Cdk9 contributes to the formation of the ternary complex between P-TEFb, Tat, and TAR, we used an RNA capture assay (Fig. 5). Briefly, we isolated ubiquitylated and nonubiquitylated P-TEFb complexes from MG-132-treated Skp2+/+ and Skp2–/– MEF cells, respectively. These P-TEFb complexes were incubated with the GST.Tat chimera and -32P-labeled TAR, which was subjected to RNA capture assay. Since P-TEFb isolated from MEF Skp2+/+ cells captured TAR RNA six times more efficient compared to P-TEFb from MEF Skp2–/– cells, the capture of TAR depended on the presence of Skp2 (Fig. 5, compare lanes 3 and 4). Predictably, it also depended on the FLAG.CycT1 and GST.Tat chimeras (Fig. 5, compare lanes 1 to 3). Note that amounts of immunoprecipitated P-TEFb were equal in all reactions (Fig. 5, lower panel, lanes 2 to 4). Thus, ubiquitylation of Cdk9 by Skp2 facilitates the formation of the ternary complex between P-TEFb, Tat, and TAR.
DISCUSSION
In this study, we provide evidence that the ubiquitylation of Cdk9 by Skp2 facilitates Tat transactivation. First, by using several independent systems, we found that Cdk9 is ubiquitylated in vivo. Second, not only was Skp2 required for Cdk9 ubiquitylation but it also contributed to optimal levels of Tat transactivation. Finally, the ubiquitylation of Cdk9 by Skp2 facilitated the formation of the ternary complex between P-TEFb, Tat, and TAR. Thus, Tat requires ubiquitylated Cdk9 to activate optimally the transcriptional elongation of HIV-1 genes.
These findings extend recent observations for the ubiquitin/proteosome system in the regulation of HIV-1 gene expression. Whereas the Hdm2 E3 ubiquitin ligase activity also facilitates Tat transactivation (1), the role of Skp2 had not been addressed in this system. Rather, a negative role of Skp2 had been implicated in the function of the major histocompatibility complex class II transactivator (CIITA), which also depends on P-TEFb (13). Therein, the expression of CycT1 which lacks the PEST sequence at its C terminus (CycT1PEST) that binds Skp2 and mediates ubiquitylation of Cdk9, increased transcription activated by CIITA. In contrast, our biochemical and genetic evidence suggests that Skp2 acts as a coactivator for Tat transactivation. Indeed, the expression of the mutant Skp2F protein or the absence of Skp2 in the MEF Skp2–/– cells resulted in a loss of Cdk9 ubiquitylation. As a consequence, levels of Tat transactivation decreased. Consistently, the expression of the mutant CycT1PEST protein decreased this activity in COS cells (data not presented). Overall, these observations indicate that multiple E3 ubiquitin ligases regulate one transcription unit. Moreover, ubiquitylation of Cdk9 could affect transcriptional units differently, which could stem from their distinct mechanisms of transcriptional activation. Indeed, CIITA recruits P-TEFb to activate transcriptional elongation of its target genes via the MHCII enhanceosome assembled on the DNA, whereas Tat achieves the same task by recruiting P-TEFb to TAR RNA (12, 27).
It remains to be established what fraction of Cdk9 is ubiquitylated in cells. Although we found slightly increased steady-state levels of Cdk9 in total cell lysates prepared from MEF Skp2–/– cells compared to MEF Skp2+/+ cells, we did not observe any changes in levels of Cdk9 upon treatment of cells with the proteosomal inhibitor MG-132 or upon the expression of the Skp2F protein. Likewise, a recent report by Garriga et al. suggested that levels of Cdk9 protein do not depend on Skp2 (8). However, our functional data underscore the importance of Cdk9 ubiquitylation for the efficient formation of the ternary RNA-protein complex, which leads to optimal levels of Tat transactivation. In light of these observations, it is tempting to speculate that Cdk9 becomes ubiquitylated only when associated with the HIV-1 promoter or other cellular promoter elements. This scenario would resemble a mechanism of transcriptional activation by SREBP-1, c-Myc, and VP16, where the ubiquitylation of these activators occurs optimally on DNA (14, 15, 22, 23, 25). Indeed, E3 ubiquitin ligase Skp2 associates with the cyclin D2 promoter in a c-Myc-dependent manner in cells (25). It remains to be determined whether this situation holds true for the HIV-1 LTR as well.
How could the ubiquitylated Cdk9 affect Tat transactivation In principle, ubiquitin moieties on Cdk9 could influence protein-protein or RNA-protein interactions. The evidence obtained from tethering experiments in which Skp2 did not affect transcription activated by heterologously tethered P-TEFb via DNA or RNA targets suggests that ubiquitylation of Cdk9 regulates positively neither its association with CycT1 to increase P-TEFb kinase activity nor its binding to Tat to increase the levels of P-TEFb on the HIV-1 promoter. Rather, by using an RNA capture assay, we found that the ubiquitylated form of Cdk9 facilitates the overall assembly of the ternary RNA-protein complex, possibly by increasing direct binding to TAR or by facilitating the relief of autoinhibitory intramolecular interactions within CycT1 (see below).
The recruitment of P-TEFb to the paused RNAPII by Tat is critical for stimulating transcriptional elongation of HIV-1 genes and viral replication. The significance of this decisive event in the course of the HIV-1 replicative cycle is reflected in the existence of several mechanisms, which regulate the formation of the ternary complex between P-TEFb, Tat, and TAR. The following model emerges from numerous studies. First, an inhibitory intramolecular interaction between the N- and C-terminal regions of CycT1, which prevents the ternary complex assembly, is relieved by the binding of the transcription elongation factor Tat-SF1 to the C-terminal region in CycT1 (4). Second, Cdk9 phosphorylation is required for high-affinity binding of Tat and P-TEFb to TAR RNA (4, 7). In addition, the ubiquitylation of Cdk9 by Skp2 increases this RNA-protein complex assembly even further. Due to this sequence of events, high amounts of P-TEFb are concentrated in close proximity to its substrates, the CTD of RNAPII and the Spt5 subunit of DSIF. Moreover, P-TEFb targets the RD subunit of NELF, which dissociates from the lower stem of TAR RNA, thus contributing to the release of paused RNAPII (6). Finally, the acetylation at a single lysine residue in the TAR RNA-binding domain of Tat by the transcriptional coactivator p300 enhances Tat transactivation, possibly by dissociating CycT1 from TAR RNA and transferring Tat onto the elongating RNAPII (11). This model illustrates nicely how the assembly and disassembly of the multisubunit RNA-protein complex is a highly dynamic and temporally regulated process, which requires the participation of a variety of cellular cofactors, as well as multiple posttranslational modifications of its primary players.
ACKNOWLEDGMENTS
We thank Qiang Zhou, Kei-ichi Nakayama, Andrea C. Carrano, and Shao-Cong Sun for reagents and the members of Peterlin laboratory for useful discussions.
M.B. was supported in part by a fellowship from the Ministry of Science and Technology, Republic of Slovenia and by grant number 106584-36-RFNT from the American Foundation for AIDS Research. This study was supported by a grant from the National Institutes of Health (RO1 AI49104).
REFERENCES
Bres, V., R. E. Kiernan, L. K. Linares, C. Chable-Bessia, O. Plechakova, C. Treand, S. Emiliani, J. M. Peloponese, K. T. Jeang, O. Coux, M. Scheffner, and M. A. Benkirane. A non-proteolytic role for ubiquitin in Tat-mediated transactivation of the HIV-1 promoter. 2003. Nat. Cell Biol. 5:754-761.
Carrano, A. C., E. Eytan, A. Hershko, and M. Pagano. SKP2 is required for ubiquitin-mediated degradation of the CDK inhibitor p27. 1999. Nat. Cell Biol. 1:193-199.
Conaway, R. C., C. S. Brower, and J. W. Conaway. 2002. Emerging roles of ubiquitin in transcription regulation. Science 296:1254-1258.
Fong, Y. W., and Q. Zhou. 2000. Relief of two built-in autoinhibitory mechanisms in P-TEFb is required for assembly of a multicomponent transcription elongation complex at the human immunodeficiency virus type 1 promoter. Mol. Cell. Biol. 20:5897-5907.
Fujinaga, K., T. P. Cujec, J. Peng, J. Garriga, D. H. Price, X. Grana, and B. M. Peterlin. 1998. The ability of positive transcription elongation factor B to transactivate human immunodeficiency virus transcription depends on a functional kinase domain, cyclin T1, and Tat. J. Virol. 72:7154-7159.
Fujinaga, K., D. Irwin, Y. Huang, R. Taube, T. Kurosu, and B. M. Peterlin. 2004. Dynamics of human immunodeficiency virus transcription: P-TEFb phosphorylates RD and dissociates negative effectors from the transactivation response element. Mol. Cell. Biol. 24:787-795.
Garber, M. E., T. P. Mayall, E. M. Suess, J. Meisenhelder, N. E. Thompson, and K. A. Jones. 2000. CDK9 autophosphorylation regulates high-affinity binding of the human immunodeficiency virus type 1 tat-P-TEFb complex to TAR RNA. Mol. Cell. Biol. 20:6958-6969.
Garriga, J., S. Bhattacharya, J. Calbo, R. M. Marshall, M. Truongcao, D. D. Haines, and X. Grana. 2003. Cdk9 is constitutively expressed throughout the cell cycle, and its steady-state expression is independent of Skp2. Mol. Cell. Biol. 23:5165-5173.
Ghosh, S., M. J. Selby, and B. M. Peterlin. 1993. Synergism between Tat and VP16 in transactivation of HIV-1 LTR. J. Mol. Biol. 234:610-619.
Ivanov, D., Y. T. Kwak, J. Guo, and R. B. Gaynor. 2000. Domains in the SPT5 protein that modulate its transcriptional regulatory properties. Mol. Cell. Biol. 20:2970-2983.
Kaehlcke, K., A. Dorr, C. Hetzer-Egger, V. Kiermer, P. Henklein, M. Schnoelzer, E. Loret, P. A. Cole, E. Verdin, and M. Ott. 2003. Acetylation of Tat defines a cyclinT1-independent step in HIV transactivation. Mol. Cell 12:167-176.
Kanazawa, S., T. Okamoto, and B. M. Peterlin. 2000. Tat competes with CIITA for the binding to P-TEFb and blocks the expression of MHC class II genes in HIV infection. Immunity 12:61-70.
Kiernan, R. E., S. Emiliani, K. Nakayama, A. Castro, J. C. Labbe, T. Lorca, K. Nakayama, and M. Benkirane. 2001. Interaction between cyclin T1 and SCF(SKP2) targets CDK9 for ubiquitination and degradation by the proteasome. Mol. Cell. Biol. 21:7956-7970.
Kim, S. Y., A. Herbst, K. A. Tworkowski, S. E. Salghetti, and W. P. Tansey. 2003. Skp2 regulates Myc protein stability and activity. Mol. Cell 11:1177-1188.
Kurosu, T., and B. M. Peterlin. 2004. VP16 and ubiquitin: binding of P-TEFb via its activation domain and ubiquitin facilitates elongation of transcription of target genes. Curr. Biol. 14:1112-1116.
Madore, S. J., and B. R. Cullen. 1993. Genetic analysis of the cofactor requirement for human immunodeficiency virus type 1 Tat function. J. Virol. 67:3703-3711.
Muratani, M., and W. P. Tansey. 2003. How the ubiquitin-proteasome system controls transcription. Nat. Rev. Mol. Cell. Biol. 4:192-201.
O'Keeffe, B., Y. Fong, D. Chen, S. Zhou, and Q. Zhou. 2000. Requirement for a kinase-specific chaperone pathway in the production of a Cdk9/cyclin T1 heterodimer responsible for P-TEFb-mediated tat stimulation of HIV-1 transcription. J. Biol. Chem. 275:279-287.
Orphanides, G., and D. Reinberg. 2002. A unified theory of gene expression. Cell 108:439-451.
Pickart, C. M. 2001. Ubiquitin enters the new millennium. Mol. Cell 8:499-504.
Price, D. H. 2000. P-TEFb, a cyclin-dependent kinase controlling elongation by RNA polymerase II. Mol. Cell. Biol. 20:2629-2634.
Salghetti, S. E., A. A. Caudy, J. G. Chenoweth, and W. P. Tansey. 2001. Regulation of transcriptional activation domain function by ubiquitin. Science 293:1651-1653.
Sundqvist, A., and J. Ericsson. 2003. Transcription-dependent degradation controls the stability of the SREBP family of transcription factors. Proc. Natl. Acad. Sci. USA 100:13833-13838.
Taube, R., K. Fujinaga, J. Wimmer, M. Barboric, and B. M. Peterlin. 1999. Tat transactivation: a model for the regulation of eukaryotic transcriptional elongation. Virology 264:245-253.
von der Lehr, N., S. S. Johansson, Wu, F. Bahram, A. Castell, C. Cetinkaya, P. Hydbring, I. Weidung, K. Nakayama, K. I. Nakayama, O. Soderberg, T. K. Kerppola, and L. G. Larsson. 2003. The F-box protein Skp2 participates in c-Myc proteosomal degradation and acts as a cofactor for c-Myc-regulated transcription. Mol. Cell 11:1189-1200.
Wada, T., T. Takagi, Y. Yamaguchi, A. Ferdous, T. Imai, S. Hirose, S. Sugimoto, K. Yano, G. A. Hartzog, F. Winston, S. Buratowski, and H. Handa. 1998. DSIF, a novel transcription elongation factor that regulates RNA polymerase II processivity, is composed of human Spt4 and Spt5 homologs. Genes Dev. 12:343-356.
Wei, P., M. E. Garber, S. M. Fang, W. H. Fischer, and K. A. Jones. 1998. A novel CDK9-associated C-type cyclin interacts directly with HIV-1 Tat and mediates its high-affinity, loop-specific binding to TAR RNA. Cell 92:451-462.
Xiao, G., E. W. Harhaj, and S. C. Sun. 2001. NF-B-inducing kinase regulates the processing of NF-B2 p100. Mol. Cell 7:401-409.
Yamaguchi, Y., T. Takagi, T. Wada, K. Yano, A. Furuya, S. Sugimoto, J. Hasegawa, and H. Handa. 1999. NELF, a multisubunit complex containing RD, cooperates with DSIF to repress RNA polymerase II elongation. Cell 97:41-51.(Matjaz Barboric, Fan Zhan)
Institute of Biochemistry, Medical Faculty of the University of Ljubljana, Vrazov trg 2, 1000 Ljubljana, Slovenia
Department of Biology, Biotechnical Faculty, University of Ljubljana, Vecna pot 111, 1000 Ljubljana, Slovenia
ABSTRACT
By recruiting the positive transcriptional elongation factor b (P-TEFb) to paused RNA polymerase II, the transactivator Tat stimulates transcriptional elongation of the human immunodeficiency virus type 1 (HIV-1) genome. We found that cyclin-dependent kinase 9 (Cdk9), the catalytic subunit of P-TEFb, is ubiquitylated in vivo. This ubiquitylation depended on the Skp1/Cul1/F-box protein E3 ubiquitin ligase Skp2. Likewise, Tat required Skp2 since its transactivation of the HIV-1 long terminal repeat decreased in primary mouse embryonic fibroblasts, which lacked Skp2. The ubiquitylation of Cdk9 by Skp2 facilitated the formation of the ternary complex between P-TEFb, Tat, and transactivation response element. Thus, our findings underscore the requirement of ubiquitylation for the coactivator function in regulating HIV-1 transcriptional elongation.
INTRODUCTION
The transcription of eukaryotic genes by the multisubunit enzyme RNA polymerase II (RNAPII) is a remarkably complex process that requires the participation of a diverse collection of proteins, e.g., numerous DNA-binding transcriptional regulators, the Mediator complex, chromatin modifying enzymes, and a plethora of general initiation and elongation factors (19). Correspondingly, RNAPII transcription is regulated at multiple levels, including those of transcriptional elongation, of which human immunodeficiency virus type 1 (HIV-1) transcription is the paradigm (24). Soon after RNAPII initiates transcription on the HIV-1 long terminal repeat (LTR), it becomes arrested, yielding predominantly short viral transcripts. The C-terminal domain (CTD) of the largest subunit of RNAPII (Rpb1), which is composed of 52 repeats of the consensus heptapeptide Tyr-Ser-Pro-Thr-Ser-Pro-Ser, is hypophosphorylated, resulting in its association with a diverse collection of negative transcription elongation factors (N-TEFs), of which DRB sensitivity-inducing factor (DSIF) and negative elongation factor (NELF) are critical components (26, 29, 21). The transition from initiation to elongation of transcription is accompanied by massive phosphorylation of the CTD. Positive transcriptional elongation factor b (P-TEFb), which consists of cyclin T1 (CycT1) and the cyclin-dependent kinase 9 (Cdk9) (21), binds with Tat the transactivation response element (TAR) RNA that forms on a nascent viral transcripts, resulting in hyperphosphorylated CTD. In addition, P-TEFb phosphorylates Spt5 (10) and RD, which are part of DSIF and NELF, resulting in their dissociation from TAR (6). As a consequence, productive elongation of HIV-1 transcription ensues.
Considering the diversity of physiological signals that regulate gene expression, it is not surprising that transcriptional activators and coactivators themselves are subject to multiple modes of regulation. Common themes in their regulation include numerous posttranslational modifications, such as protein phosphorylation, acetylation, methylation, and ubiquitylation. Indeed, a number of recent reports indicate that the ubiquitin/proteasome system regulates RNAPII transcription (3, 17). Whereas in the classical ubiquitin-dependent proteolysis pathway, where ubiquitin molecules are repetitively conjugated to lysine residues in target proteins through a multienzyme cascade, thus licensing them for degradation by the 26S proteasome, the conjugation of ubiquitin also leads to many unanticipated biological consequences, including transcriptional activation. The latter function is carried out by a growing cast of E3 ubiquitin ligases, which together with E2 ubiquitin-conjugating enzymes and E1 ubiquitin-activating enzymes constitute the multienzyme ubiquitin-dependent proteolysis cascade (20). For example, whereas the yeast E3 enzyme Met30 mediates transcriptional activation by VP16 (22), c-Myc activity requires the mammalian E3 ubiquitin ligase Skp2 (14, 25). While the F-box proteins Met30 and Skp2 bind to their specific proteins, a degenerate, approximately 40-amino-acid sequence motif called the F-box binds the adaptor Skp1, which in turn recruits target proteins to a heterodimer between the Cullin protein family member Cul1 and the RING finger protein Rbx1 (Cul/Rbx1), resulting in their ubiquitylation. Since they retain the ability to bind their corresponding substrates but are no longer part of the Skp1/Cul1/Rbx1 protein complex, the F-box deleted mutant F-box proteins (F) act as dominant-negative proteins in vivo (25).
Recently, the ubiquitin/proteasome system has been implicated in the regulation of HIV-1 transcription. On one hand, the E3 ubiquitin ligase Hdm2 binds Tat, mediates its ubiquitylation, and acts as a positive cofactor for Tat transactivation of the HIV-1 promoter (1). On the other hand, Skp2 binds the CycT1 subunit of P-TEFb to promote the ubiquitylation of Cdk9 (13). In the present study, we investigated whether the ubiquitylation of P-TEFb itself would regulate positively HIV-1 transcription analogously to the ubiquitylation of Tat. We found that Cdk9 is ubiquitylated in vivo by the action of Skp2. Critically, this ubiquitylation increased transcriptional elongation of the HIV-1 genome due to the formation of a stronger ternary RNA-protein complex between P-TEFb, Tat, and TAR.
MATERIALS AND METHODS
Cell culture and cell lines. Monkey embryonic kidney COS and human embryonic kidney 293 cells were maintained in Dulbecco modified Eagle medium containing 10% fetal calf serum, 100 mM L-glutamine, and 50 μg each of penicillin and streptomycin per ml. B4 cells were a gift from Q. Zhou and were cultured as described previously (18). Primary mouse embryonic fibroblast (MEF) cells were a gift from K. Nakayama. All cells were grown at 37°C with 5% CO2.
Plasmid DNAs. Plasmid reporters pHIVSCAT, pSLIIBCAT, and pG6TAR were described previously (9, 16). Plasmids coding for HA.Cdk9, Tat, Rev.Tat, and Gal4.CycT1 were described previously (5, 16). The plasmids pcDNA3.1-Skp2, pcDNA3.1-Skp2F, and pcDNA3.1--TrCP were gift from C. Carrano and were described previously (2). The plasmid pcDNA3.1-Myc.Ub was a gift from S. C. Sun and was described previously (28). The plasmid pGEM7WT was described previously (5).
Chemicals and immunoreagents. MG-132 was obtained from Calbiochem (St. Louis, MO). Mouse monoclonal anti-Cdk9, anti-ubiquitin, rabbit polyclonal anti-HA, anti-Myc, anti-Cdk9, and goat polyclonal anti-actin antibodies were obtained from Santa Cruz Biotechnology (Santa Cruz, CA). The mouse monoclonal anti-FLAG M2 antibody was purchased from Sigma.
Transient transfection and CAT reporter gene assay. COS cells were seeded into 50- or 100-mm-diameter petri dishes approximately 12 h prior to transfection and transiently transfected with Lipofectamine according to the manufacturer's instructions (Gibco-BRL, Rockville, Md.). MEF cell lines were seeded into six-well plates approximately 24 h prior to transfection and transiently transfected with FuGENE6 reagent according to the manufacturer's instructions (Gibco-BRL). Chloramphenicol acetyltransferase (CAT) enzymatic assays were performed as described previously (5). In all transfections, the amount of DNA was equilibrated with a corresponding empty vector. A cytomegalovirus--galactosidase plasmid reporter (Gibco-BRL) was used to monitor transfection efficiency. The activity of the reporter plasmid alone is given as 1. The data are representative of three independent transfections, which were performed in duplicates. Error bars give the standard errors of the mean.
in vivo ubiquitylation assay, immunoprecipitation assay, and Western blot analysis. 293 and B4 cells were mock treated or treated for 4 h with MG-132 (50 μM) and then lysed in 0.4 ml of lysis buffer A (1% [vol/vol] NP-40, 10 mM Tris-HCl [pH 7.4], 150 mM NaCl, 2 mM EDTA, 0.1% protease inhibitor) for 1 h at 4°C. Total cell lysates were examined for the presence of hemagglutinin (HA) epitope-tagged Cdk9 species by Western blotting with the rabbit polyclonal anti-HA antibody. COS cells were transfected with indicated plasmid vectors. About 36 h posttransfection, the cells were mock treated or treated for 4 h with MG-132 (50 μM) and then lysed in 0.8 ml of lysis buffer A for 1 h at 4°C. Total cell lysates were immunoprecipitated with the rabbit polyclonal anti-HA antibody and examined for the presence of ubiquitin conjugates by Western blotting with the rabbit polyclonal anti-Myc antibody. MEFs were treated for 2 h with MG-132 (50 μM) and then lysed in 0.4 ml of lysis buffer A for 1 h at 4°C. Total cell lysates were immunoprecipitated with the rabbit polyclonal anti-Cdk9 antibody and examined for the presence of ubiquitin conjugates [(Ub)n Conj.] by Western blotting with the mouse monoclonal anti-ubiquitin antibody. An immunoprecipitation assay was performed as described previously (12). Western blotting was performed according to standard protocols.
Preparation of TAR RNA and RNA capture assay. -32P-labeled TAR RNA was prepared from the pGEM7WT plasmid by transcribing the linearized plasmid template (HindIII) with T7 RNA polymerase in the presence of [-32P]UTP by using the MEGAshortscript T7 kit (Ambion, Austin, Tex.). TAR RNA was excised from a 5% denaturing polyacrylamide gel, eluted (1.95 M ammonium acetate, 1% sodium dodecyl sulfate, 60 μg of tRNA/ml) for 6 h at 60°C, purified by using a phenol-chloroform extraction, and precipitated with 2.5 volumes of ethanol. Prior to use, the RNA pellet was dissolved in RNA capture buffer (50 mM HEPES-KOH [pH 7.6], 50 mM KCl, 0.1 mM EDTA, 5% glycerol, 7 mM MgCl2, 10 mM dithiothreitol). Total cell lysates were prepared from MEF cells, which expressed transiently the FLAG epitope-tagged human CycT1 protein, by using the lysis buffer A. P-TEFb complexes were immunoprecipitated onto protein A-Sepharose beads (Pharmacia, Piscataway, N.J.) with the mouse monoclonal anti-FLAG M2 antibody, washed three times in the wash buffer (50 mM Tris-HCl [pH 7.4], 0.25 M NaCl, 0.1% Triton X-100, 5 mM EDTA, 10 mM dithiothreitol, 1 mM sodium metabisulfite, 0.2 mM phenylmethylsulfonyl fluoride) and two times in RNA capture buffer before resuspension in 300 μl of RNA capture buffer. RNA capture assays were performed in the presence or absence of recombinant GST-Tat protein (500 ng), -32P-labeled TAR RNA and tRNA (10 ng/ml). Binding reaction mixtures were incubated at room temperature for 1 h with gentle rotation and then washed extensively with RNA capture buffer. Capture TAR RNA from the resulting RNA-protein complexes was extracted with TRIzol reagent (Gibco-BRL, Rockville, Md.), concentrated by precipitation (0.3 M sodium acetate, 20 μg of tRNA, 2.5 volumes of ethanol), resolved on 10% denaturing polyacrylamide gel, and analyzed by autoradiography.
RESULTS
Cdk9 is ubiquitylated in vivo. To determine whether Cdk9 is ubiquitylated, we performed an in vivo ubiquitylation assay (Fig. 1). Briefly, we coexpressed the hybrid HA.Cdk9 protein (Cdk9) with a Myc epitope-tagged ubiquitin or an empty plasmid effector pcDNA3.1 in COS cells and incubated these cells in the absence or presence of the proteasomal inhibitor MG-132. After isolation of Cdk9 by immunoprecipitation, the polyubiquitylated Cdk9 protein was detected by Western blotting with the anti-Myc antibody. In the absence of MG-132, we detected a light smear of high-molecular-weight ubiquitin conjugates in Cdk9 immunoprecipitations (Fig. 1, lane 1). Importantly, whereas in the presence of MG-132, this smear increased significantly, it was absent in cells where Cdk9 was not expressed (Fig. 1, compare lanes 2 and 3). In contrast, we found that the hybrid CycT1.HA protein was not ubiquitylated in vivo (data not presented). Overall, in agreement with the findings of Kiernan et al. (13), these data demonstrate that Cdk9 but not CycT1 is ubiquitylated in vivo.
Skp2 promotes ubiquitylation of Cdk9. To identify the E3 ubiquitin ligase that targets Cdk9, we developed a ubiquitylation assay where we examined a smear of high-molecular-weight ubiquitin conjugates of Cdk9 (Fig. 2). Briefly, we used B4 cells, which are human embryonal kidney 293 cells that express the HA epitope-tagged Cdk9 protein (Cdk9) and the parental 293 cells as controls. Cells were untreated or treated with MG-132, and cell lysates were examined for the presence of Cdk9 by Western blotting with the anti-HA antibody. We detected Cdk9 in B4 but not in parental 293 cells (Fig. 2a, compare lanes 1 and 2 with lanes 3 to 6). Since Cdk9 is ubiquitylated in vivo (Fig. 1), we also observed a smear of high-molecular-weight ubiquitin.Cdk9 conjugates in the presence of MG-132 only in cell lysates prepared from B4 but not in lysates prepared from parental 293 cells (Fig. 2a, compare lanes 2 and 4). Therefore, we expressed a number of wild-type and mutant F proteins in B4 cells and examined their effects on the ubiquitylation of Cdk9. The expression of Skp2 increased amounts of high-molecular-weight ubiquitin.Cdk9 conjugates significantly in MG-132-treated B4 cells. In contrast, the expression of the mutant dominant-negative Skp2F protein resulted in the loss of this ubiquitylation (Fig. 2a, compare lane 4 with lanes 5 and 6). Since the expression of another F-box protein, -TrCP, had no effect on Cdk9 ubiquitylation (data not presented), these effects were specific. Thus, our analysis identified Skp2 as the E3 ubiquitin ligase for Cdk9.
Next, we wanted to demonstrate the requirement of Skp2 for Cdk9 ubiquitylation in an endogenous system. We took advantage of primary MEFs which do or do not express Skp2 (Skp2+/+ and Skp2–/– MEFs, respectively) and performed in vivo ubiquitylation assays in these cells (Fig. 2b). Briefly, we treated cells with MG-132, prepared total cell lysates, immunoprecipitated the endogenous Cdk9 protein, and detected the polyubiquitylated Cdk9 protein by Western blotting with the anti-ubiquitin antibody. In the presence of Skp2, we detected the polyubiquitylated Cdk9 protein. On the contrary, this ubiquitin-Cdk9 (Ub.Cdk9) conjugate was not detected in the absence of Skp2 (Fig. 2b, upper panel, compare lanes 1 and 2). Moreover, while we observed slightly lower levels of Cdk9 in total cell lysates prepared from Skp2+/+ MEF cells compared to Skp2–/– MEF cells, those of actin remained unaffected (Fig. 2b, lower panel, compare lanes 1 and 2). Taken together, these studies confirmed that Skp2 is required for the ubiquitylation of Cdk9.
Skp2 facilitates Tat transactivation. To determine the importance and specificity of Skp2-mediated Cdk9 ubiquitylation for transcriptional activation, we tested its effects on Tat transactivation (Fig. 3). First, we expressed the plasmid reporter pHIVSCAT, together with Tat and a plasmid effector for Skp2 or -TrCP in COS cells, and measured levels of Tat transactivation with CAT reporter assays. Tat activated the pHIVSCAT 26-fold. The expression of Skp2 increased this activation to 54-fold. In contrast, when we expressed -TrCP instead of Skp2, the activity of pHIVSCAT remained unaffected (Fig. 3a). We conclude that Skp2 increases Tat transactivation specifically. Moreover, since the expression of Skp2 enhanced this transcription, Skp2 may play a positive role in Tat transactivation.
To address this stimulatory role of Skp2 for Tat transactivation, we next coexpressed the mutant dominant-negative Skp2F protein, Tat, and pHIVSCAT (Fig. 3b). Increasing amounts of the mutant Skp2F protein resulted in a dose-dependent repression of Tat transactivation (Fig. 3b, compare bar 2 with bars 3 and 4). Importantly, when we coexpressed the highest amount of the mutant Skp2F protein with increasing amounts of the wild-type Skp2 protein, we restored Tat transactivation to normal levels (Fig. 3b, compare bars 3 and 4 with bars 5 and 6). These data confirmed that Skp2 plays a positive role in Tat transactivation.
To explore the specificity of the inhibitory effect of the mutant Skp2F protein on Tat transactivation, we mimicked the natural recruitment of P-TEFb to the HIV-1 LTR by using two distinct artificial tethering systems of P-TEFb to DNA or RNA elements in the HIV-1 LTR and tested the effects of the mutant Skp2F protein on transcriptional activation by P-TEFb in these systems (Fig. 3c and d). The DNA-tethering system consists of the hybrid Gal4.CycT1 protein and the pG6TAR reporter plasmid, containing six high-affinity Gal4 DNA-binding sites in the HIV-1 LTR, whereas the RNA-tethering system takes advantage of the binding between the hybrid Rev.Tat protein and the Rev response element (RRE), which was fused to the stem of TAR, thus creating the pSLIIBCAT reporter plasmid (12, 24). In sharp contrast to Tat transactivation, neither the activation of the pG6TAR reporter plasmid by the DNA-tethered hybrid Gal4.CycT1 protein, nor the activation of pSLIIBCAT reporter plasmid by the RNA-tethered hybrid Rev.Tat protein was affected by the mutant Skp2F protein (Fig. 3c and d, compare bars 2 with 3). Thus, effects observed in Tat transactivation were specific for TAR.
Next, we wanted to confirm the above findings genetically in an endogenous system. For these experiments, we used wild-type and Skp2-deficient MEF cell lines (Skp2+/+ and Skp2–/– MEFs, respectively; Fig. 4). We coexpressed pHIVSCAT and Tat alone or in the presence of a human FLAG epitope-tagged CycT1 protein in these cells and measured CAT activity. Whereas Tat activated the pHIVSCAT 7-fold in the Skp2+/+ MEF cells, this activation decreased 2.5-fold in the Skp2–/– MEF cells (Fig. 4a, compare bars 3 and 4). Furthermore, we used the artificial tethering systems of P-TEFb to DNA or RNA elements and tested the requirement of Skp2 for transcriptional activation by P-TEFb in these systems (Fig. 4b). Consistent with the findings presented in Fig. 3c and d, the loss of Skp2 did not result in a decreased activation of the pG6TAR and pSLIIBCAT reporter plasmids by the hybrid Gal4.CycT1 and Rev.Tat proteins, respectively, underscoring again the specificity of effects observed in Tat transactivation (Fig. 4b, compare bars 3 with 4 and 7 with 8). Thus, Skp2 facilitates optimal Tat transactivation. Since the transcriptional activation by the Rev.Tat chimera was unaffected by the absence of Skp2 and the ubiquitylation of Cdk9 depended on Skp2 (Fig. 2b), the ubiquitylation of Cdk9 could contribute to the formation of the ternary complex between P-TEFb, Tat, and TAR.
Skp2 facilitates the formation of the RNA-protein complex between P-TEFb, Tat, and TAR. To examine the prediction that Skp2-mediated ubiquitylation of Cdk9 contributes to the formation of the ternary complex between P-TEFb, Tat, and TAR, we used an RNA capture assay (Fig. 5). Briefly, we isolated ubiquitylated and nonubiquitylated P-TEFb complexes from MG-132-treated Skp2+/+ and Skp2–/– MEF cells, respectively. These P-TEFb complexes were incubated with the GST.Tat chimera and -32P-labeled TAR, which was subjected to RNA capture assay. Since P-TEFb isolated from MEF Skp2+/+ cells captured TAR RNA six times more efficient compared to P-TEFb from MEF Skp2–/– cells, the capture of TAR depended on the presence of Skp2 (Fig. 5, compare lanes 3 and 4). Predictably, it also depended on the FLAG.CycT1 and GST.Tat chimeras (Fig. 5, compare lanes 1 to 3). Note that amounts of immunoprecipitated P-TEFb were equal in all reactions (Fig. 5, lower panel, lanes 2 to 4). Thus, ubiquitylation of Cdk9 by Skp2 facilitates the formation of the ternary complex between P-TEFb, Tat, and TAR.
DISCUSSION
In this study, we provide evidence that the ubiquitylation of Cdk9 by Skp2 facilitates Tat transactivation. First, by using several independent systems, we found that Cdk9 is ubiquitylated in vivo. Second, not only was Skp2 required for Cdk9 ubiquitylation but it also contributed to optimal levels of Tat transactivation. Finally, the ubiquitylation of Cdk9 by Skp2 facilitated the formation of the ternary complex between P-TEFb, Tat, and TAR. Thus, Tat requires ubiquitylated Cdk9 to activate optimally the transcriptional elongation of HIV-1 genes.
These findings extend recent observations for the ubiquitin/proteosome system in the regulation of HIV-1 gene expression. Whereas the Hdm2 E3 ubiquitin ligase activity also facilitates Tat transactivation (1), the role of Skp2 had not been addressed in this system. Rather, a negative role of Skp2 had been implicated in the function of the major histocompatibility complex class II transactivator (CIITA), which also depends on P-TEFb (13). Therein, the expression of CycT1 which lacks the PEST sequence at its C terminus (CycT1PEST) that binds Skp2 and mediates ubiquitylation of Cdk9, increased transcription activated by CIITA. In contrast, our biochemical and genetic evidence suggests that Skp2 acts as a coactivator for Tat transactivation. Indeed, the expression of the mutant Skp2F protein or the absence of Skp2 in the MEF Skp2–/– cells resulted in a loss of Cdk9 ubiquitylation. As a consequence, levels of Tat transactivation decreased. Consistently, the expression of the mutant CycT1PEST protein decreased this activity in COS cells (data not presented). Overall, these observations indicate that multiple E3 ubiquitin ligases regulate one transcription unit. Moreover, ubiquitylation of Cdk9 could affect transcriptional units differently, which could stem from their distinct mechanisms of transcriptional activation. Indeed, CIITA recruits P-TEFb to activate transcriptional elongation of its target genes via the MHCII enhanceosome assembled on the DNA, whereas Tat achieves the same task by recruiting P-TEFb to TAR RNA (12, 27).
It remains to be established what fraction of Cdk9 is ubiquitylated in cells. Although we found slightly increased steady-state levels of Cdk9 in total cell lysates prepared from MEF Skp2–/– cells compared to MEF Skp2+/+ cells, we did not observe any changes in levels of Cdk9 upon treatment of cells with the proteosomal inhibitor MG-132 or upon the expression of the Skp2F protein. Likewise, a recent report by Garriga et al. suggested that levels of Cdk9 protein do not depend on Skp2 (8). However, our functional data underscore the importance of Cdk9 ubiquitylation for the efficient formation of the ternary RNA-protein complex, which leads to optimal levels of Tat transactivation. In light of these observations, it is tempting to speculate that Cdk9 becomes ubiquitylated only when associated with the HIV-1 promoter or other cellular promoter elements. This scenario would resemble a mechanism of transcriptional activation by SREBP-1, c-Myc, and VP16, where the ubiquitylation of these activators occurs optimally on DNA (14, 15, 22, 23, 25). Indeed, E3 ubiquitin ligase Skp2 associates with the cyclin D2 promoter in a c-Myc-dependent manner in cells (25). It remains to be determined whether this situation holds true for the HIV-1 LTR as well.
How could the ubiquitylated Cdk9 affect Tat transactivation In principle, ubiquitin moieties on Cdk9 could influence protein-protein or RNA-protein interactions. The evidence obtained from tethering experiments in which Skp2 did not affect transcription activated by heterologously tethered P-TEFb via DNA or RNA targets suggests that ubiquitylation of Cdk9 regulates positively neither its association with CycT1 to increase P-TEFb kinase activity nor its binding to Tat to increase the levels of P-TEFb on the HIV-1 promoter. Rather, by using an RNA capture assay, we found that the ubiquitylated form of Cdk9 facilitates the overall assembly of the ternary RNA-protein complex, possibly by increasing direct binding to TAR or by facilitating the relief of autoinhibitory intramolecular interactions within CycT1 (see below).
The recruitment of P-TEFb to the paused RNAPII by Tat is critical for stimulating transcriptional elongation of HIV-1 genes and viral replication. The significance of this decisive event in the course of the HIV-1 replicative cycle is reflected in the existence of several mechanisms, which regulate the formation of the ternary complex between P-TEFb, Tat, and TAR. The following model emerges from numerous studies. First, an inhibitory intramolecular interaction between the N- and C-terminal regions of CycT1, which prevents the ternary complex assembly, is relieved by the binding of the transcription elongation factor Tat-SF1 to the C-terminal region in CycT1 (4). Second, Cdk9 phosphorylation is required for high-affinity binding of Tat and P-TEFb to TAR RNA (4, 7). In addition, the ubiquitylation of Cdk9 by Skp2 increases this RNA-protein complex assembly even further. Due to this sequence of events, high amounts of P-TEFb are concentrated in close proximity to its substrates, the CTD of RNAPII and the Spt5 subunit of DSIF. Moreover, P-TEFb targets the RD subunit of NELF, which dissociates from the lower stem of TAR RNA, thus contributing to the release of paused RNAPII (6). Finally, the acetylation at a single lysine residue in the TAR RNA-binding domain of Tat by the transcriptional coactivator p300 enhances Tat transactivation, possibly by dissociating CycT1 from TAR RNA and transferring Tat onto the elongating RNAPII (11). This model illustrates nicely how the assembly and disassembly of the multisubunit RNA-protein complex is a highly dynamic and temporally regulated process, which requires the participation of a variety of cellular cofactors, as well as multiple posttranslational modifications of its primary players.
ACKNOWLEDGMENTS
We thank Qiang Zhou, Kei-ichi Nakayama, Andrea C. Carrano, and Shao-Cong Sun for reagents and the members of Peterlin laboratory for useful discussions.
M.B. was supported in part by a fellowship from the Ministry of Science and Technology, Republic of Slovenia and by grant number 106584-36-RFNT from the American Foundation for AIDS Research. This study was supported by a grant from the National Institutes of Health (RO1 AI49104).
REFERENCES
Bres, V., R. E. Kiernan, L. K. Linares, C. Chable-Bessia, O. Plechakova, C. Treand, S. Emiliani, J. M. Peloponese, K. T. Jeang, O. Coux, M. Scheffner, and M. A. Benkirane. A non-proteolytic role for ubiquitin in Tat-mediated transactivation of the HIV-1 promoter. 2003. Nat. Cell Biol. 5:754-761.
Carrano, A. C., E. Eytan, A. Hershko, and M. Pagano. SKP2 is required for ubiquitin-mediated degradation of the CDK inhibitor p27. 1999. Nat. Cell Biol. 1:193-199.
Conaway, R. C., C. S. Brower, and J. W. Conaway. 2002. Emerging roles of ubiquitin in transcription regulation. Science 296:1254-1258.
Fong, Y. W., and Q. Zhou. 2000. Relief of two built-in autoinhibitory mechanisms in P-TEFb is required for assembly of a multicomponent transcription elongation complex at the human immunodeficiency virus type 1 promoter. Mol. Cell. Biol. 20:5897-5907.
Fujinaga, K., T. P. Cujec, J. Peng, J. Garriga, D. H. Price, X. Grana, and B. M. Peterlin. 1998. The ability of positive transcription elongation factor B to transactivate human immunodeficiency virus transcription depends on a functional kinase domain, cyclin T1, and Tat. J. Virol. 72:7154-7159.
Fujinaga, K., D. Irwin, Y. Huang, R. Taube, T. Kurosu, and B. M. Peterlin. 2004. Dynamics of human immunodeficiency virus transcription: P-TEFb phosphorylates RD and dissociates negative effectors from the transactivation response element. Mol. Cell. Biol. 24:787-795.
Garber, M. E., T. P. Mayall, E. M. Suess, J. Meisenhelder, N. E. Thompson, and K. A. Jones. 2000. CDK9 autophosphorylation regulates high-affinity binding of the human immunodeficiency virus type 1 tat-P-TEFb complex to TAR RNA. Mol. Cell. Biol. 20:6958-6969.
Garriga, J., S. Bhattacharya, J. Calbo, R. M. Marshall, M. Truongcao, D. D. Haines, and X. Grana. 2003. Cdk9 is constitutively expressed throughout the cell cycle, and its steady-state expression is independent of Skp2. Mol. Cell. Biol. 23:5165-5173.
Ghosh, S., M. J. Selby, and B. M. Peterlin. 1993. Synergism between Tat and VP16 in transactivation of HIV-1 LTR. J. Mol. Biol. 234:610-619.
Ivanov, D., Y. T. Kwak, J. Guo, and R. B. Gaynor. 2000. Domains in the SPT5 protein that modulate its transcriptional regulatory properties. Mol. Cell. Biol. 20:2970-2983.
Kaehlcke, K., A. Dorr, C. Hetzer-Egger, V. Kiermer, P. Henklein, M. Schnoelzer, E. Loret, P. A. Cole, E. Verdin, and M. Ott. 2003. Acetylation of Tat defines a cyclinT1-independent step in HIV transactivation. Mol. Cell 12:167-176.
Kanazawa, S., T. Okamoto, and B. M. Peterlin. 2000. Tat competes with CIITA for the binding to P-TEFb and blocks the expression of MHC class II genes in HIV infection. Immunity 12:61-70.
Kiernan, R. E., S. Emiliani, K. Nakayama, A. Castro, J. C. Labbe, T. Lorca, K. Nakayama, and M. Benkirane. 2001. Interaction between cyclin T1 and SCF(SKP2) targets CDK9 for ubiquitination and degradation by the proteasome. Mol. Cell. Biol. 21:7956-7970.
Kim, S. Y., A. Herbst, K. A. Tworkowski, S. E. Salghetti, and W. P. Tansey. 2003. Skp2 regulates Myc protein stability and activity. Mol. Cell 11:1177-1188.
Kurosu, T., and B. M. Peterlin. 2004. VP16 and ubiquitin: binding of P-TEFb via its activation domain and ubiquitin facilitates elongation of transcription of target genes. Curr. Biol. 14:1112-1116.
Madore, S. J., and B. R. Cullen. 1993. Genetic analysis of the cofactor requirement for human immunodeficiency virus type 1 Tat function. J. Virol. 67:3703-3711.
Muratani, M., and W. P. Tansey. 2003. How the ubiquitin-proteasome system controls transcription. Nat. Rev. Mol. Cell. Biol. 4:192-201.
O'Keeffe, B., Y. Fong, D. Chen, S. Zhou, and Q. Zhou. 2000. Requirement for a kinase-specific chaperone pathway in the production of a Cdk9/cyclin T1 heterodimer responsible for P-TEFb-mediated tat stimulation of HIV-1 transcription. J. Biol. Chem. 275:279-287.
Orphanides, G., and D. Reinberg. 2002. A unified theory of gene expression. Cell 108:439-451.
Pickart, C. M. 2001. Ubiquitin enters the new millennium. Mol. Cell 8:499-504.
Price, D. H. 2000. P-TEFb, a cyclin-dependent kinase controlling elongation by RNA polymerase II. Mol. Cell. Biol. 20:2629-2634.
Salghetti, S. E., A. A. Caudy, J. G. Chenoweth, and W. P. Tansey. 2001. Regulation of transcriptional activation domain function by ubiquitin. Science 293:1651-1653.
Sundqvist, A., and J. Ericsson. 2003. Transcription-dependent degradation controls the stability of the SREBP family of transcription factors. Proc. Natl. Acad. Sci. USA 100:13833-13838.
Taube, R., K. Fujinaga, J. Wimmer, M. Barboric, and B. M. Peterlin. 1999. Tat transactivation: a model for the regulation of eukaryotic transcriptional elongation. Virology 264:245-253.
von der Lehr, N., S. S. Johansson, Wu, F. Bahram, A. Castell, C. Cetinkaya, P. Hydbring, I. Weidung, K. Nakayama, K. I. Nakayama, O. Soderberg, T. K. Kerppola, and L. G. Larsson. 2003. The F-box protein Skp2 participates in c-Myc proteosomal degradation and acts as a cofactor for c-Myc-regulated transcription. Mol. Cell 11:1189-1200.
Wada, T., T. Takagi, Y. Yamaguchi, A. Ferdous, T. Imai, S. Hirose, S. Sugimoto, K. Yano, G. A. Hartzog, F. Winston, S. Buratowski, and H. Handa. 1998. DSIF, a novel transcription elongation factor that regulates RNA polymerase II processivity, is composed of human Spt4 and Spt5 homologs. Genes Dev. 12:343-356.
Wei, P., M. E. Garber, S. M. Fang, W. H. Fischer, and K. A. Jones. 1998. A novel CDK9-associated C-type cyclin interacts directly with HIV-1 Tat and mediates its high-affinity, loop-specific binding to TAR RNA. Cell 92:451-462.
Xiao, G., E. W. Harhaj, and S. C. Sun. 2001. NF-B-inducing kinase regulates the processing of NF-B2 p100. Mol. Cell 7:401-409.
Yamaguchi, Y., T. Takagi, T. Wada, K. Yano, A. Furuya, S. Sugimoto, J. Hasegawa, and H. Handa. 1999. NELF, a multisubunit complex containing RD, cooperates with DSIF to repress RNA polymerase II elongation. Cell 97:41-51.(Matjaz Barboric, Fan Zhan)