The cyclin dependent kinase (CDK)9 and its cyclin T partner have received much attention since this complex was proposed as a putative cellular cofactor of the HIV-1 Tat transactivator (reviewed in [1-3]). Although Tat is able to transactivate the HIV promoter in the absence of any viral encoded protein , a considerable body of evidence indicates that Tat interacts with cellular proteins. A critical role for cellular cofactor in Tat transactivation was originally suggested by the inability of Tat to function in rodent cells [5,6].
Tat appears to act almost exclusively at the level of transcriptional elongation. In the presence of Tat elongation is very effective, and hyperphosphorylation of the RNA polymerase IIcarboxy terminal domain (RNAPII CTD) has been suggested as the molecular event underlying Tat transactivation [1,3,4], and it has been shown that removal of CTD blocks the response of RNAPII to Tat [7-9]. In addition, Tat-mediated activation is effectively inhibited by 5,6-dichloro-1β-D-ribofuranosylbenzymidazole (DRB), a nucleoside analogue that inhibits the kinase activity required for elongation by RNAPII. In accordance with the proposal that Tat requires a cofactor, a cellular kinase complex called TAK (Tat activating kinase) was identified , and the kinase component of TAK was then shown to be identical to a previously identified cdc2-related kinase named PITALRE  (for a recent review see ) which has been renamed CDK9 [13,14].
As for other CDKs a cyclin partner of CDK9, called cyclin T, has been isolated [13,14]. It has been shown that CDK9 associates specifically with multiple cyclin subunits: cyclin T1, cyclin T2a and T2b, the latter two being generated by alternative splicing . Cyclin T1 specifically binds the Tat protein both as a subunit of the cyclinT1/CDK9 complex and on its own [14,15], and cyclin T1 increases the affinity of Tat for TAR [14,15]. The ability of cyclinT1 to restore Tat activity in rodent cells and its localization on chromosome 12 provided strong evidences that cyclin T1, probably as a component of the TAK/P-TEFb complex, is a bona fide human cellular cofactor for Tat [16-18]. In addition, it has been shown recently that unlike human cyclin T1 mouse cyclin T1 does not bind to Tat; it fails to support Tat transactivation [16-18], and it was shown recently that cyclins T2 fail to bind Tat and rescue Tat function in rodent cells [18,19].
Here we report that like cyclin T1, cyclin T2 can associate in vitro with CDK9; however, unlike cyclin T1, cyclin T2 is unable to bind Tat and in agreement with the recent reports [18,19], we found that Tat function can be rescued in rodent cells only by cyclin T1. Most notably, enforced expression of cyclin T2 inhibits Tat transactivation.
Material and methods
Tissue culture and transfections
Human 293T and rodent Chinese hamster ovary (CHO) cells were grown at 37°C in Dulbecco‚s modified Eagle‚s medium supplemented with 10% foetal calf serum (Gibco, Life Technologies, Gaithersburg, MD, USA). Subconfluent cell cultures were transfected by the calcium phosphate method using different amounts of reporter and effector plasmids, as described in the text, while maintaining the quantity of DNA transfected constant at 10μg by the addition of empty vector DNA. For normalization of transfections efficiencies, 200ng of Renilla (sea pansy) luciferase expression plasmid was included in the transfections (pRL-CMV, Promega, Madison, Wisconsin, USA). Cells were harvested 48 h after addition of the precipitates, and extracts were assayed for luciferase activity using Dual-Luciferase Reporter assay (Promega) according to the manufacturer‚s instructions. The experimental reporter luciferase activity was then normalized to transfection efficiency as measured by the activity deriving from pRL-CMV. Western blotting was performed on transfected CHO cells with the anti-HA antibody (12CA5) and binding was visualized by enhanced chemiluminescence (ECL-plus kit, Amersham, Arlington Heights, IL, USA).
The reporter plasmid HIV-Luc contains the long terminal repeat (LTR) sequences inserted into the SmaI-HindIII sites of pGL3-Basic Vector (Promega). The glutathione S-transferase (GST)-Tat, pSV-Tat, GST-CDK9 and CMV-CDK9 plasmids have been described previously [20-22]. All cyclins T expression plasmids were made by inserting the indicated regions into a cytomegalovirus (CMV)-driven promoter from the pCDNAI (Invitrogen, Carlsbad, CA, USA) with an HA tag at the N terminus. The CMV-CycT1, T2a and T2b were as described . The fragments encoding the indicated protein regions of both cyclin T1 and T2 were isolated using appropriate restriction sites from the full-length cyclin T1 and T2b cDNAs  and inserted into the pCDNAI HA-tagged vector. All constructs were confirmed by DNA sequencing. Full details of each construction are available upon request.
Stationary phase cultures of Escherichia coli HB101 cells, transformed with the GST and GST expression plasmids derivatives, were diluted 1:10 in Luria Bertani broth to an OD600 0.4 and induced with 1mM isopropyl thio-β-D-galactoside for 3 h at 37°C. The cultures were harvested and resuspended in 5 ml cold NETN buffer [20mM Tris-HCl pH8, 100mM NaCl, 1mM EDTA, 0.5% Nonidet P-40, 1mM phenylmethanesulfonyl fluoride and protease inhibitors]. The cells were broken by sonication, ahd the lysate was resuspended in NETN buffer and rocked at 4°C for 5 h with 250ml of glutathione-Sepharose beads. The resin was washed six times with NETN buffer and resuspended in a final volume of 250μl of NETN buffer. The plasmids CMV-CycT1(1-726), CMV-CycT1(1-290), CMV-CycT1 (290-726), CMV-CycT1 (1-188), CMV-CycT2a(1-730), CMV-CycT2 (1-290) and CMV-Luc were used as templates to produce the in vitro 35S-labelled proteins by using the coupled transcription and translation system from rabbit reticulocyte lysate (T7 TNT system, Promega) according to the manufacturer‚s protocol.
In vitro binding assays
For in vitro binding 3μl [35S]methionine-labelled proteins were incubated with glutathione-Sepharose beads (GST, GST-CDK9 and GST-Tat, respectively) by rocking at 4°C for 3-5 h and subsequently spun down in a microcentrifuge. The supernatant was removed and the precipitates were washed extensively using NENT buffer. The resulting proteins were resolved on sodium lauryl sulphate/7% polyacrylamide gels. Dried gels were visualized and quantified by using the Molecular Dynamics PhosphorImager System (Molecular Dynamics, Sunnyvale, CA, USA).
Functional interaction between Tat and cyclins T
It has been shown recently that the TAK/P-TEFb complex is a cellular cofactor for HIV-1 Tat. Most importantly, P-TEFb was found to be essential for Tat transactivation both in vitro and in vivo [23,24]. The CDK9 kinase is the catalytic subunit of P-TEFb and it can be found associated with either cyclinT1 or with the related cyclin T2a or cyclin T2b, the latter two being generated by alternative splicing . A direct interaction between Tat and cyclin T1 has been demonstrated recently and such interaction enhances the affinity and specificity of Tat binding to TAR [14,18]. As P-TEFb consisting of CDK9 kinase and any one of the cyclin T subunits is able to enhance transcription in vitro and in vivo [13-15], the role of cyclin T2a and T2b proteins in Tat transactivation it is not clear.
To address the role of these different cyclins T in Tat function we tested the ability of the three cyclins to co-operate with Tat in transient transfections assays in rodent cells. Tat protein is a potent activator of HIV-1 LTR transcription in primate cells but only poorly functional in rodent cells [4-6]. However, Tat-mediated activation can be rescued by enforced expression of cyclin T1 [14-19]. To determine whether the same activity might also be attributable to the cyclin T2, rodent CHO cells were transfected with an HIV-1 LTR-luciferase reporter plasmid in the presence of Tat together with expression plasmids for cyclin T1, T2a and T2b, respectively. As shown in Fig. 1, it was found that Tat activates LTR-driven transcription only poorly. In contrast, when cyclin T1 was co-expressed with Tat, LTR-driven expression was increased in a dose-dependent manner (Fig. 1). Importantly, co-expression of either cyclin T2a or T2b failed to augment Tat activity. Western blot analysis of transfected cyclins (Fig. 1b) reveals that all three proteins were expressed at comparable levels in transfected cells, thus demonstrating that the efficient rescue of Tat activity by cyclin T1 was not due to different protein levels in transfected CHO cells.
To determine the relevant region of cyclin T1 responsible for Tat functional rescue in rodent cells a series of cyclin T1 expression vectors was constructed. These CMV-HA-tagged expression vectors are reported schematically in Fig. 2. Rodent CHO cells were then transfected with an HIV-1 LTR-luciferase reporter plasmid in the presence of Tat and with the indicated cyclin T1 expression vector. It was found that the cyclin homology region (amino acids 1-290) [13,14] was required and sufficient for efficient rescue of Tat transactivation. Further truncation to amino acid 180 completely eliminated the ability of cyclin T1 to enhance Tat activity (Fig. 2).
Cyclins T binding to Tat and CDK9
The results reported above strongly suggested that cyclin T1 co-operates specifically with Tat and that the cyclin box homology region (amino acids 1-290) is strictly required and sufficient for such functional interaction. In contrast, cyclin T2 failed to augment Tat activity in rodent cells. As both cyclins can be found associated with the CDK9 kinase , it is likely that the inability of cyclin T2 to enhance Tat transactivation might be due to the lack of association between Tat and cyclin T2. To address this point the ability of cyclin T1 and T2 to bind to Tat and the catalytic subunit CDK9 was compared by binding assays in vitro. The GST-Tat and GST-CDK9 proteins obtained from E. coli were incubated with the 35S-labelled in vitro translated products of cyclin T1 and T2, respectively. Various regions of the cyclin T1 and T2 cDNA were used for in vitro binding assays as represented schematically in Fig. 3. The specificity of the interactions was ascertained by the observation that both cyclin T1 and T2 did not adhere to GST resin (Fig. 3) and no binding of a control protein, 35S-labelled luciferase, was observed under identical conditions (data not shown). It was found that both cyclin T1 and T2 proteins bind efficiently to CDK9 kinase, and the cyclin box was required and sufficient for such interactions. In contrast, it was found that whereas cyclin T1 binds specifically to Tat, lack of significant binding between Tat and cyclin T2b proteins was detected (Fig. 3). Moreover, both cyclin T2 and T2b proteins gave identical results (data not shown). Interestingly, it was found that whereas the cyclin T1 homology region (amino acids 1-290) was required and sufficient for both Tat and CDK9 binding, further truncation to amino acid 189 eliminated completely the ability of cyclin T1 to bind to Tat but this cyclin T1 region still retained the ability to associate with CDK9 kinase. Thus, the region amino acids 189-290 of cyclin T1 contains residues critical for Tat binding. Notably, the lack of binding to Tat of the T1 region amino acids 1-189 correlates with the inability of this region to enhance Tat activation in rodent cells (Fig. 2). Taken together these findings strongly suggest that enhancement of Tat transactivation is correlated with the ability of cyclin T1 to associate with Tat. Conversely, the functional inability of cyclin T2 to enhance Tat function is probably due to the lack of association between cyclin T2 and Tat. A similar conclusion has been reported independently by other laboratories [16-19].
Enforced expression of cyclin T2 represses Tat transactivation
From the transient expression assays and in vitro binding data it appears that the lack of interaction between Tat and cyclin T2 is responsible for the inability of cyclin T2 to enhance Tat activation in rodent cells. However, like cyclin T1 cyclin T2 binds efficiently to CDK9 kinase (Fig. 3), and most importantly it has been found that cyclin T1 and T2 are associated with CDK9 kinase in a mutually exclusive manner . As CDK9 is strictly required for Tat activity [1-3], it was then queried whether overexpression of cyclin T2 might interfere with Tat function: an HIV-LTR reporter plasmid was co-transfected into rodent CHO cells with cyclin T1 and Tat in the presence of increasing amounts of cyclin T2b. As shown in Fig. 4, Tat enhancement mediated by cyclin T1 was partially abrogated by co-expression of cyclin T2b in a dose-dependent manner (Fig. 4a). Similar results were also obtained using cyclin T2a (data not shown). Thus, it appears that enforced expression of cyclin T2 proteins can cause an inhibition of cyclin T1-dependent enhancement of Tat activity in rodent cells, probably as a result of competition for binding to endogenous CDK9 kinase.
Whether cyclin T2 might interfere with Tat function in human cells was then tested by co-transfection of human 293T cells with HIV-LTR reporter plasmid and Tat in the presence of increasing amounts of either cyclin T1 or T2b. It was found that overexpression of cyclin T1 did not affect Tat activity significantly. In contrast, overexpression of cyclin T2b induced a significant reduction of the Tat transactivation. Notably, the cyclin box region (amino acids 1-286) gave a similar reduction of Tat transactivation (Fig. 4b). These findings indicate that enforced expression of the cyclin T2 N terminal region, encompassing the cyclin homology box, is sufficient for the induction of a significant inhibition in Tat transactivation. As we have determined that the N terminal region of cyclin T2 efficiently binds to CDK9 kinase (Fig. 3), it is conceivable that overexpression of cyclinT2 disrupts the stoichiometry of the P-TEFb cellular complex required for Tat transactivation.
In this study we determined that Tat transactivation in rodent cells can be restored by human cyclin T1, and the cyclin homology region (amino acids 1-290) is required and sufficient for efficient Tat transactivation in rodent cells. In contrast, human cyclin T2 does not support Tat transactivation in rodent cells. A similar finding has been reported recently by other laboratories [17,19]. Additionally, we found that overexpression of cyclin T2 partially decreases the effect of cyclin T1 on Tat activation in rodent cells and that it represses Tat transactivation in human cells. From our transient expression assays in both rodent and human cells together with our in vitro binding data, we conclude that diverse CDK9 kinase associated cyclins T exert different effects on Tat activity.
Our in vitro studies clearly demonstrated that both cyclin T1 and T2 proteins can bind the CDK9 protein specifically, and that the N-terminal regions of both T1 and T2 proteins, encompassing the cyclin homology region, are sufficient for binding. In sharp contrast, cyclin T1 but not T2 can bind Tat specifically. Moreover, we determined that the region of cyclin T1 able to bind in vitro to Tat (amino acids 1-290) can also restore Tat function in rodent cells. Conversely, the cyclin T1 region (amino acids 1-189), which retains efficient binding to CDK9 but it does not bind to Tat, failed to enhance Tat transactivation. Our results are fully consistent with recent independent reports showing delineation of the minimal domain of cyclin T1 required to bind Tat and TAR and support Tat function in rodent cells [16-18]. Thus, the combination of our data together with similar reported findings [16-19], strongly suggests that the ability of cyclin T1 to bind Tat correlates with the enhancement of Tat transactivation in rodent cells. Although we cannot exclude other possibilities, it is conceivable that the inability of cyclin T2 to enhance Tat activation in rodent cells is due to the lack of binding to Tat.
While this report was in preparation work from different laboratories demonstrated that, unlike the human cyclin T1, the mouse cyclin T1 fails to support Tat transactivation, and that the different effect is correlated to a single amino acid difference between human and rodent cyclin T1 - a tyrosine in place of a cysteine at residue 261 [16-18]. In accordance with the proposal that the cysteine 261 plays a critical role in Tat transactivation it was found that in cyclin T2 this residue is an asparagine.
Cyclin T1, as well as cyclin T2a and T2b proteins have been found associated with the CDK9 kinase in the multi-protein complex TAK/P-TEFb that is able to hyperphosphorylate the C-terminal domain of RNAPII, which is required for Tat transactivation [10,13-15]. Moreover, in vitro studies demonstrated that any of the two cyclin proteins can form an active complex with the CDK9 kinase that phosphorylates the CTD of RNAPII and induces the transition from abortive elongation to productive elongation . The ability of overexpressed cyclin T2 to inhibit Tat function suggests that enforced expression of cyclin T2 might change the relative stoichiometry of the P-TEFb functional complex. Consequently, Tat transactivation appears to require a dedicated P-TEFb complex containing the cyclin T1/CDK9 kinase. The crucial and specific role of human cyclin T1 protein in Tat transactivation has received further support from recent reports demonstrating that the levels of cyclin T1 are upregulated upon activation of peripheral blood lymphocytes and in purified CD4 primary T cells [25,26]. Notably, it has been shown that regulation of cyclin T2 expression is significantly different, with no change in protein levels following 4β-phorbol 12-myristate 13-acetate induction of HL-60 and U937 cells or in peripheral blood lymphocytes being observed [25,26]. These findings suggested that transcription of HIV-1 in activated T-cells might be due to the specific upregulation of the cyclin T1/CDK9 complex .
Our results showing that diverse CDK9-kinase- associated cyclins T exert opposite effects on Tat activity provide additional support for the proposal that the human cyclin T1/CDK9 kinase complex plays a fundamental role in Tat transactivation. Moreover, our findings showing that overexpression of cyclin T2 can interfere with Tat transactivation may provide an additional specific target for selective inhibition of Tat activity in infected cells.
The authors thank M. Giacca and D. Price for the gift of plasmids.
1. Jones KA. Taking a new TAK on tat transactivation. Genes Dev
2. Emerman M, Malin M. HIV-1 regulatory/accessory genes: keys to unraveling viral and host cell biology. Science
3. Cullen BR. HIV-1 auxiliary proteins: making connections in a dying cell. Cell
4. Jones KA, Peterlin BM. Control of RNA initiation and elongation at the HIV-1 promoter. Annu Rev Biochem
5. Alonso A, Derse D, Peterlin BM. Human chromosome 12 is required for optimal interactions between Tat and TAR of human immunodeficiency virus type 1 in rodent cells. J Virol
6. Hart CE, Galphin JC, Westhafer MA, Schochetman G. TAR loop-dependent human immunodeficiency virus trans activation requires factors encoded on human chromosome 12. J Virol
7. Chun RF, Jeang K-T. Requirements for RNA polymerase II carboxy-terminal domain for activated transcription of human retroviruses human T-cell lymphotropic virus I and HIV-1. J Biol Chem
8. Okamoto H, Sheline C, Corden J, Jones K, Peterlin B. Trans- activation by human immunodeficiency virus Tat requires the C-terminal domain of RNA polymerase II. Proc Natl Acad Sci USA
9. Parada C, Roeder R. Enhanced processivity of RNA polymerase II triggered by Tat-induced phosphorylation of its carboxy- terminal domain. Nature
10. Herrmann CH, Rice AP. Lentivirus Tat proteins specifically associate with a cellular protein kinase, TAK, that hyperphosphorylates the carboxyl-terminal domain of the large subunit of RNA polymerase II: candidate for a Tat cofactor. J Virol
11. Grana X, De Luca A, Sang N, et al. PITALRE, a nuclear CDC2-related protein kinase that phosphorylates the retinoblastoma protein in vitro. Proc Natl Acad Sci USA
12. De Falco G, Giordano A. CDK9 (PITALRE): A multifuntional cdc2-related kinase
. J Cell Physiol
13. Peng J, Zhu Y, Milton JT, Price DH. Identification of multiple cyclin subunits of human P-TEFb. Genes Dev
14. Wei P, Garber ME, Fang SM, Fischer WH, Jones KA. 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
15. Zhou Q, Chen D, Pierstorff E, Luo K. Transcription elongation factor P-TEFb mediates Tat activation of HIV-1 transcription at multiple stages. EMBO J
16. Garber ME, Wei P, Kewal Ramani VN, et al. The interaction beween HIV-1 Tat and human cyclin T1 requires zinc and a critical cysteine residue that is not conserved in the rodent CycT1 protein. Genes Dev
17. Bieniasz PD, Grdina TA, Bogerd HP, Cullen BR. Recruitment of a protein complex containing Tat and cyclin T1 to TAR governs the species specifity of HIV-1 Tat. EMBO J.
18. Fujinaga K, Taube R, Wimmer J, Cujec, TP, Peterlin BM. Interactions between human T, Tat, and the transactivation response elemnt (TAR) are disrupted by a cysteine to tyrosine substitution found in mouse cyclin T. Proc Natl Acad Sci USA
19. Winner J, Fujinaga K, Taube R, et al. Interactions between Tat and TAR and immunodeficiency virus replication are facilitated by human cyclin T1 but not cyclins T2a or T2b. Virology
20. Garriga J, Segura E, Mayol X, Grubmeyer C, Grana X. Phosphorylation site specificity of the CDC2-related kinase PITALRE. Biochem J
21. Majello B, Napolitano G, Lania L. Recruitment of the TATA-binding protein to the HIV-1 promoter is a limiting step for Tat transactivation. AIDS
22. Marzio G, Tyagy M, Gutierrez MI, Giacca M. HIV-1 Tat transactivator recruits p300 and CREB-binding protein histone acetyltransferase to the viral promoter. Proc Natl Acad Sci USA
23. Zhu Y, Pe‚ery T, Peng J, et al. Transcription elongation factor P-TEFb is required for HIV-1 tat transactivation in vitro. Genes Dev
24. Mancebo HS, Lee G, Flygare J, et al. P-TEFb kinase is required for HIV Tat transcriptional activation in vivo and in vitro. Genes Dev
25. Herrmann CH, Carroll RG, Wei P, Jones KA, Rice, AP. Tat-associated kinase, TAK, activity is regulated by distinct mechanisms in peripheral blood lymphocytes and promonocytic cell lines. J Virol
26. Garriga J, Peng J, Parreno M, Price DH, Henderson EE, Grana X. Upregulation of cyclin T1/CDK9 complexes during T cell activation. Oncogene