Cyclin K/CPR4 inhibits primate lentiviral replication by inactivating Tat/positive transcription elongation factor b-dependent long terminal repeat transcription

The positive transcription elongation factor b complexes comprise CDK9 and a C-type cyclin, required for the efficient expression of both eukaryotic and primate lentivirus-encoded genes. Cyclin K/CPR4 is the least studied of the positive transcription elongation factor b-forming cyclins. Here, we demonstrate that cyclin K/CPR4-containing positive transcription elongation factor b complexes are unresponsive to Tat and HEXIM1-mediated inactivation. Enhancing expression of cyclin K/CPR4 inhibited the human and simian immunodeficiency viral replication. These data indicate that cyclin K/CPR4 functions as a natural inhibitor of primate lentiviruses.

Positive transcription elongation factor b (P-TEFb) complexes comprise CDK9 and a C-type cyclin (cyclin T1, T2 or K) and play a role in transcriptional initiation in/or elongation of cellular genes ^[1-6]. The P-TEFb complex is critical for Tat-dependent transcription from the human and simian immunodeficiency virus (HIV-1 and SIV) long terminal repeat (LTR) promoter. By interacting with the P-TEFb complex, Tat recruits the P-TEFb complex to the trans-activation-responsive (TAR) viral RNA element, which is located close to the transcriptional start site to which Tat binds in complex with P-TEFb complex. CDK9 phosphorylates the carboxy-terminal domain of RNA polymerase II, which boosts transcription from the proviral genome ^[7]. Among the P-TEFb complex-forming cyclins, only CCNT1 has the Tat responsive motif (TRM). Given the structural and functional similarities between CCNT2 and CCNK ^[8], expression of CCNK has been assumed to negatively affect Tat-dependent transcription, thus primate lentiviral replication. This possibility, however, has not yet been examined. Recently, it was found that the P-TEFb complex containing CCNT1 or CCNT2 is inactivated by the 7SK RNA-HEXIM1 complex ^[9-12]. The HEXIM1-mediated inhibition of CCNK-containing P-TEFb complexes remains to be tested.

We constructed mammalian expression plasmids for wild type and N-terminal FLAG-tagged or green fluorescent protein (GFP)-tagged human CCNK (WT, fCCNK or GFP-fCCNK). The CCNK function is unaffected by N-terminal epitope tagging ^[2,13,14]. GFP-fCCNK was used to examine the intracellular localization of fCCNK in NP2 cells. GFP-fCCNK accumulated in the discrete nuclear speckle compartments, similar to GFP-CCNT1, known to be sites of RNA Pol II-mediated transcription and co-transcriptional premRNA processing ^[15] and exhibited co-distribution with hemaglutinin eipope tagged CDK9 (data not shown). These data suggest functional similarities between CCNK and CCNT1.

We examined the physical association between CCNK and the transcriptional regulators HEXIM1 and HIV-1 Tat. Under the same conditions, endogenously expressed HEXIM1 did not co-immunoprecipitate with fCCNK, but co-immunoprecipitated successfully with FLAG-tagged CCNT1 (fCCNT1; Fig. 1a). Similarly, FLAG-tagged Tat could not co-immunoprecipitate CCNK under conditions in which it did precipitate GFP-CCNT1 (Fig. 1a). These data suggest that the CCNK-containing P-TEFb complex is unresponsive to HEXIM1 or Tat and, thus, it would appear that CCNK is unique among P-TEFb complex-forming cyclins.

Fig. 1

To test whether CCNK expression reduces HIV-1 production at the level of transcription, we performed a transient transfection assay to measure HIV-1 production in the presence of different levels of CCNK expression. HIV-1 proviral DNA was co-introduced into 293 T cells with increasing amounts of fCCNK expression plasmid. It was found that the viral production was decreased dramatically in a dose-dependent manner (Fig. 1b). Similarly, the production of SIV was reduced (data not shown). Western blot analysis revealed that the level of p24 antigen in cell lysates dropped significantly with increasing levels of fCCNK (Fig. 1b). By contrast, the level of the internal control Bip was unaffected (Fig. 1b). Real-time (RT)-PCR analysis revealed that viral mRNA levels fell with increasing fCCNK expression, whereas the internal control cyclophilin A mRNA levels were unaffected (Fig. 1b). The reduction in viral mRNA was comparable with that observed for viral protein. Specifically, RT-PCR demonstrated that 80 ng of pfCCNK exerted the maximum reduction in viral mRNA levels (5.1% relative to the control), a finding similar to the viral production levels quantified by enzyme-linked immunosorbent assay (3.3%, Fig. 1b). These data strongly suggest that CCNK inhibits HIV-1 production by specifically limiting Tat-dependent LTR transcription.

Finally, we investigated the potential inhibitory activity of CCNK on HIV-1 replication in a physiologically relevant system. We established human T cell lines that constitutively overexpressed fCCNK or luciferase (Luc) by infecting MOLT-4, MT-4 and M8166 cells with murine leukemia viral vectors carrying fCCNK-IRES-GFP or Luc-IRES-GFP cassettes and GFP-positive cells were collected as described previously to examine the role of HEXIM1 in the primate lentiviral replication ^[16]. Expression of fCCNK was verified by western blot analysis using an anti-CCNK antibody and the expression of cellular proteins such as CCNT1, CDK9, HEXIM1, Bip, and cell surface receptors for HIV-1 (CD4 and CXCR4) was comparable between fCCNK-expressing and Luc-expressing cells. In addition, no differences in rate of cell proliferation were detected between fCCNK-expressing and Luc-expressing cells (data not shown). We measured HIV-1 and SIV replication in these T cell lines and found that the viral replication was strongly inhibited in all the three fCCNK-expressing T cell lines (data not shown). These data directly demonstrate that CCNK functions as a negative regulator of primate lentiviral replication. In addition, we verified that constitutive fCCNK expression did not block the early phase of the viral life cycle by conducting a real-time Alu-LTR PCR assay to quantify the integrated viral genome in MOLT-4 and MT-4 cells infected with VSV-G-pseudotyped replication-incompetent HIV-1 vector (data not shown).

Taken together, our results indicate that CCNK is able to limit the replication of primate lentiviruses by competing with CCNT1 for CDK9, which results in a reduction of CCNT1-containing Tat-responsive P-TEFb complexes, thereby inactivation of Tat/P-TEFb-dependent transcription of the viral LTR promoter. Our report is the first demonstration that constitutive upregulation of a C-type cyclin CCNK limits primate lentiviral replication in human T cell lines without conferring any detectable effect on cell proliferation. The lack of any detectable effect on cell proliferation could be because CCNK-containing P-TEFb complexes can complement the role played by the CCNT1-containing P-TEFb complex in cellular gene transcription. These data highlight the importance of Tat-dependent LTR transcription for the efficient propagation of the primate lentivirus and the potential protective role of CCNK in the pathogenesis of primate lentiviruses.

Acknowledgements

This work was supported in part by the Japan Health Science Foundation, the Japanese Ministry of Health, Labor and Welfare (H18-AIDS-W-003) and the Japanese Ministry of Education, Culture, Sports, Science and Technology (18689014 and 18659136).