Tax interacts with P-TEFb in a novel manner to stimulate human T-lymphotropic virus type 1 transcription

Abstract

Human T-lymphotropic virus type 1 (HTLV-1) encodes a transcriptional activator, Tax, whose function is essential for viral transcription and replication. Tax transactivates the viral long-terminal repeat through a series of protein-protein interactions which facilitate CREB and CBP/p300 binding. In addition, Tax dissociates transcription repressor histone deacetylase 1 interaction with the CREB response element. The subsequent events through which Tax interacts and communicates with RNA polymerase II and cyclin-dependent kinases (CDKs) are not clearly understood. Here we present evidence that Tax recruits positive transcription elongation factor b (P-TEFb) (CDK9/cyclin T1) to the viral promoter. This recruitment likely involves protein-protein interactions since Tax associates with P-TEFb in vitro as demonstrated by glutathione S-transferase fusion protein pull-down assays and in vivo as shown by co-immunoprecipitation assays. Functionally, small interfering RNA directed toward CDK9 inhibited Tax transactivation in transient assays. Consistent with these findings, the depletion of CDK9 from nuclear extracts inhibited Tax transactivation in vitro. Reconstitution of the reaction with wild-type P-TEFb, but not a kinase-dead mutant, recovered HTLV-1 transcription. Moreover, the addition of the CDK9 inhibitor flavopiridol blocked Tax transactivation in vitro and in vivo. Interestingly, we found that Tax regulates CDK9 kinase activity through a novel autophosphorylation pathway.

FIG. 1.

CDK9 kinase activity is required for Tax transactivation in vivo and in vitro. (A) Tax transactivation is decreased by CDK9 RNAi treatment and restored by CDK9 overexpression. HeLa cells were transfected with CDK9 siRNA or control CAT siRNA, and Tax transactivation was then assayed by analysis of HTLV luciferase activity. Each result was shown as an average of four experiments with the standard error indicated (error bars). +, presence; −, absence. (B) Western blot analysis of CDK9 in whole-cell lysates from HeLa cells subjected to RNAi treatments. +, presence; −, absence. nt, nucleotide. (C) CDK9 kinase activity is required for Tax transactivation in vitro. P-TEFb containing WT or a kinase-dead D167N CDK9 was added to the CDK9-depleted extracts as indicated. The in vitro transcription assays were performed by incubating HTLV-1 templates with these extracts. The radiolabeled transcripts were then analyzed by electrophoresis on 6% denaturing polyacrylamide gels followed by autoradiography. nt, nucleotide. (D) CDK9 inhibitor flavopiridol (flavo) inhibits Tax transactivation in vitro. In vitro transcription assays were performed by incubating HTLV-1 templates with HeLa nuclear extracts in the absence (−) or presence (+) of Tax. Flavopiridol was added to the transcription reactions to give the indicated concentrations. The radiolabeled transcripts were fractionated by electrophoresis on 6% denaturing polyacrylamide gels and detected by autoradiography. (E) Flavopiridol inhibits Tax transactivation in vivo. pA-18G-BHK-21/Tax cells were cultured in the absence or presence of flavopiridol for 2 days. Cell cultures were then rinsed twice with PBS and lysed. β-Galactosidase activity was measured with the Galacto-light system (Applied Biosystems). Each result was shown as an average of four experiments with the standard error indicated (error bars). (F) Effect of flavopiridol on cell growth. pA-18G-BHK-21/Tax cells were cultured in the absence or presence of flavopiridol for 2 days. The number of viable cells was then determined by the MTT method (Sigma). Each result was shown as an average of four experiments with the standard error indicated (error bars).

FIG. 2.

Recruitment of P-TEFb to the HTLV-1 PICs in the presence of Tax and interaction of Tax with P-TEFb in vitro and in vivo. (A) P-TEFb is recruited to the HTLV-1 PICs in the presence (+) of Tax. HTLV-1 PICs were assembled by incubating biotinylated HTLV-1 templates with HeLa nuclear extracts in the absence (−) or presence of Tax and then purified with streptavidin-coated magnetic beads. The protein components of the purified PICs were analyzed by Western blotting with α-CDK9 or α-CDK7 antibody. (B and C) P-TEFb interacts with the carboxyl terminus of Tax. A total of 400 ng of GST-Tax, GST-Tax truncations, GST-M22, GST-M47, or GST was incubated with 200 ng of purified p-TEFb proteins in 50 μl GST-binding buffer at 4°C for 4 h. Glutathione-Sepharose (50% slurry; Amersham), precoated with 2 μg/μl of BSA, was then added and incubated overnight at 4°C. Complexes bound on glutathione-Sepharose beads were washed four times with GST washing buffer. Components of complexes were subjected to 4 to 20% SDS-PAGE and analyzed by Western blotting using α-CDK9 antibody. (D) Tax interacts with P-TEFb in vivo. Nuclear extracts from HTLV-1-transformed C81 cells were immunoprecipitated with different antibodies. Immunoprecipitates were then analyzed by Western blotting with α-CDK9 (upper panel) or α-Tax (lower panel) antibody. (E) Colocalization of Tax and P-TEFb. HeLa or HCT116 cells were transfected with pcTax plasmid. Cells were fixed and then immunostained with α-Tax and/or α-CDK9 antibodies.

FIG. 3.

P-TEFb associates with HTLV-1 transcription complexes by ChIP assays. The SP cell, which contains a single active integrated copy of the HTLV-1 proviral genome, was treated with flavopiridol (50 nM) overnight and subjected to ChIP assays. Antibodies specific for Tax, CDK7, CDK9, or RNAP II were used for immunoprecipitation. Real-time PCRs were then carried out to analyze precipitated DNA using primers specific for the promoter region (LTR) and the Tax coding region. The antibodies for immunoprecipitation were labeled on the x axis. The y axis represents the percentage of input material. Each ChIP result was shown as an average of four experiments with the standard error indicated (error bars).

FIG. 4.

Tax regulates CDK9 kinase activity in vitro. (A) Effect of Tax on CDK9 kinase activity. In vitro kinase assays were performed by incubating GST-CTD and P-TEFb with [γ-³²P]ATP in the absence (−) or presence of Tax or CREB. The phosphorylated GST-CTD was precipitated with glutathione-Sepharose beads and fractionated by electrophoresis on 8% SDS-polyacrylamide gels followed by autoradiography. The hypophosphorylated (CTDa) and hyperphosphorylated (CTDo) forms of CTD are indicated. (B) Tax inhibits Ser 2 phosphorylation of CTD by P-TEFb. In vitro kinase assays were performed by incubating GST-CTD and P-TEFb with 100 μM ATP in the absence or presence of Tax or CREB. CTD phosphorylation was detected with the H5 antibody which specifically recognizes phosphorylated Ser 2. The hyperphosphorylated (CTDo) form of CTD is indicated. (C) Effect of Tax on CDK9 autophosphorylation. In vitro kinase assays were performed by incubating P-TEFb with [γ-³²P]ATP in the absence or presence of Tax or CREB. ³²P-labeled CDK9 was immunoprecipitated with α-CDK9 antibody and analyzed by electrophoresis on 4 to 20% SDS-polyacrylamide gels followed by autoradiography.

FIG. 5.

Mechanism of CDK9 regulation by Tax. (A) Alignment of CDK9 to other CDKs. (B) Effect of CDK9 T29A or T29E on CTD phosphorylation. In vitro kinase assays were performed by incubating GST-CTD and P-TEFb with [γ-³²P]ATP in the absence (−) or presence of Tax. The phosphorylated GST-CTD was precipitated with glutathione-Sepharose beads and fractionated by electrophoresis on 8% SDS-polyacrylamide gels followed by autoradiography. The hypophosphorylated (CTDa) and hyperphosphorylated (CTDo) forms of CTD are indicated. (C) Effect of CDK9 T29A or T29E on Ser 2 phosphorylation of CTD. In vitro kinase assays were performed by incubating GST-CTD and P-TEFb with 100 μM ATP in the absence (−) or presence of Tax. CTD phosphorylation was detected with the H5 antibody which specifically recognizes phosphorylated Ser 2. The hyperphosphorylated (CTDo) form of CTD is indicated. (D) Effect of CDK9 T29A or T29E on CDK9 autophosphorylation. In vitro kinase assays were performed by incubating P-TEFb with [γ-³²P]ATP in the absence (−) or presence of Tax. ³²P-labeled CDK9 was immunoprecipitated with specific α-CDK9 antibody and analyzed by electrophoresis on 4 to 20% SDS-polyacrylamide gels followed by autoradiography. (E) Western blot analysis of CDK9 WT and T29A and T29E mutants. (F) Mutation of T29 in CDK9 did not alter the binding affinity of P-TEFb with Tax. A total of 400 ng of GST-Tax or GST was incubated with 200 ng of purified p-TEFb proteins in 50 μl GST-binding buffer at 4°C for 4 h. Glutathione-Sepharose (50% slurry; Amersham) pre coated with 2 μg/μl of BSA was then added and incubated overnight at 4°C. Complexes bound on glutathione-Sepharose beads were washed four times with washing buffer. Components of complexes were subjected to 4 to 20% SDS-PAGE and analyzed by Western blotting using α-CDK9 antibody.

FIG. 6.

Effect of CDK9 inhibition on Tax transactivation. P-TEFb containing WT, T29A, or T29E CDK9 was added back to CDK9-depleted extracts. The in vitro transcription assays were then performed by incubating HTLV-1 templates with these extracts. The radiolabeled transcripts were fractionated by electrophoresis on 6% denaturing polyacrylamide gels and detected by autoradiography. +, with; −, without.

FIG. 7.

A model for CDK9 regulation by phosphorylation in Tax transactivation.