The HTLV-1 tax protein cooperates with phosphorylated CREB, TORC2 and p300 to activate CRE-dependent cyclin D1 transcription

Abstract

Adult T-cell leukemia/lymphoma is a fatal malignancy etiologically linked to infection with the human T-cell leukemia virus (HTLV-1). The virally encoded oncoprotein Tax activates the transcription of HTLV-1 and cellular genes by cooperating with cellular transcription factors. Cyclin D1 is a pivotal regulator of cell cycle progression, and increased expression strongly correlates with malignant transformation. Here, we characterize the mechanism of Tax transactivation of cyclin D1. We find that cyclin D1 transcript levels are elevated in HTLV-1 infected cells and that Tax physically associates with the cyclin D1 gene in vivo. Tax binds the cyclin D1 promoter-proximal cyclic AMP response element (CRE) in the presence of phosphorylated CREB (pCREB) in vitro, and together the Tax-pCREB complex recruits the cellular co-activator p300 to the promoter through this unconventional Tax-responsive element. We further show that the transducer of regulated CREB 2 (TORC2) cooperates with Tax to further enhance p300 recruitment to the cyclin D1 promoter in vitro. Tax and TORC2 in combination stimulate cyclin D1 expression in vivo, demonstrating the functional outcome of the binding interactions. Together, our findings support a model in which Tax-induced accumulation of cyclin D1 shortens the G1 phase of the cell cycle, promotes mitotic replication of the virus, and drives selection and expansion of malignant T-cells.

Figure 1

The human cyclin D1 gene is transcriptionally activated by Tax. (a) Cyclin D1 transcript levels are elevated in HTLV-1 infected T-cell lines. Cyclin D1 RNA from uninfected (Jurkat, CEM) and HTLV-1 infected (SLB-1, C81 [C816645]) cells was amplified by RT-PCR and subjected to agarose gel electrophoresis to determine relative transcript levels. Thymidine kinase (TK) levels are shown as an internal control. (b) Schematic diagram of cyclin D1 promoter. Relevant cis-acting elements and the deletion constructs used in the assay are shown. (c) Tax transactivates cyclin D1 through the −66 proximal promoter. Jurkat cells were cotransfected with the indicated deletion constructs of the human cyclin D1 promoter/Luc (100 ng) (Albanese et al., 1995), in the absence or presence of a Tax expression plasmid (100 ng) (pSG-Tax). Luciferase activity was measured 24 h after transfection and fold-induction was quantified. The graph represents the average of three independent experiments. (d) The NF-κB and ATF/CREB pathways are utilized in Tax activation of cyclin D1. Jurkat cells were transiently transfected with the cyclin D1 −66/Luc reporter in the absence or presence of expression plasmids for wild-type Tax (pSG-Tax), or the M22 or M47 Tax mutants (pSG-M22, pSG-M47) as indicated. M22 Tax is defective for activation through the NF-κB pathway and M47 Tax is defective for activation through ATF/CREB proteins (Smith and Greene, 1990).

Figure 2. Chromatin immunoprecipitation assays reveal that Tax is associated with the cyclin D1 promoter in HTLV-1 infected cells

Chromatin from Jurkat (uninfected) and SLB-1 (infected) cells was immunoprecipitated using antibodies against the indicated proteins. Graphical representation of the ChIP data following immunoprecipitation and real-time PCR. Tax binding at the IL-2 promoter in SLB-1 cells was used as a negative control. Data shown represents the average of three independent experiments.

Figure 3. The CRE is required for Tax binding at the cyclin D1 promoter in vitro

(a) Schematic diagram of the immobilized cyclin D1 core promoter template. The cyclin D1 core promoter was bound to streptavidin-coated magnetic beads via incorporation of a biotin group at the upstream end of the fragment. (b) Tax interacts with the cyclin D1 core promoter. The immobilized template assay was performed with a wild-type (wt) cyclin D1 promoter fragment or a mutant fragment carrying substitutions in both the κB and CRE binding sites (ΔCRE/ΔκB) (2 pmol each). Binding reactions were performed with CEM nuclear extract (NE) (300 ug) in the absence or presence of purified Tax (10 pmol). Samples were analyzed by western blot using antibodies against Tax or CREB, as indicated. (c) The CRE is required for Tax binding to the cyclin D1 promoter. The immobilized template assay was performed with the wild-type (wt) or mutant cyclin D1 promoter fragments (ΔCRE, ΔκB), as indicated. Binding reactions were performed with nuclear extract and purified Tax, as described in panel B, and analyzed by western blot using an anti-Tax antibody.

Figure 4. Tax and pCREB together recruit p300 to the cyclin D1 promoter

The immobilized cyclin D1 core promoter (5 pmol) was incubated with purified full-length p300 (4 pmol) in the presence of purified Tax, CREB (C), or Ser¹³³-phosphorylated CREB (pC) (10 pmol each), as indicated. Protein binding was detected by western blot using anti-Tax, anti-CREB, anti-Ser¹³³CREB, or anti-p300 antibody, as indicated. As a positive control, analysis of protein binding to the immobilized vCRE was performed in parallel with the cyclin D1 core promoter.

Figure 5. Tax forms a stable complex with pCREB and KIX on the cyclin D1 CRE

(a) The cyclin D1 CRE supports Tax binding in the presence of pCREB and KIX. Electrophoretic mobility shift assays (EMSA) were performed using purified pCREB (24 nM), GST-KIX (5 μM), and increasing concentrations of Tax (0.8, 3.1, 12, 50, 200, 400 nM). Proteins were incubated with the ³²P-labeled HTLV-1 vCRE or cyclin D1 CRE (0.15 nM), as indicated. Binding reactions were resolved by 5% nondenaturing PAGE and analyzed by autoradiography. Saturating amounts of GST-KIX were used to visualize the Tax-dependent transition from the ternary complex (pCREB/KIX/CRE) into the slower migrating quaternary complex (Tax/pCREB/KIX/CRE). A dashed line was added to assist in evaluating the migration of the nucleoprotein complexes. The reduced amount of Tax/pCREB/DNA complex shown in lane 4 is anomalous, and thus not representative of the band typically produced by Tax/pCREB binding to the vCRE (for example, see (Kim et al., 2007). (b) Relative binding affinities of Tax for the cyclin D1 CRE and vCRE in the presence of pCREB and KIX. From the data shown in panel A, we quantified and plotted the percent quaternary complex formed with each probe against the total Tax concentration (nM) in each reaction. The concentration of Tax at the mid-point (50%) of the transition from the ternary to the quaternary complex was used to determine the apparent Kd of Tax for vCRE/pCREB/KIX and cyclin D1 CRE/pCREB/KIX, as indicated. The average values from three independent experiments are shown.

Figure 6. Tax and TORC2 cooperate to activate cyclin D1 via enhanced recruitment of p300

(a) TORC2 stimulates cyclin D1 expression and facilitates Tax transactivation. Jurkat cells were co-transfected with the cyclin D1 −66/Luc reporter construct (100 ng) and Tax and/or TORC2 expression plasmids (100 ng), as indicated. (b) Tax and TORC2 cooperate to enhance p300 recruitment. Immobilized cyclin D1 core promoter DNA (5 pmol) (see Fig. 3A) was incubated with purified full-length p300 (4 pmol) and pCREB (10 pmol), in the absence or presence of purified Tax and TORC2 (10 pmol each). Binding reactions were analyzed by western blot using antibodies against the indicated proteins. (c) The cyclin D1 CRE is required for Tax-containing complex formation. The immobilized template assay was performed using the wild-type (wt) and ΔCRE cyclin D1 promoter fragments, as described in panel B.

Figure 7. Model depicting the Tax-containing transcription factor complex on the cyclin D1 CRE

Incorporation of Tax into the pCREB/TORC2 complex at the cyclin D1 promoter-proximal CRE promotes enhanced p300 binding, with subsequent association of RNA polymerase II (RNAPII) and activation of the cyclin D1 gene. The schematic shows Tax association with the promoter DNA, however, the extent of Tax-DNA interaction is not known.