Binding of Tat to TAR and recruitment of positive transcription elongation factor b occur independently in bovine immunodeficiency virus

Abstract

Transcriptional transactivators (Tat) from many lentiviruses interact with their cognate transactivation response RNA structures (TAR) to increase rates of elongation rather than initiation of transcription. For several of them, the complex of Tat and a species-specific cyclin T1 must be formed before the binding to TAR can occur with high affinity and specificity. In sharp contrast, Tat from the bovine immunodeficiency virus (BIV) binds to its TAR without the help of the cyclin T1. This binding depends on the upper stem and 5' bulge, but not the central loop in TAR. Moreover, cyclins T1 from different species can mediate effects of this Tat in cells. Unlike the situation with other lentiviruses, Tat transactivation can be rescued simply by linking a heterologous promoter to TAR in permissive cells. Thus, lentiviruses have evolved different strategies to recruit Tat and the positive transcription elongation factor b to their promoters, and interactions between Tat and TAR are independent from those between Tat and the cyclin T1 in BIV.

FIG. 1

The N-terminal 300 residues in hCycT1, eCycT1, and mCycT1 bind to bTat in vitro. ³⁵S-labeled hCycT1, eCycT1, and mCycT1 were incubated with GST alone (lanes 2, 5, and 8) or with the hybrid GST-bTat protein (lanes 3, 6, and 9) and selected on glutathione-Sepharose beads. Bound cyclins T1 were separated on SDS-PAGE and subjected to autoradiography. The input of each cyclin T1 was equal in all reactions and represented 25% of the amount used for the binding assay (lanes 1, 4, and 7). The arrow to the left points to cyclins T1. Twenty-five percent of the input GST alone (lane 1) and the hybrid GST-bTat protein (lane 2) were comparable and are presented in the Coomassie blue-stained SDS-PAGE panel at the bottom. Arrows to the left indicate the presence of the hybrid proteins.

FIG. 2

bTat transactivation in lapine cells is inhibited by low levels of basal transcription from bLTR. (A) Schematic representation of plasmid targets. pBLTRCAT contains the CAT reporter gene under the control of bLTR. pHIVSCATbTAR contains the hLTR linked to bTAR instead of hTAR. pA represents the polyadenylation signal. (B) bTat requires a strong promoter in lapine EREp cells. Cells were transfected with 0.3 μg of pBLTRCAT alone (lane 1, white bar) or together with 0.1 μg of pbTat (lane 2, white bar). pHIVSCATbTAR (0.1 μg) was expressed alone (lane 3, black bar) or together with 0.1 μg of pbTat (lane 4, black bar). Where no bTat was added, the amount of DNA was equilibrated with 0.1 μg of pEFBOS (lanes 1 and 3, white and black bars). The CAT enzymatic activity of pBLTRCAT or pHIVSCATbTAR alone was set to 1. Standard errors of the mean from three independent transfections are shown by error bars.

FIG. 3

Cyclins T1 from different species increase bTat transactivation in serum-starved Cf2Th cells. (A) bTat transactivates bLTR via different cyclins T1. Cells were serum starved before and after transfection (lanes 1, and lanes 3 to 6), before transfection only (lane 7), or grown in the medium with serum (lane 2). pBLTRCAT (0.3 μg) was expressed alone (lane 1, white bar) or together with 0.1 μg of pbTat (lanes 2 to 7). To bTat were added effector plasmids (1.0 μg) phCycT1, peCycT1, and pmCycT1 (lanes 4 to 6, striped bars). In all transfections, the amount of DNA was equilibrated with pEFBOS. Values are as in Fig. 2B. (B) Amounts of exogenously expressed cyclins T1 determined by Western blotting. Numbers under the Western blot correspond to lanes from the transient transfection assay (A).

FIG. 4

bTat binds to the upper stem and 5′ bulge in bTAR without the help of the cyclin T1. Where indicated, bacterially expressed GST, hybrid GST-bTat, and GST-hCycT1 proteins were used in EMSA. α-³²P-labeled wild-type bTAR was present in all reactions. For competition experiments, three different unlabeled competitor bTAR transcripts were used: bTARWT (lane 6), bTARΔS, which is mutated in the upper stem in bTAR (lane 7), and bTARΔL, which is mutated in the central loop in bTAR (lane 8). The resulting RNA-protein complexes were resolved on a 6% nondenaturing polyacrylamide gel and analyzed by autoradiography. Arrows to the right indicate the free bTAR probe and the presence of distinct RNA-protein complexes.

FIG. 5

Model for the formation of the tripartite complex between bTat, P-TEFb, and bTAR. The binding of bTat to bTAR (step 1) and the recruitment of P-TEFb (step 2) occur independently in BIV. The ARM from bTat interacts with the upper stem and 5′ bulge, but not the central loop, in bTAR. The activation domain from bTat binds to the cyclin T1. The binding of bTat to bTAR is of sufficient affinity and specificity to recruit P-TEFb from different species to bLTR for efficient elongation of transcription.