Report

Role for the Amino-Terminal Region of Human TBP in U6 snRNA Transcription

See allHide authors and affiliations

Science  21 Feb 1997:
Vol. 275, Issue 5303, pp. 1136-1140
DOI: 10.1126/science.275.5303.1136

Abstract

Basal transcription from the human RNA polymerase III U6 promoter depends on a TATA box that recruits the TATA box-binding protein (TBP) and a proximal sequence element that recruits the small nuclear RNA (snRNA)-activating protein complex (SNAPc). TBP consists of a conserved carboxyl-terminal domain that performs all known functions of the protein and a nonconserved amino-terminal region of unknown function. Here, the amino-terminal region is shown to down-regulate binding of TBP to the U6 TATA box, mediate cooperative binding with SNAPc to the U6 promoter, and enhance U6 transcription.

The TATA box-binding protein is a central transcription factor required for transcription by all three RNA polymerases. The highly conserved COOH-terminal domain performs all of the TBP functions examined so far, including binding to the TATA box and interacting with TBP-associated factors, general transcription factors, and activators (1). In vivo, this domain can functionally replace the full-length protein for all promoters tested in mammalian systems (2); and in yeast, strains carrying a TBP missing the NH2-terminal domain are viable (3). The role of the nonconserved NH2-terminal domain is unknown.

Most RNA polymerase III promoters consist of gene-internal elements and recruit TBP as part of the TBP-containing transcription factor IIIB (TFIIIB), through protein-protein interactions with the DNA-binding TFIIIC (4). However, in some unusual cases, exemplified by the human U6 snRNA promoter, the promoter elements are located upstream of the transcription start site (5) and appear to recruit neither TFIIIC (6) nor the same TFIIIB complex recruited by RNA polymerase III promoters with gene-internal elements (7). Instead, the U6 promoter contains two basal promoter elements, a proximal sequence element (PSE), which recruits a multisubunit complex referred to as the SNAP complex (SNAPc) or PSE transcription factor (8) and a TATA box, which recruits TBP (9).

To test whether SNAPc and TBP might bind to the U6 promoter cooperatively, as suggested by previous results (10), we performed an electrophoretic mobility shift analysis (EMSA) (11) with a probe corresponding to the basal human U6 promoter but containing the mouse U6 PSE, which is a higher affinity binding site for SNAPc. A barely detectable complex formed when full-length human TBP fused to glutathione- S-transferase (G-hTBP), was incubated with the probe, but a prominent complex formed when a fraction highly enriched in SNAPc was incubated with the probe (Fig. 1A). When G-hTBP and SNAPc were co-incubated with the probe, a prominent complex of slower mobility was observed, and its formation was dependent on an intact PSE and an intact TATA box (Fig. 1A). This complex contained both G-hTBP and SNAPc, as determined by antibody supershift experiments (12). Similar results were obtained when hTBP was used instead of the G-hTBP fusion protein (12). Thus, SNAPc appeared to enhance binding of TBP to the U6 promoter probe. The COOH-terminal core of TBP lacking the NH2-terminal region (G-hTBPΔN) bound much more efficiently to the TATA box than did G-hTBP on its own, and G-hTBPΔN binding was not substantially increased in the presence of SNAPc bound to the PSE (Fig. 1A). Thus, although both factors could co-occupy the U6 probe, in this case they did not influence each other's ability to bind DNA. Similarly, in assays with the TATA box of the adenovirus 2 major late (Ad2 ML) promoter, G-hTBP bound more weakly than did G-hTBPΔN (12). These results suggest that the NH2-terminal region of human TBP down-regulates binding to the TATA box and allows SNAPc to mediate recruitment of TBP to the U6 TATA box.

Fig. 1.

SNAPc-enhanced binding of full-length human TBP but not TBPΔN to the TATA box. (A) The EMSA was performed with a probe containing a high-affinity PSE (mouse U6 PSE) and a TATA box, or mutated versions thereof, and G-hTBP (lanes 2, 3, 5, and 6), or G-hTBPΔN (lanes 9 through 12) or SNAPc or both, as indicated above the lanes. Proteins were obtained as described (11). Each titration contained approximately 80 and 160 ng of TBP. The positions of the free probe and the protein-probe complexes are indicated. (B) The binding reactions for the DNase I footprinting assay contained probe alone (lane 1), G-hTBP alone (lanes 2 through 4; 30, 70, and 110 ng, respectively), the same amounts of G-hTBP and a constant amount of SNAPc (lanes 5 through 7) or SNAPc alone (lane 8). (C) The binding reactions contained probe alone (lane 1 and 11), G-hTBPΔN alone (lanes 2 through 5; 15, 30, 70 and 110 ng, respectively), the same amounts of G-hTBPΔN and a constant amount of SNAPc (lanes 6 through 9), or SNAPc alone (lane 10). The locations of the mouse U6 PSE and the TATA box on the probe are shown on the left. The dashed line indicates vector sequences on the probe. The arrow marks a hypersensitive site observed in the presence of SNAPc.

To determine whether the same effect could be observed at equilibrium in solution, we performed a deoxyribonuclease I (DNase I) footprinting experiment (13). Addition of increasing amounts of human G-hTBP to the U6 probe did not result in any detectable footprint, indicating that under the conditions used, full-length G-hTBP did not bind stably to the probe (Fig. 1B). In contrast, in the presence of SNAPc, which gave a footprint over the high-affinity mouse U6 PSE on its own, a clear footprint over the TATA box was observed, confirming that SNAPc recruits and stabilizes TBP on the TATA box. Recruitment of TBP by SNAPc was also observed with non-glutathione-S-transferase (GST) TBP purchased from Promega (12). In contrast, G-hTBPΔN generated a clear footprint over the TATA box in the absence of SNAPc, and this footprint was not increased in the presence of SNAPc (Fig. 1C). The results confirm that G-hTBPΔN binds more efficiently to the TATA box than does G-hTBP and cannot be further stabilized on the DNA by SNAPc.

To test whether TBP stabilized SNAPc on the PSE, we assayed binding of SNAPc and G-hTBP to the wild-type human U6 promoter, which contains a PSE with very low affinity for SNAPc. The SNAPc bound efficiently to the mouse U6 PSE but not to the human U6 PSE (Fig. 2). However, upon addition of increasing amounts of G-hTBP, which did not bind effectively to the probe on its own, we observed very efficient formation of a SNAPc/G-hTBP complex on the human U6 promoter. Remarkably, G-hTBPΔN bound very efficiently to the probe but was unable to recruit SNAPc to the PSE (Fig. 2). Yeast TBP, whose NH2-terminal domain bears no homology to human TBP, behaved like hTBPΔN (12). Thus, the NH2-terminal region of human TBP not only down-regulates binding of full-length TBP to the TATA box, but also mediates cooperative binding with SNAPc.

Fig. 2.

Enhanced recruitment of SNAPc to a low-affinity human PSE by full-length hTBP but not hTBPΔN. The EMSA was performed with a U6 probe containing either the high-affinity mouse U6 PSE (lane 1) or the low-affinity human U6 PSE (lanes 2 through 7) and a TATA box. Either G-hTBP (lanes 3, 4, and 7) or G-hTBPΔN (lanes 5 and 6) and a constant amount of SNAPc, as indicated above the lanes, was used. Each titration contained 40 and 100 ng of TBP.

The nonconserved NH2-terminal region of TBP can be divided into three segments: (i) a run of glutamines at position 55 through 95 (segment II), (ii) the segment that precedes it (segment I, amino acids at position 1 through 54), and (iii) the segment that follows it (segment III, amino acids at position 96 through 158) (Fig. 3A). We first generated two constructs that either contained only the third segment (ΔN+III) or only the first two segments (ΔN+I+II) of the NH2-terminal region, and compared them with full-length TBP and hTBPΔN in their ability to recruit SNAPc to the low-affinity human PSE. We tested both GST TBP fusion proteins and non-GST TBP proteins (Fig. 3B). SNAPc alone bound relatively weakly to the PSE, and full-length G-hTBP or hTBP did not show detectable binding. However, SNAPc and G-hTBP together or SNAPc and hTBP together bound very efficiently to the human U6 promoter. Both G-hTBPΔN and hTBPΔN bound to the probe on their own but did not recruit SNAPc to the PSE. G-TBP and TBP proteins containing only the third segment of the NH2-terminal region (ΔN+III) also bound to the TATA box on their own and also failed to enhance recruitment of SNAPc to the PSE. In contrast, proteins containing only the first two segments (ΔN+I+II) did not bind efficiently on their own but were able to bind cooperatively with SNAPc (Fig. 3B). The TBP molecules lacking the GST moiety were also tested for their ability to bind to the Ad2 ML TATA box, and in this assay too, full-length hTBP and ΔN+I+II bound less efficiently to the TATA box than did hTBPΔN and ΔN+III (12). These results suggest that the first two segments down-regulate binding of human TBP to the TATA box and recruit SNAPc to the PSE.

Fig. 3.

(A) Structure of NH2-terminal deletion mutants of hTBP and their effect on DNA binding, ability to interact with SNAPc, and basal U6 and Ad2 ML transcription. The positions of segments I, II, and III in wild-type TBP and the various TBP mutants are indicated. The EMSA and transcription results in (B) and (C) are summarized on the right. (B) The first two segments of the NH2-terminal domain of TBP can recruit SNAPc to the PSE. The upper panel shows an experiment performed with GST fusion proteins and the lower panel shows an experiment performed with TBP molecules lacking the GST moiety. The NH2-terminal deletion mutants of hTBP used are indicated above the lanes. The positions of the free probes and of the various DNA-protein complexes are indicated. In the upper panel, lanes 4, 7, 10, 13, 16, and 19 contained 40 ng of G-TBP and lanes 3, 5, 6, 8, 9, 11, 12, 14, 15, 17, 18, and 20 contained 80 ng of G-TBP. In the lower panel, lanes 4, 7, 10, and 13 contained 8 ng of hTBP and lanes 3, 5, 6, 8, 9, 11, 12, and 14 contained 16 ng of hTBP. The SNAPc preparation used in the upper panel was very concentrated; the autoradiogram is therefore a much shorter exposure than that in Fig. 2, hence the much weaker signals corresponding to the G-hTBPΔN complex. (C) The first two segments of the NH2-terminal domain of hTBP are required for efficient in vitro transcription from the human U6 but not from the Ad2 ML promoter. Mock-depleted (lane 1) or TBP-depleted (lanes 2 through 20) whole-cell extracts (10 μl) were programmed with the U6 template hU6/Hae/RA.2 (22) containing a mutated enhancer (upper panel) or with the p119ML(C2A) template (22) containing the Ad2 ML promoter in front of a G-less cassette (lower panel). Where indicated, the reactions were supplemented with 20, 60, and 180 ng of G-hTBP or truncated versions thereof, as illustrated above the lanes. RNA derived from the U6 template was analyzed by ribonuclease T1 protection with the U6/RA.2/143 probe, according to Lobo et al. (7). A correctly initiated RNA signal from the U6 promoter and from the Ad2 ML promoter is indicated. Transcription from U6 was severely reduced by mutations in the PSE or the TATA box (12).

We then tested two GST fusion constructs containing either segment I or segment II fused directly to the COOH-terminal domain (constructs ΔN+I and ΔN+II; Fig. 3A). Whereas G-hTBPΔN+I was unable to bind effectively to the U6 TATA box on its own (like full-length and ΔN+I+II TBP), G-hTBPΔN+II could bind to the TATA box (Fig. 3B). These results suggest that segment I down-regulates binding of TBP to the TATA box. Both proteins, however, were capable of binding cooperatively with SNAPc to the U6 promoter, although less efficiently than either G-hTBP or G-hTBPΔN+I+II (Fig. 3B). This suggests that segments I and II both contribute to cooperative binding with SNAPc.

To determine whether the NH2-terminal region of TBP plays a role in snRNA gene transcription, we compared the ability of G-hTBP and the various NH2-terminal truncations to direct basal RNA polymerase III transcription from the human U6 promoter, which has a low-affinity PSE, and basal RNA polymerase II transcription from the Ad2 ML promoter, in an extract immunodepleted of TBP (14). Depletion of TBP inhibited both U6 and Ad2 ML transcription, but addition of increasing amounts of G-hTBP or any of its mutant derivatives restored transcription from the Ad2 ML promoter (Fig. 3C). This indicates that none of the TBP mutants were inherently defective, including those that bound only weakly to the TATA box, and also suggests that the TBP mutants that do not bind DNA effectively can be stabilized on the Ad2 ML TATA box by other members of the Ad2 ML initiation complex, such as TFIIA or TFIIB. Indeed, all TBP mutants bound to the Ad2 ML TATA box in an EMSA in the presence of TFIIA or TFIIB (12). Transcription from the U6 promoter could be restored very efficiently by addition of increasing amounts of G-hTBP or G-TBPΔN+I+II and somewhat less efficiently by addition of increasing amounts of G-hTBPΔN+I or G-hTBPΔN+II. Remarkably however, neither G-hTBPΔN+III nor G-hTBPΔN was capable of restoring U6 transcription at any of the concentrations tested (Fig. 3C). Thus, the abilities of the different mutant TBPs to support basal transcription from the human U6 promoter correlate with their abilities to recruit SNAPc to the human PSE, and not with their intrinsic abilities to bind DNA or to restore basal RNA polymerase II transcription. These results suggest that recruitment of SNAPc by the NH2-terminal domain of TBP is essential for basal transcription from the human U6 promoter, which has a low-affinity PSE. The results may also explain the observation of Lescure et al. that a monoclonal antibody directed against an epitope corresponding to the junction between regions I and II inhibits RNA polymerase III transcription from a U6 promoter but not from TATA-less promoters in vitro (15).

How does the NH2-terminal domain of TBP down-regulate binding to the TATA box and recruit SNAPc to the PSE? Perhaps this domain masks part of the conserved COOH-terminal DNA-binding domain and undergoes a conformational change during cooperative binding of SNAPc and TBP to the U6 promoter, so that it becomes engaged in direct protein-protein interactions with SNAPc. This putative interaction would be dependent on both factors binding to DNA, because both an intact PSE and an intact TATA box are required for formation of the SNAPc/TBP complex. Are the functions we describe restricted to human TBP? Although the TBP NH2-terminal domain has diverged much more than the COOH-terminal domain, it is highly conserved among vertebrate TBPs (16, 17). Segments I and III are 78 and 73% identical, respectively, among humans (18), mice (16), hamsters (19), two different vipers (whose TBPs are identical except for the number of Gln residues in segment II) (20), and Xenopus laevis (17) TBPs, with large blocks of amino acids conserved 100% (Fig. 4). The main variation is the length of segment II, which indeed is polymorphic in humans (21). Similarly, snRNA promoters have diverged widely between, for example, yeast and humans, but are quite conserved among vertebrates (5). Perhaps the TBP-SNAPc interaction, like the NH2-terminal region of TBP, has remained conserved among vertebrate species.

Fig. 4.

Deduced amino acid sequences from the NH2-terminal regions of human (18), mouse (16), hamster (19), viper (20), and Xenopus (17) TBP. Identical residues are shown against a black background. A dash indicates a gap. Abbreviations for the amino acid residues are as follows: A, Ala; C, Cys; D, Asp; E, Glu; F, Phe; G, Gly; H, His; I, Ile; K, Lys; L, Leu; M, Met; N, Asn; P, Pro; Q, Gln; R, Arg; S, Ser; T, Thr; V, Val; W, Trp; and Y, Tyr.

REFERENCES AND NOTES

  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
View Abstract

Stay Connected to Science

Navigate This Article