Nov 28, 1994 - Communicated by Robert G. Roeder, The Rockefeller University, New York; NY November 28 ..... Johnson, P. F. & McKnight, S. L. (1989) Annu.
Proc. Natl. Acad. Sci. USA Vol. 92, pp. 1955-1959, March 1995 Biochemistry
Cooperation between core promoter elements influences transcriptional activity in vivo JOHN COLGAN* AND JAMES L. MANLEY Department of Biological Sciences, Columbia University, New York, NY 10027
Communicated by Robert G. Roeder, The Rockefeller University, New York; NY November 28, 1994
(18). When both elements are present in the same promoter, the rate of complex formation is similar to that seen with a TATA-only promoter, but the absolute levels of basal and activated transcription are increased, suggesting a cooperative interaction between the two elements (6, 18-21). Evidence for interactions between factors bound at TATA and Inr elements also comes from experiments showing that TBP and TFII-I can bind cooperatively to Inr-containing DNA fragments (22). These results support the idea that, at least in vitro, Inr can influence transcription complex formation, possibly by increasing or altering the association of one or more of the general transcription factors. In this study, a transient cotransfection assay was used to examine the interactions between prototypical TATA and Inr sequences in vivo. In most contexts, the TATA box was the dominant element, but the addition of an Inr significantly enhanced the response to activators. An interesting exception was the activator Spl, which functioned most effectively with the Inr, and the addition of a TATA box had little effect. However, in all cases the presence of the TATA and Inr elements together facilitated an interaction involving the general transcription factor TFIIB.
ABSTRACT Core promoters for RNA polymerase II frequently contain either (or both) of two consensus sequence elements, a TATA box and/or an initiator (Inr). Using test promoters consisting of prototypical TATA and/or Inr elements, together with binding sites for sequence-specific activators, we have analyzed the function of TATA and Inr elements in vivo. In the absence of activators, the TATA element was significantly more active than the Inr, and the combination of elements was only slightly more effective than the TATA-only promoter. In the presence of any of several coexpressed activator proteins, the TATA element was again most active, but here addition of the Inr allowed significant increases in activity, indicating a cooperative interaction between the two elements. An interesting exception was observed with the activator Spl, which was more effective with the Inr-only promoter, and addition of a TATA box did not enhance activity. Finally, in all cases the TATA plus Inr promoters were found to be partially or completely resistant to the dominant negative effects of a transcription factor TFIIB mutant previously shown to interfere with expression from TATA-only promoters. This result strengthens the conclusion that TATA and Inr elements can cooperate in vivo.
MATERIALS AND METHODS
Promoters for genes transcribed by RNA polymerase II (RNAP II) commonly contain two classes of DNA sequence elements. The first, the core or basal promoter elements, consists of sequences that alone can support transcription in vitro (1, 2). The second class includes sequences recognized by gene-specific regulatory proteins, which function to modulate the level of transcription initiated from a core element (3, 4). Two types of core promoter elements have been identified, the TATA box, which is found 25-30 bp upstream of the start site of transcription, and the initiator (Inr) (5), which fits a loose consensus sequence that overlaps the transcription start site (6). RNAP II promoters may contain one or both of these core elements (1, 2). Factors capable of recognizing both types of core promoter elements have been identified. All TATA elements are apparently bound by a single factor, the TATA-binding protein (TBP), which is a subunit of the general transcription factor TFIID (reviewed in ref. 7). TFIID consists of TBP and a number of associated factors calls TAFs, which are essential for activated, but not basal, transcription in vitro (eg, 8, 9). In contrast to TATA sequences, a number of factors appear capable of binding to Inrs, including a component of the TFIID complex (10, 11), RNAP 11 (12), and several possibly gene-specific regulatory proteins such as TFII-I (13) and YY1 (14). Despite the lack of a TATA box, TFIID is also essential for transcription from Inrcontainina promoters in vitro (15-17). Biochemical studies have uncovered differences in transcription complex assembly that can be attributed to Inr elements. Experiments with crude extracts to support transcription in vitro have indicated that the initial steps in assembly can occur faster on an Inr compared to a TATA box
Recombinant Plasmids. All chloramphenicol acetyltransferase (CAT) plasmids, except those used with Spl, were derived from G5 TATA CAT, which has been described (23). G5 TATA Inr CAT was created by inserting a 23-bp fragment containing the terminal deoxynucleotidyltransferase (TdT) Inr (5, 6) into G5 TATA CAT. G5 CAT and G5 Inr CAT were created by deleting the TATA box. The Spl-responsive CAT plasmids, Act GAL4ftzQ, Act Spl, and Act TFIIBAC202, as well as all other expression plasmids, have been described (23-27). DNA Transfection and Transient Expression Assays. Transient cotransfections were performed essentially as described (27). DNA precipitates contained 2 jig of the appropriate reporter plasmid, 2 ,ug of copia-lacZ internal control plasmid, the amount of expression vector indicated in the figures, and GEM3 as needed to reach a total of 10 ,g. Each experiment was repeated at least four times. Variations in transfection efficiencies within a given experiment were almost always
TATA + Inr
-fold inhibition by Act flu-TFIIBAC202
FIG. 4. A TATA plus Inr promoter is resistant to a dominant negative form of TFIIB. (A) Effects on basal and GAL4-ftzQ-activated promoter activity of 2 ,ug of Act TFIIB/C202. Reporter plasmids are shown on the left. Promoter activities are expressed relative to that obtained with G5 TATA CAT alone. Bar graph displays the -fold inhibition due to ActTFIIBAC202 plus low (10 ng) or high (200 ng) amounts of Act GAL4-ftzQ. (B) RNase protection analysis of total RNA isolated from cells transfected with G5 TATA CAT plus Act GAL4-ftzQ with (+) or without (-) ActTFIIBAC202. Mock represents analysis of RNA from cells transfected with GEM3 in place of reporter plasmid.
binding sites are frequently found associated with TATAlacking promoters (30). Given the difference in behavior of the core promoter elements with Spl relative to other activators (or no activators), it was of interest to determine the sensitivity of Splactivated expression to TFIIBAC202. Results of cotransfections with Act TFIIBAC202 (Fig. 5) allow several conclusions. First, expression from both the TATA-only and Inr-only promoters was inhibited by AC202, with the latter being somewhat more sensitive. However, the inhibitions detected were much lower than with GAL-ftzQ and were more similar to those observed with basal expression. Second, the -fold inhibition did not decrease as the Act Spl concentration was increased. This is distinct from what was observed with GAL4-ftzQ and, as discussed below, is consistent with the existence of a direct interaction between TFIIB and GAL4ftzQ but not between TFIIB and Spl. Finally, despite these differences, the TATA plus Inr promoter was almost completely resistant to AC202. These results indicate that the presence of both TATA and Inr elements in a promoter affects interactions involving TFIIB, which, depending on the condiAct Spl, ng: Act flu-TFIIBAC202:
tions, may or may not lead to higher levels of transcription than are obtained with either element alone.
DISCUSSION An important conclusion from our studies is that TATA and Inr elements can cooperate in vivo. This is reflected both in heightened responsiveness to several transcriptional activators and in resistance to a dominant negative TFIIB mutant. These results are consistent with previous in vitro experiments indicating that factors bound at TATA and Inr elements can function together to facilitate transcription complex assembly (19-21) and that an Inr can affect an early step in transcription complex formation (18, 22, 31). The Inr-binding protein TFII-I can bind cooperatively with TBP to DNA fragments containing TATA and Inr elements and, in the absence of a TATA box, both recruit TBP and facilitate binding of TFIIB. TFIID, presumably via a TAF, can also recognize the TdT Inr (10, 11, 17), and this interaction can be influenced by TFIIB (10). Indeed, recognition of the Inr by TFIID is sufficient to direct basal transcription in vitro (17). Although it is unclear whether it is a subunit of TFID or another factor that is responsible for the functions of the Inr we observed, our experiments show that recognition of the Inr in vivo can 10
1230 1210 I
Promoter elements TATA
TATA + Inr
10 15 -fold inhibition by Act flu-TFIIBAC202 5
FIG. 5. Spl functions preferentially with an Inr, but the TATA plus Inr promoter remains resistant to TFIIBAC202. Expression from the indicated reporter plasmids plus Act Spl (0, 10, or 200 ng) and the effect on expression caused by cotransfection of Act TFIIBAC202. All reporter plasmids
contain 6 Spl binding sites. Bar graph displays the -fold inhibition due to transfecting ActTFIIBA&C202 plus 10 or 200 ng of Act Spl. Promoter activities are shown relative to the value from the TATA only plasmid alone, which was set equal to 1.0.
Biochemistry: Colgan and Manley influence a step in transcription complex assembly involving TFIIB, consistent with these in vitro studies. We have shown both here and previously (23) that basal expression from a TATA-only promoter is weakly inhibited by the TFIIB mutant AC202, while expression activated by GAL4-ftzQ is strongly inhibited. Our finding here that the presence of an Inr together with a TATA box reduces or eliminates this inhibition, while at the same time enhancing activation, suggests that the Inr facilitates a step in preinitiation complex assembly such that the complex formed is both refractive to the AC202 mutant and also more responsive to activators. How might an Inr exert such effects? We suggest that the binding of TFIID to a TATA box together with recognition of the Inr by a TAF and/or Inr-binding protein favors the recruitment of full-length TFIIB relative to AC202. Given that the region deleted in AC202 (amino acids 202-315) is important for binding of TFIIB to template-bound TBP in vitro (32-36), an interaction involving the Inr could facilitate binding of full-length TFTIIB by maximizing interactions with this region. In the absence of the Inr, perhaps TFIID interacts less efficiently with the C terminus of TFII3, which allows a fraction of AC202 to enter into preinitiation complexes, rendering them nonproductive. Why does this more efficient recruitment of TFIIB by the TATA plus Inr promoter not lead to a significant enhancement of basal transcription in vivo as it does in vitro? This is likely because a step other than recruitment of TFIIB remains limiting in vivo, an idea consistent with experiments examining the mechanism of activation in vitro (37). However, increased binding of TFIIB can help explain why the TATA plus Inr promoter is generally more responsive to activators. We suggest that recognition of the Inr results in a distinct proteinDNA complex that can cooperate more effectively with activators in the recruitment of TFIIB and/or alter TFIIB's conformation so that it can interact better with RNAP II or other general transcription factors. The fact that TFIIBAC202-mediated inhibition of GAL4ftzQ activation is partially, but not fully, relieved by the Inr is probably due to a combination of two different effects. First, as argued above, factor interactions involving the Inr favor the entry of full-length TFIIB relative to AC202, providing the core promoter with resistance to the mutant and increasing interactions between the activator and promoter-bound TFIIB. Second, we propose that the partial inhibition that remains reflects a direct protein-protein interaction between GAL4-ftzQ and AC202, as suggested by our earlier experiments (23) and more recently demonstrated in vitro (38). Consistent with this, transcription activated by Spl was more weakly inhibited by AC202 when TATA- or Inr-only promoters were used and was almost totally resistant with a TATA plus Inr promoter. We suggest this is because Spl does not appear to interact directly with TFIIB (37), a conclusion supported by our data that increasing concentrations of Spl could not overcome inhibition by AC202. Therefore, like basal expression, Spl-activated expression is not affected by AC202 when both TATA and Inr elements are present. Our data also indicate that Spl can function preferentially through an Inr element. Previous studies involving transfection of mammalian cells suggested that Spl works equally well with TATA or Inr elements (39,40). This apparent discrepancy with our results may reflect differences in cell type and/or the promoter constructs used, and it is possible that our conclusion that Spl functions preferentially through an Inr may apply only in certain contexts. With this in mind, we suggest a possible mechanism for the Spl-Inr interaction. In vitro experiments have suggested that binding of Spl to a specific TAF (TAFI1 110) is likely important for Spl activation (41). We suggest that the conformation of TFIID when bound to the promoter
Proc. Nati Acad Sci. USA 92 (1995)
through an Inr is different than when it is bound to the TATA box and that the configuration of TAFI 110 in the former case is such that its ability to interact with Spl is enhanced. Regardless of the precise mechanism, however, our results indicate that TATA and Inr elements can have distinct, activator-dependent properties in vivo. We are grateful to K. Han for providing plasmids, K. Han and J. D. Kohtz for technical advice and useful discussions, and S. Smale for communicating results prior to publication. This work was supported by National Institutes of Health Grant GM 37971. 1. Smale, S. T. (1994) in Transcription, Mechanisms and Regulation, eds. Conaway, R. C. & Conaway, J. W. (Raven, New York), pp. 63-81. 2. Weis, L. & Reinberg, D. (1992) FASEB J. 6, 3300-3309. 3. Johnson, P. F. & McKnight, S. L. (1989) Annu. Rev. Biochem. 58, 799-839. 4. Tjian, R. & Maniatis, T. (1994) Cell 245, 371-378. 5. Smale, S. T. & Baltimore, D. (1989) Cell 57, 103-113. 6. Javahery, R., Khachi, A., Lo, K., Zenzie-Gregory, B. & Smale, S. T. (1994) Mol. Cell. Biol. 14, 116-127. 7. Hernandez, N. (1993) Genes Dev. 7, 1291-1302. 8. Dynlacht, B. D., Hoey, T. & Tjian, R. (1991) Cell 66, 563-575. 9. Tanese, N., Pugh, B. F. & Tjian, R. (1991) Genes Dev. 5, 2212-2224. 10. Kaufmann, J. & Smale, S. T. (1994) Genes Dev. 8, 821-829. 11. Purnell, B. A., Emanuel, P. A. & Gilmour, D. S. (1994) Genes Dev. 8, 830-842. 12. Carcamo, J., Buckbinder, L. & Reinberg, D. (1991) Proc. Natl. Acad. Sci. USA 88, 8052-8056. 13. Roy, A. L., Meisternerst, M., Pogononec, P. & Roeder, R. G. (1991) Nature (London) 354, 245-248. 14. Seto, E., Shi, Y. & Shenk, T. (1991) Nature (London) 354, 241-245. 15. Pugh, B. F. & Tjian, R. (1991) Genes Dev. 5, 1935-1945. 16. Zhou, Q., Lieberman, P. M., Boyer, T. G. & Berk, A. J. (1992) Genes Dev. 6, 1964-1974. 17. Martinez, E., Chiang, C.-M., Ge, H. & Roeder, R. G. (1994) EMBO J. 13, 3115-3126. 18. Zenzie-Gregory, B., O'Shea-Greenfield, A. & Smale, S. T. (1992) J. Biol. Chem. 267, 2823-2830. 19. Conaway, J. W., Travis, E. & Conaway, R. C. (1990) J. Biol. Chem. 265, 7564-7569. 20. Smale, S. T., Schmidt, M. C., Berk, A. J. & Baltimore, D. (1990) Proc. Natl. Acad. Sci. USA 87, 4509-4513. 21. O'Shea-Greenfield, A. & Smale, S. T. (1992) J. Biol. Chem. 267, 1391-1402. 22. Roy, A. L., Malik, S., Meisternerst, M. & Roeder, R. G. (1993) Nature (London) 365, 355-359. 23. Colgan, J., Wampler, S. & Manley, J. L. (1993) Nature (London) 362,
24. 25. 26. 27. 28.
29. 30. 31.
37. 38. 39. 40. 41.
549-553. Colgan, J. & Manley, J. L. (1992) Genes Dev. 6, 304-315. Courey, A. J. & Tjian, R. (1988) Cell 55, 887-898. Han, K. & Manley, J. L. (1993) Genes Dev. 7, 491-503. Han, K., Levine, M. S. & Manley, J. L. (1989) Cell 56, 573-583. Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989) Molecular Cloning: A Laboratory Manual (Cold Spring Harbor Lab. Press, Plainview, NY), 2nd Ed. Sadowski, I., Ma, J., Triezenberg, S. & Ptashne, M. (1988) Nature (London) 335, 563-564. Dynan, W. S. (1986) Trends Genet. 2, 196-197. Roy, A. L., Carruthers, C., Gutjahr, T. & Roeder, R. G. (1993) Nature (London) 365, 359-361. Barberis, A., Muller, C. W., Harrison, S. C. & Ptashne, M. (1993) Proc. Natl. Acad. Sci. USA 90, 5628-5632. Buratowski, S. & Zhou, H. (1993) Proc. Natl. Acad. Sci. USA 90, 5633-5637. Ha, I., Roberts, S., Maldonado, E., Sun, X., Kim, L.-U., Green, M. & Reinberg, D. (1993) Genes Dev. 7, 1021-1032. Hisatake, K., Roeder, R. G. & Horikoshi, M. (1993) Nature (London) 363, 744-747. Yamashita, S., Hisatake, K., Kokubo, T., Doi, K., Roeder, R. G., Horikoshi, M. & Nakatani, Y. (1993) Science 261, 463-466. Choy, B. & Green, M. R. (1993) Nature (London) 366, 531-536. Colgan, J., Ashali, H. & Manbey, J. L. (1995) Mol. Cell. Biol. 15, in press. Zenzie-Gregory, B., Khachi, A., Garraway, I. P. & Smale, S. T. (1993) Mol. Cell Biol. 13, 3841-3849. Chang, C. & Gralla, J. D. (1993) Mol. Cell. Biol. 13, 7469-7475. Weinzierl, R. 0. J., Dynlacht, B. D. & Tjian, R. (1993) Nature (London) 362, 511-517.