Prvc. Natl. Acad. Sci. USA Vol. 92, pp. 11952-11954, December 1995 Commentary Biochemistry meets genetics in the holoenzyme Arnold J. Berk Molecular Biology Institute, Department of Microbiology and Molecular Genetics, University of California, Los Angeles, CA 90095-1570

Two quite different types of research have vator proteins bind to regulatory DNA lian cells (17-20). To explore the function been invaluable in elucidating principles sequences called enhancers and increase of the CTD through genetic methods, in the field of molecular biology: biochem- the frequency of transcription initiation at Young and his colleagues (17) identified istry and genetics. When those two ap- neighboring promoters, even when the the minimal number of heptapeptide re- proaches intersect, as they have in recent closest promoter is 50 kb away in the linear peats that would allow a single cell to grow work on eukaryotic transcription control, DNA sequence (6). Activators stimulate into a colony on a plate of rich medium, the resulting reverberation has the ring of transcription in cell-free in vitro reactions 1O, whereas the number found in wild-type truth. In a recent issue of these Proceedings, using complex extracts prepared from iso- Saccharomyces cerevisiae is about 26. They Li et al. (1) presented the latest example of lated cell nuclei. However, they fail to found that, unlike cells with the full-length the convergence of biochemical and genetic stimulate transcription in reactions with CTD, cells with this mutant form of RNA studies that have led to the recent discovery the purified general transcription factors. polymerase II were unable to form colonies of a eukaryotic RNA polymerase holoen- Working in the budding yeast system, at low temperature, most likely because of zyme. The holoenzyme is a multiprotein Roger Kornberg and co-workers went generalized defects these mutants exhibit in complex containing the multimeric RNA back to the complex nuclear extract to regulating transcription of multiple genes. polymerase II protein that transcribes pro- search for an activity that would allow This cold sensitivity of cells with a partially tein coding genes, most of the general tran- activators to stimulate transcription in the deleted CTD allowed Young and his col- scription factors required for the poly- in vitro reaction with purified general tran- leagues to take advantage of the genetic merase to locate eukaryotic promoters and scription factors. They succeeded in iden- technique of identifying suppressing muta- initiate transcription, and a newly discov- tifying such an activity, calling it mediator tions. Such suppressors can result from mu- ered multimeric protein complex called (7). As they purified and analyzed mediator tations in additional proteins involved in the mediator required for the proper temporal (8), it became apparent that they were char- same process. regulation of transcription in response to acterizing the same multiprotein complex The identification of suppressing muta- regulatory transcription factors. Together that was concurrently discovered by Richard tions that allow the CTD partial deletion with the transcription initiation factors Young and colleagues (24) through what mutant to grow at low temperature (21- TFIID and TFIIA, the holoenzyme forms was initially a genetic strategy for the anal- 23) led to the discovery of nine genes an approximately ribosome-sized complex ysis of transcription control. This was the called SRB2, -4, . . . -11 (for suppressor of of more than 50 polypeptides that assem- initial intersection of the biochemical and RNA polymerase B). When the Young bles on eukaryotic promoters. This huge genetic studies that led to the discovery of group isolated these genes and used them preinitiation complex integrates signals the holoenzyme. to prepare specific antibody against the from regulatory transcription factors Young and colleague's genetic strategy encoded Srb discovered bound at enhancer and silencer DNA se- began with an analysis of a unique struc- proteins, they quences to determine how frequently tural aspect of RNA polymerase II, a that all the Srb proteins were part of the RNA polymerase II initiates transcription specific C-terminal domain (CTD) asso- same, very large, macromolecular com- of the neighboring gene. ciated with its largest subunit (9, 10). The plex that included the subunits of RNA Even though RNA polymerase II is a CTD is composed of 26-52 tandem copies polymerase II and most of the subunits of complex multimeric protein of 14 sub- (depending on the organism) of nearly the general transcription factors (22, 24). units (2), by itself it is incapable of perfect repeats of a seven amino acid [However, the critical TFIID factor that starting transcription at the specific ini- residue sequence, Tyr-Ser-Pro-Thr-Ser- binds to the TATA-box sequence of eu- tiation sites used in intact cells. Specifi- Pro-Ser. When RNA polymerase II is karyotic promoters was not present in the cation of the transcription initiation site assembled into the massive preinitiation complex (24)]. The multiprotein complex is a critical aspect of transcription con- complex at eukaryotic promoters, its CTD they purified from yeast nuclear extracts trol, since it determines which portions is unmodified. As the polymerase initiates was very similar to the complex identified of genomic DNA are transcribed into transcription and then transcribes away by Kornberg and co-workers during their mRNA. Consequently, the initial efforts from the initiation site on the template, purification of mediator activity (8). Both of biochemists went toward identifying a the CTD is phosphorylated at multiple groups agreed to call the complex the set of protein factors called general tran- sites (11-14). A number of protein kinases holoenzyme. While there are potentially scription factors (named TFIIA, -B, -D, have been identified that can phosphory- important differences between the com- -E, and -H) that direct the polymerase to late the CTD, but the Cdk7 subunit of position of the holoenzymes described by initiate transcription at sites in DNA that TFIIH appears to be responsible for most the two groups, the many similarities seem are used in the cell (3-5). However, CTD phosphorylations in in vitro tran- more significant than the differences. In although this complex set of general scription reactions (15). In these in vitro experiments with cells having tempera- transcription factors can act together to reactions, Cdk7 phosphorylation of the ture-sensitive srb4 or srb6 mutations, the specify the correct starts for RNA poly- CTD is required for transcription from transcription of multiple genes shut down merase II, they are not sufficient to some promoters (16) but not others (15). rapidly when the cells were shifted to the reproduce the regulation of transcription Genetic experiments using the methods nonpermissive temperature (25). This im- initiation observed in cells. of site-directed mutagenesis revealed that portant observation argues that the ho- A combination of genetic and biochem- the CTD is essential for the viability of loenzyme is required for the transcription ical studies established that, in vivo, acti- budding yeast, Drosophila, and mamma- of most, if not all, yeast genes.

Downloaded by guest on September 23, 2021 11952 Commentary: Berk Proc. Natl. Acad. Sci. USA 92 (1995) 11953 The discovery of the holoenzyme was Li et al. (1) focused on the genes SIN4 number of potential regulatory interac- surprising because earlier work had shown and RGRI because they realized that the tions in such large complexes seems stag- that the purified general transcription fac- published characteristics of cells with mu- gering. But we should not have expected tors and RNA polymerase II can be as- tations in these genes were similar to those less from the mechanisms that control sembled onto promoter DNA in a specific, ofgall] mutants (32-34). Indeed, specific transcription initiation, processes central ordered, stepwise fashion (3-5). Purifica- antibody to Sin4 protein revealed that it to the development of organisms as com- tion of the holoenzyme indicates that, in purifies with the holoenzyme, and peptide plex as ourselves. the cell, most of the general transcription microsequencing revealed that another factors are preassembled into holoenzyme previously unidentified polypeptide in the 1. Li, Y., Bjorklund, S., Jiang, Y. W., Kim, complexes with RNA polymerase II and a complex is encoded by the RGRI gene (1). Y.-J., Lane, W. S., Stillman, D. J. & Korn- mediator/Srb complex. Consequently, as- Thus, earlier studies indicating that the berg, R. D. (1995) Proc. Natl. Acad. Sci. sembly of a preinitiation complex on pro- Galll, Sin4, and Rgrl proteins are re- USA 92, 10864-10868. moter DNA may involve only two binding quired for maximal transcriptional activa- 2. Sentenac, A., Riva, M., Thuriaux, P., Buh- steps: binding of TFIID and TFIIA to tion by a number of different enhancers ler, J.-M., Treich, I., Caries, C., Werner, promoter DNA, followed by binding of can now be explained by the association of M., Ruet, A., Huet, J., Mann, C., Chian- the holoenzyme to the TFIID-TFIIA- these proteins with the holoenzyme, which nilkulchai, N., Stettler, S. & Mariotte, S. promoter DNA complex. Until recently, is required for the transcription of many, (1992) in Transcriptional Regulation, eds. the holoenzyme had been observed only if not all, genes (25). Sugl, another McKnight, S. L. & Yamamoto, K. (Cold in budding yeast. But fundamental cellular polypeptide first identified through ge- Spring Harbor Lab. Press, Plainview, netic studies (35), is also a component of NY), pp. 27-54. processes such as transcription are gener- 3. Sawadogo, M. & Sentenac, A. (1990) ally highly conserved among all eu- the holoenzyme (1). Moreover, as Li et al. point out, the association of Sin4 and Rgrl Annu. Rev. Biochem. 59, 711-754. karyotes, and a recent paper reports the 4. Zawel, L. & Reinberg, D. (1993) Prog. characterization of a holoenzyme from with the holoenzyme has significant addi- tional implications for the function of the Nucleic Acid Res. Mol. Biol. 44, 67-108. mammalian liver cells (26). Thus, it is a 5. Conaway, R. C. & Conaway, J. W. (1993) good bet that the yeast mediator/Srb com- mediator/Srb complex in the holoenzyme. Sin4 and Rgrl proteins are also required Annu. Rev. Biochem. 62, 61-109. plex proteins will have functional coun- 6. Tjian, R. & Maniatis, T. (1994) Cell 77,5-8. terparts in higher cells. The current de- for repression by a number of different repressors (32, 33, 35-37). Similarly, re- 7. Flanagan, P. M., Keleher, R. J. I., Sayre, scriptions of holoenzyme preparations dif- M. H., Tschochner, H. & Kornberg, R. D. fer in the number of associated general cent genetic studies implicate the SrblO (1991) Natuire (London) 350, 436-438. transcription factors (8, 24, 26). It will be and Srbll proteins in mechanisms of re- 8. Kim, Y.-J., Bjorklund, S., Li, Y., Sayre, a challenging problem to determine pre- pression by several different repressors M. H. & Kornberg, R. D. (1994) Cell 77, cisely whichl general transcription factors (38-40). Consequently, the mediator/Srb 599-608. associate with the holoenzyme in vivo subcomplex of the holoenzyme likely 9. Corden, J. L. & Ingles, C. J. (1992) before it binds to promoter DNA. The functions in negative as well as positive in Transcriptional Regulation, eds. Mc- holoenzyme isolated from mammalian control of transcription initiation. The Knight, S. L. & Yamamoto, K. (Cold SrblO and Srbll proteins are particularly Spring Harbor Lab. Press, Plainview, cells appears to include TFIID, raising the interesting because they have sequence possibility that preinitiation complex as- NY), pp. 81-107. homology to Cdk-cyclin protein kinases 10. Dahmus, M. E. (1995) Biochim. Biophlys. sembly on promoter DNA may involve (23), which regulate target proteins by a single protein-DNA-binding step Acta 1261, 171-182. only phosphorylating them at specific sites. 11. Laybourn, P. J. & Dahmus, M. E. (1990))J. (26, 27). Our understanding of the RNA poly- Biol. Chem. 265, 13165-13173. Now, Li et al. (1) report additional merase II preinitiation complex has grown 12. Lu, H., Flores, O., Weinmann, R. & Rein- instances where biochemistry and genetics enormously over the past few years. In berg, D. (1991) Proc. Natl. Acad. Sci. USA cross paths in the holoenzyme, providing addition to the 14 polypeptides of the 88, 10004-10008. further insight into its function. When the polymerase (2) and the 20 polypeptides 13. Kang, M. E. & Dahmus, M. E. (1993) J. holoenzyme is treated with a monoclonal of the general transcription factors TFIIA, Biol. Chem. 268, 25033-25040. antibody that binds to the CTD, 20 -B, -E, -F, and -H (3-5), there are multiple 14. O'Brien, T., Hardin, S., Greenleaf, A. & Lis, polypeptides are displaced in a subcom- TAFs required for activated transcription J. T. (1994) Nature (London) 370, 75-77. plex with mediator activity (8). The sub- (41) that associate with the TATA-box- 15. Makela, T. P., Parvin, P. D., Kim, J., Hu- complex includes the three subunits of binding polypeptide to form the multisub- ber, L. J., Sharp, P. A. & Weinberg, R. A. yeast TFIIF and at least four of the Srb unit TFIID general transcription factor. (1995) Proc. Natl. Acad. Sci. USA 92, proteins. Another of these polypeptides is And now we can add to the preinitiation 5174-5178. encoded by the GALl1 gene (8), a gene complex the 20 additional polypeptides 16. Akoulitchev, S., Makela, T. P., Weinberg, discovered in 1980 because mutations-in it in the mediator/Srb complex. The picture R. A. & Reinberg, D. (1995) Nature (Lon- reduce the ability of yeast cells to ferment that emerges is one of an intricate molec- don) 377, 557-560. galactose (28). Normal fermentation of ular machine positioned at a transcription 17. Nonet, M., Sweetzer, D. & Young, R. A. galactose requires the function of the Gal4 start site consisting of some 60-70 differ- (1987) Cell 50, 909-915. activator protein, which stimulates tran- 18. Allison, L. A., Wong, J. K., Fitzpatrick, ent polypeptides. This appropriately V. M. & Ingles, C. J. (1988) scription of genes encoding enzymes re- named preinitiation "complex" functions D., Moyle, to metabolize In Mol. Cell. Biol. 8, 321-329. quired galactose. gall] to integrate multiple positive and negative 19. Bartolomei, M. S., Halden, N. F., Cullen, mutants, Gal4 activation is impaired (29). regulatory signals from other multiprotein C. A. & Cordern, J. L. (1988) Mol. Cell. Additional studies revealed that galll mu- complexes bound to enhancer DNA se- Biol. 8, 330-339. tations have a general defect that reduces quences (e.g., see ref. 42) to control the 20. Zehring, W. A., Lee, J. M., Weeks, J. R., the ability of multiple activators to stim- timing of transcription initiation. Still Jokerst, R. S. & Greenleaf, A. L. (1988) ulate transcription (30, 31). With this pre- other coactivator proteins not tightly as- Proc. Natl. Acad. Sci. USA 85, 3698-3702. cedent in mind, Li et al. (1) tested whether sociated with DNA-bound protein com- 21. Nonet, M. & Young, R. A. (1989) Genetics any of the unidentified mediator/Srb plexes also participate (e.g., see ref. 43). In 123, 715-724. complex polypeptides might be encoded addition to this complexity, chromatin pro- 22. Thompson, C. M., Koleske, A. J., Chao, by other genes which have generalized teins and their states of modification con- D. M. & Young, R. A. (1993) Cell 73, effects on transcription. tribute to transcription control (44, 45). The 1361-1375. Downloaded by guest on September 23, 2021 11954 Commentary: Berk Proc. Natl. Acad. Sci. USA 92 (1995)

23. Liao, S.-M., Zhang, J., Jeffrey, D. A., sumoto, K. & Fukasawa, T. (1990) Proc. 38. Kuchin, S., Yeghiayan, P. & Carlson, M. Koleske, A. J., Thompson, C. M., Chao, Natl. Acad. Sci. USA 87, 5373-5377. (1995) Proc. Natl. Acad. Sci. USA 92, D. M., Viljoen, M., van Vuuren, H. J. J. & 31. Fassler, J. S. & Winston, F. (1988) Genet- 4006-4010. Young, R. A (1995) Nature (London) 374, ics 118, 203-212. 39. Surosky, R. T., Strich, R. & Esposito, R. W. 193-196. 32. Jiang, Y. & Stillman, D. J. (1992) Mol. (1994) Mol. Cell. Biol. 14, 3446-3458. 24. Koleske, A. J. & Young, R. A. (1994) Cell. Biol. 12, 4503-4514. 40. Wahi, M. & Johnson, A. D. (1995) Genet- Nature (London) 368, 466-469. 33. Jiang, Y. & Stillman, D. J. (1995) Genetics ics 140, 79-90. 25. Thompson, C. M. & Young, R. A. (1995) 140, 103-114. 41. Goodrich, J. A. & Tjian, R. (1994) Curr. Proc. Natl. Acad. Sci. USA 92, 4587-4590. 34. Jiang, Y. W., Dohrmann, P. R. & Still- Opin. Cell Biol. 6, 403-409. 26. Ossipow, V., Tassan, J.-P., Nigg, E. A. & man, D. J. (1995) Genetics 140, 47-54. 42. Thanos, D. & Maniatis, T. (1995) Cell 80, Schibler, U. (1995) Cell 83, 147-155. 35. Swaffield, J. C., Bromberg, J. F. & 529-532. 27. Carey, M. F. (1995) Curr. Biol. 5, 1003- Johnston, S. A. (1992) Nature (London) 43. Luo, Y. & Roeder, R. G. (1995) Mol. Cell. 1005. 357, 698-700. Biol. 15, 4115-4124. 28. Nogi, Y. & Fukasawa, T. (1980) Curr. 36. Chen, W., West, R. W. J., Johnson, S. L., 44. Felsenfeld, G. (1992) Nature (London) Genet. 2, 115-120. Gans, H., Kruger, B. & Ma, J. (1993) Mol. 355, 219-224. 29. Suzuki, Y., Nogi, Y., Abe, A. & Fukasawa, Cell. Biol. 13, 831-840. 45. Hecht, A., Laroche, T., Strahl-Bolsinger, T. (1988) Mol. Cell. Biol. 9, 4991-4999. 37. Covitz, P. A., Song, W. & Mitchell, A. P. S., Gasser, S. M. & Grunstein, M. (1995) 30. Nishizawa, M., Suzuki, Y., Nogi, Y., Mat- (1994) Genetics 138, 577-586. Cell 80, 583-592. Downloaded by guest on September 23, 2021