Variants of the TATA-Binding Protein Can Distinguish Subsets of RNA Polymerase I, II, and

Michael C. Schultz, Ronald H. Reeder, and Steven Hahn Hutchinson Cancer Research Center 1124 Columbia Street Seattle, Washington 98104 Summary Transcription extracts prepared from yeast that are de- ficient in the TATA-binding protein (TBP or TFIID) are also impaired in specific promoter recognition by all three nuclear RNA polymerases (pol I, II, and Ill). Spe- cific initiation can be rescued by the addition of puri- fied recombinant TBP, demonstrating that pol I, II, and Ill all require this factor. A mutation of TBP has been identified that will function with pol I but not with pol II or Ill. Conversely, another mutation, which inactivates TATA element binding in vitro, will function with pol I and Ill promoters but is inactive for a pol II promoter. Thus, it is possible to identify TBP variants that will only function on different subsets of all nuclear pro- moters. Introduction One of the hallmarks distinguishing eukaryotes from pro- karyotes is the use of three distinct RNA polymerase sys- tems for transcription of the nuclear genes. RNA polymer- ase I (pol I) is reserved for transcription of precursor to the large ribosomal RNAs (and occasional transcripts of the rDNA spacer, of unknown function). RNA polymerase II (pol II) transcribes all of the protein-coding messenger RNA. RNA polymerase Ill (pol Ill) transcribes a number of small, relatively stable RNAs, including transfer RNA and the 5S RNA of ribosomes. The reason for this subdivision of labor is not clear, but it is an ancient arrangement, pres- ent from archaebacteria to humans. Much previous work has focused on the differences that exist between the pol I, II, and Ill transcription systems. Although the polymerases share some subunits (Sen- tenac, 1985) each polymerase also contains unique sub- units, has unique inhibitor sensitivities, and appears to interact with a unique assemblage of trans-acting tran- scription factors in order to locate promoters and initiate transcription. For example, pol I interacts with promoters complexed with UBF and a second factor (variously named SLl , TIFIB, factor D, Rib1 ; for review see Reeder, 1992). Pol II requires the basal transcription factors TFIID, A, B, E, F, H, and J (reviewed in Zawel and Reinberg, 1992) and also requires a bewildering array of upstream activators for different individual promoters. Pol Ill requires TFIIIB and C for most of its promoters plus TFIIIA for the 5s gene promoters (reviewed in Geiduschek and Kassavetis, 1992). Although there has been some speculation about possible regulatory interaction between the three polymer-