Proc. NatL Acad. Sci. USA Vol. 79, pp. 7385-7389, December 1982 Genetics The yeast his3 promoter contains at least two distinct elements (eukaryotic gene expression/transcription/chromatin/RNA polymerase II/deletion mutants) KEVIN STRUHL* Medical Research Council Laboratory of Molecular Biology, Hills Road, Cambridge, England CB2 2QH; and Department of Biological Chemistry, Harvard Medical School, Boston, Massachusetts 02115 Communicated by Sydney Brenner, September 3, 1982 ABSTRACT Phenotypic analysis of65 mutations indicates that proposed as a promoter element because it is (i) highly con- the yeast his3 promoter is composed of at least two separate re- served before most eukaryotic genes and (ii) homologous in gions ofDNA. Each is necessary, but neither is sufficient for wild- DNA sequence to the E. coli Pribnow box (11). However, de- type levels of his3 expression. Deletion mutations that destroy leting the TATA box ofthe simian virus 40 (SV40) tumor antigen either promoter element express his3 poorly or not at all. The gene seems to have no effect on the expression level (6), and upstream element is located between 112 and 155 base pairs be- similar deletions of the sea urchin histone 2b gene (5) or the fore the site of transcriptional initiation (nucleotides -112 to herpesvirus thymidine kinase gene (8) (assayed after microin- -155). A comparison of derivatives strongly suggests that the downstream element maps somewhere between nucleotides -32 jection into frog oocytes) decrease transcription by a factor of and -52 and includes a sequence between nucleotides -45 and 1A at most, a relatively minor effect. In view ofthese apparently -52. This location coincides with sequences conserved before conflicting results, it is difficult to draw firm conclusions about most eukaryotic genes (the TATA box region). By using derivatives the nature of a eukaryotic promoter. in which hN3 sequences are replaced by a small fragment of coli- This paper continues previous genetic analysis of the his3 phage M13 DNA, three properties of the hiN3 promoter were es- promoter ofbaker's yeast, Saccharomyces cerevisiae (4, 7). The tablished. First, his3 TATA box deletions fail to express hiN3 be- key feature ofthis approach is that the phenotypes ofall mutant cause they lack specific sequences and not because they disrupt genes are determined under true physiological conditions; i.e., spacing relationships between other sequences. Second, the TATA cloned mutant DNAs are introduced back into genetically de- box region can be replaced functionally by the one orientation of fined yeast strains such that each resulting cell contains one the M13 DNA fragment that contains a TATA-like sequence. copy at the normal his3 chromosomal location. Previously, I Third, the distance between the two elements (normally 90 base showed that the his3 promoter is surprisingly large in that it pairs) can be varied between 40 and 160 base pairs without mark- includes a sequence located 112-155 base pairs before the start edly affecting promoter function. These results strongly suggest of transcription (7). Here, I describe 41 additional mutations that yeast RNA polymerase H, unlike its Escherichia coli coun- that lead to the following conclusions: (i) The his3 promoter terpart, does not bind simultaneously to both promoter elements, contains at least two distinct elements. (ii) the his3 TATA box and they add further support to the view that the upstream ele- region suffices for the downstream element. (iii) The distance ment is not part ofa transcriptionally competent binding site. This ability of the his3 upstream promoter element to act at a long and between upstream and downstream elements is not a critical variable distance is similar to properties of viral enhancer se- factor for promoter function. quences and is reminiscent of position effects in yeast. MATERIALS AND METHODS Promoters are sites of DNA necessary for the transcription of The wild-type his3 gene used in these experiments is a 6.1-kilo- structural genes (1). Escherichia coli promoters are composed base-pair (kb) fragment (Sc2605) that contains the entire gene of two noncontiguous sequences (reviewed in refs. 2 and 3). as well as 2.5 kb flanking the 5' end and 2.8 kb flanking the 3' These are located approximately 32-37 base pairs (the -35 se- end (12). All the deletion mutations described here derive from quence) and 6-12 base pairs (the Pribnow box) upstream from Sc2605. The schemes for generating mutants are sketched in the start of transcription. E. coli RNA polymerase interacts di- the text and in Fig. 2; the details concerning isolation, char- rectly with each of these elements, thus leading to the conclu- and DNA will be else- sion that the prokaryotic promoters may be equated with RNA acterization, sequence analysis published polymerase binding sites that are necessary and sufficient for where. transcription. The procedures used to assay the phenotypes of mutant and Transcription ofeukaryotic genes by RNA polymerase II also wild-type genes have been published (7, 13). Briefly, his3 DNA requires DNA sequences flanking the 5' end ofthe mRNA cod- fragments cloned in the ura3V vector YRpL4 (4) are introduced ing region (4-6). However, unlike the situation of E. coli, the into a ura3 hs3- yeast strain. Uracil-independent transform- phenotypes ofmutant genes depend upon the assay. When phe- ants are selected, and those containing one copy of the trans- notypes are determined in vivo, sequences more than 100 base forming DNA integrated at the his3 locus are analyzed for his3 pairs upstream from the mRNA coding region are implicated expression. The phenotypic test is the ability of the relevant as promoter elements (6-8). On the contrary, when phenotypes transformants to grow in the absence of histidine. Those that are assayed by transcription in vitro of purified DNAs, these grow (His') express his3, whereas thoseialing to grow (His-) upstream sequences seem to play no role. Here, the critical do not express the gene. For some mutations, the level ofgene sequence, the TATA box, is located much closer to the site of expression is quantitated by measuring the enzyme activity of transcriptional initiation (9, 10). The TATA box was initially Abbreviations: kb, kilobase pair(s); His+ and His-, histidine indepen- The publication costs ofthis article were defrayed in part by page charge dence and dependence; SV40, simian virus 40. payment. This article must therefore be hereby marked "advertise- * Present address: Dept. of Biological Chemistry, Harvard Medical ment" in accordance with 18 U. S. C. §1734 solely to indicate this fact. School, Boston, MA 02115. 7385 Downloaded by guest on September 28, 2021 7386 Genetics: Struhl Proc. Natl. Acad. Sci. USA 79 (1982) the his3 gene product (imidazoleglycerol-phosphate dehydra- his3 tase). _0- --% I Sc 2605 RESULTS The hMs3 Promoter Contains at Least Two Elements. Pre- -200 -150 -100 -5 0 +50 Endpoints His IGPD viously, I described 24 deletion mutations that successively en- _ a.M *'Ell wt + 1.0 croach upon the 5' end of the hiMO gene (7). These derivatives Upstream TATA RNA fall into three distinct phenotypic classes. Derivatives that re- ,4 f.9 I -32 -32 + 0.8 tain more than 155 base pairs upstream from the start of his3 -32 -20 + 0.5 transcriptional initiation (class I) are phenotypically indistin- 436= -32 -8 + 1.1 guishable from the wild-type allele. Class II derivatives retain a39 -80 -8 - '0.03 between 60 and 113 upstream nucleotides and express his3 a4l -106 -8 - NT poorly, while class III derivatives, which contain less than 45 44, -112 -8 - NT upstream base pairs, fail to express hMs3 -119 -8 - "0.03 443 -80 -20 - NT From these results, two strong conclusions can be drawn. g -106 -20 - NT First, the entire hMs3 promoter is located within the 155-base- J45 -112 -20 - NT pair region adjacent to the 5' end ofthe structural gene. Second, -119 -20 - NT the most upstream promoter element includes a sequence be- 44. -80 -32 - '0.03 tween -113 and -155. In addition, the distinction between 449a471 -106 -32 - '0.03 derivatives that hMs3 or not at all suggests that 1 -112 -32 - NT express partially r -d49-- -119 -32 - NT the promoter may contain a second element. -80 -53 + 0.9 To prove that the hisM promoter contains two distinct ele- I "m I'. -106 -53 + 1.1 ments, it is necessary to show that the-upstream element is not 4430 -119 -53 + 0.2 sufficient for wild-type levels ofgene expression. Nine deletion 427l0 -80 -73 + 0.6 mutations retain the entire upstream element but nevertheless -106 -73 + 0.2 fail to express the gene (see Fig. 1). Thus, it is possible to delete 4S37 -119 -73 +/- 0.05 either of two separate regions of DNA and greatly reduce hisM FIG. 1. Structures and phenotypes ofhis3 deletion mutations; The promoter function. This indicates that the promoter contains top line represents a 6.1-kb fragment of wild-type (wt) yeast his3 DNA at least two elements, both ofwhich are necessary but neither (Sc2605), from which the deletion mutations described here were de-. ofwhich is individually sufficient. rived. The second line is an expanded view ofthe his3 promoter region. The TATA Box Region Can Serve as the Downstream Pro- The scale is in nucleotides and position zero marks the boundary be- moter Element The extent of the smallest deletion tween transcribed(+) nucleotides and nontranscribed (-) nucleotides. his3-A38, The nucleotide sequence for the promoter region between nucleotides that leaves the upstream element but eliminates promoter func- -160 and +10 (probable initiation site underlined) is 5'-TAGTACAC- tion, indicates that the downstream element includes a se- TCTATATTTTTTTATGCCTCGGTAATGATTTTCATTTTTTT- quence located 32-80 base pairs upstream from the mRNA cod- TTTTCCACTAGCGGATGACTCTTTTTTTTTCTTAGCGATTG- ing region (nucleotides -32 to -80).