A predominantly pyrimidine-rich sequence (purine in the template strand, 32 of 37 bases) is located between a functional TATA element and the corresponding transcription start site region of the Saccharomyces cerevisiae iso-1-cytochrome c (CYCI) gene. By using linker deletions and gene fusion techniques, the functional characteristics of this pyrimidine sequence were examined. Results indicate that the function of this element is to limit the accumulation of full-length mRNAs with 5' ends which map upstream of the pyrimidine-rich sequence. Data suggest that the 5'-noncoding region of the CYCI gene possesses signals for mRNA 3'-end processing.Studies which have examined the expression of the Saccharomyces cerevisiae CYCI gene indicate that transcription generates CYCI mRNAs possessing a number of different 5' ends (>20). CYCJ gene expression involves four to five functional TATA elements, each responsible for a subset of mRNAs (8,13). Within the 5'-noncoding region of the CYCJ gene, a pyrimidine-rich sequence (32 of 37 bases; 62% T and 24% C [23]) is located at position -78 to -114 between functional TATA elements and the corresponding transcription initiation region. Similar sequences have been observed between TATA elements and the transcription start site region for a number of S. cerevisiae genes (6), including SUC2 (27), PGK (6), GAL] (29), ADR2 (19), CARl (26), and PYK (5). To ascertain whether any functional properties relating to gene expression can be attributed to the CYC1 pyrimidine sequence, gene fusion techniques were utilized and linker deletion mutations were constructed. Results indicate that this sequence element possesses signals required for the formation of mature 3' ends that map to the 5'-noncoding region of the CYCI gene. The CYC1 pyrimidine-rich element may function independently or in cooperation with other functional elements. The function and the relative position of the pyrimidine element within the 5'-noncoding region suggest that it defines the 5' boundary of the CYCI transcription unit.
MATERIALS AND METHODSBacteria and S. cerevisiae strains. The Escherichia coli strains used were JF1754 (hsdR lac gal metB leuB hisB [14]) and JM101 [A(lac-pro) supE thi F'traD36 proAB lacPZ AM15 (15)]. The S. cerevisiae strains used were GM-3C-2 (a leu2-3 leu2-112 trpl-J his4-519 cycl-1 cyp3-1 [7]) and DBY746 (a his3-AJ leu2-3 leu2-112 ura3-52 trpl-289a).Construction of linker deletions and gene fusions. HIS3 and CYCJ PstI linker deletions were constructed as described previously (13), and the position of each PstI deletion endpoint is shown (Fig. 1). Gene fusions, involving promoter sequences derived from the HIS3 gene, and transcription start site and coding regions derived from the CYCI gene were ligated together at a PstI site and cloned into the yeast shuttle vector Yepl3 (4) as described before (13).Nucleic acid hybridizations. For Northern (RNA) analysis, RNAs were electrophoresed in a TBE nondenaturing gel system (11). The RNA gel was dried and prehybridized (6 h) in 5x SSC (lx SSC is 0.15 M ...