The role of TATA elements in the expression of a mammalian promoter was investigated in the fission yeast Schizosaccharomyces pombe, by studying the human adenovirus E2-early promoter. This is a unique dual promoter with two nonconsensus TATA elements directing transcription from two cap sites, +1 and -26. A sequence TTAAGA provides the TATA box function for the +1 promoter, whereas a sequence TαTT, with a closer resemblance to the consensus (TAT-AA/TA) provides this function for the -26 promoter. Yet, in human cells, the +1 promoter is transcribed about 20 fold more efficiently than the -26 promoter. We found that both promoters are transcribed faithfully in S.pombe with start sites identical or close to those found in human cells. Surprisingly, the relative ratio of expression for the +1 and -26 promoters was exactly reversed in S.pombe cells. This reversal appeared to be due to the relatively weak binding of S.pombe TATA binding protein to the TTAAGA motif, rather than to its rate of dissociation. Furthermore, we show that in S.pombe, promoter expression correlates well with the nucleotide sequence of the TATA element rather than the context in which it is placed. By contrast, it is the context of the TATA element, rather than its nucleotide sequence that appears to be critical for promoter expression in human cells. Our data suggest the existence of one or more additional factors in human cells that permit the utilization of nonconsensus TATA elements. S.pombe appears to lack these factors.
ASJC Scopus subject areas