Search

We have determined the nucleotide sequence of two Bacillus subtilis promoters (veg and tms) that are utilized by the principal form of B. subtilis RNA polymerase found in vegetative cells (sigma 55-RNA polymerase) and have compared our sequences to those of several previously reported Bacillus promoters. Hexanucleotide sequences centered approximately 35 (the "--35" region) and 10 (the "--10" region) base pairs upstream from the veg and tms transcription starting points (and separated by 17 base pairs) corresponded closely to the consensus hexanucleotides (TTGACA and TATAAT) attributed to Escherichia coli promoters. Conformity to the preferred --35 and --10 sequences may not be sufficient to promote efficient utilization by B. subtilis RNA polymerase, however, since three promoters (veg, tms and E. coli tac) that conform to these sequences and that are utilized efficiently by E. coli RNA polymerase were used with highly varied efficiencies by B. subtilis RNA polymerase. We have also analyzed mRNA sequences in DNA located downstream from eight B. subtilis chromosomal and phage promoters for nucleotide sequences that might signal the initiation of translation. In accordance with the rules of McLaughlin, Murray and Rabinowitz (1981), we observe mRNA nucleotide sequences with extensive complementarity to the 3' terminal region of B. subtilis 16S rRNA, followed by an initiation codon and an open reading frame.

The Bacillus subtilis veg promotor site was shown to be composed of two neighboring RNA polymerase binding sites, only one of which ("Site I") produces a transcript. The relationship between these sites has been investigated following insertion of ten base pairs of DNA between the sites and subsequent cloning of each site independent of the other, and deletion of DNA sequences upstream from the productive site. The effect of DNA insertion was to permit transcription initiation from the previously nonproductive Site II, in either the presence or absence of Site I. This could be attributed to insertion of a purine residue seven base pairs downstream from the Pribnow box of Site II. Transcription from Site I was slightly increased in the presence of Site II in cis, but was still efficient in the absence of Site II, indicating that there is no absolute dependence on Site II for in vitro transcription from Site I.

The Escherichia coli lacUV5 promoter is used inefficiently by the major vegetative form of Bacillus subtilis RNA polymerase, despite very close adherence in the -35 and -10 regions to consensus sequences for promoters recognized by this enzyme. To select derivatives of this promoter with increased activity in B. subtilis, the lacUV5 promoter was fused to a promoter-less chloramphenicol acetyltransferase gene and mutagenized by passage through an E. coli mutD5 mutator strain. Derivatives that conferred resistance to chloramphenicol in B. subtilis were isolated. Twenty-three independent isolates each contained single mutations in the 207 bp lac fragment. These mutations, which were in ten different positions, fell in two clusters. One set of mutations, located between positions -18 and -14, resulted in greater homology to a consensus sequence previously noted for this region in B. subtilis vegetative promoters. The remaining mutations were located near the transcription initiation site. The effects of these mutations and additional mutations constructed by oligonucleotide-directed mutagenesis on expression in B. subtilis and E. coli was determined by measurements of chloramphenicol acetyltransferase activities directed by these promoters. While most mutations had little effect on expression in E. coli, the increase in activity in B. subtilis was as much as 28-fold.