Normal cell development requires regulation of transcription initiation and activation in order to express appropriate genes at appropriate times. We study regulation of transcription initiation using a simple prokaryotic system: E. coli RNA polymerase, a five subunit complex comprised of a core (2 alphas, beta, beta?, and omega) and a sigma specificity factor. The sigma70 factor specifies transcription from promoters that are responsible for basal gene expression during vegetative growth. When sigma70 is present within polymerase, two of its domains, 2.4 (an internal region) and 4.2 (the C-terminal region), interact with sequences within the ?10 and ?35 regions, respectively, of host promoter DNA. One focus of our group has been to understand how the components of RNA polymerase affect promoter utilization. The sigma70 subunit of RNA polymerase interacts with promoter DNA only after binding to core RNA polymerase and forming the holoenzyme. Free sigma70 is inhibited from promoter binding by the N-terminal 100 amino acids of sigma70 (domain 1.1), which block domain 4.2 interaction with DNA. Additionally, in free sigma70, domains 2.4 and 4.2 are too close together for the distance between the ?10 and ?35 promoter regions. Upon binding to core, sigma70 undergoes a conformational change, which unmasks domain 4.2 and increases the spacing between domains 2.4 and 4.2. Previous work in many labs has suggested that there is a fairly rigid requirement for the interactions of sigma domain 2.4 and 4.2 with their respective recognition elements at a distance of 10 and 35 base pairs from the start of transcription, respectively. We have previously demonstrated that polymerase lacking sigma70 domain 1.1 (sigma Del100) forms stable transcription complexes more slowly with some promoters, but much more rapidly from a promoter called Pminor. Pminor has poor matches to the expected ?10 and ?35 sequence elements. However, a consensus ?35 sequence (TTGAAA) is centered at ?28.5 relative to the Pminor transcriptional +1. To investigate whether this sequence is critical for Pminor expression, we constructed and assayed a series of Pminor deletions. Our results, in vitro with polymerases reconstituted using either full length sigma70 or sigma Del100, indicate that the TTGAAA sequence centered at ? 28.5 is required for Pminor activity, while the actual ?35 sequence is dispensable. In addition, altering the -28 region to perfectly match the consensus -35 sequence (TTGACA) greatly improved Pminor transcription by polymerase containing either full length sigma70 or sigma Del100. However, perfecting the -10 region (TATGGG to TATAAT) increased Pminor expression only when polymerase containing the full length sigma was used. Potassium permanganate footprinting assays were carried out to examine the Pminor transcription bubble. Although polymerase containing sigma Del 100 formed open complexes more readily than did polymerase with full length sigma, with either polymerase, the Pminor transcription bubble occurred farther downstream than is typical. In addition, slightly altered sensitivity patterns seen when the deleted sigma was present indicated that the promoter-polymerase architecture differs when domain 1.1 is missing. We conclude that polymerase recognizes Pminor by interacting with a -35 canonical sequence located at ?28.5 and that in the context of this """"""""shorter"""""""" promoter, polymerase lacking domain 1.1 is better suited for open complex formation. Our results suggest that RNA polymerase has more flexibility in utilizing promoter sequences than has been previously recognized and that domain 1.1 of sigma 70 is involved in regulating this flexibility.
