The marRAB multiple antibiotic resistance operon of Escherichia coli is autorepressed by MarR. MarR binds to two palindromic sequences in vitro: Site I lies between and overlaps the ?35 and ?10 hexamers for RNA polymerase binding; Site II lies between the transcription start site and the GTG initiation codon of marR. To assess the importance of these sites in vivo, the effects of mutant sites on transcription were analysed using appropriate fusions to lacZ. When both sites were wild-type, transcription was repressed by MarR about 20-fold; when only Site I or Site II was wild-type, transcription was repressed 4.3 ? and 2.6?fold, respectively, indicating that repression at one site is largely independent of repression at the other. Fusions of the wild-type promoter to lacZ demonstrated that marR translation proceeds at only 4.5% of the transcription rate. Analysis of translational fusions with mutant leader sequences demonstrated that the principal reason for inefficient translation is a weak Shine-Dalgarno sequence (SD), AGG(G). Although the SD is located within the potential stem-loop structure of Site II, no evidence for occlusion of the SD was found in the wild-type strain. However, a single bp mutation that would strengthen the stem-loop structure drastically reduced the translational efficiency. Substitution of ATG for GTG as the initial codon did not have a significant effect. Increasing the spacing between the SD and the GTG codon by one to four bases further reduced the translational efficiency by two- to four-fold. We suggest (Martin & Rosner, 2004) that the inefficient translation of marR strongly destabilizes the mRNA resulting in an apparent drop in transcription. The transcriptional activator, MarA, interacts with RNA polymerase (RNAP) to activate promoters of the mar regulon. We identified the interacting surfaces of MarA and of the carboxy-terminal domain of the a subunit of RNAP (a-CTD) by NMR-based chemical shift mapping (Dangi et al 2004). Spectral changes were monitored for a MarA-DNA complex upon titration with a-CTD, and for a-CTD upon titration with MarA-DNA. The mapping results were confirmed by mutational studies and retention chromatography. A model of the ternary complex shows that a-CTD uses a ?265-like determinant? to contact MarA at a surface distant from the DNA. This is unlike the interaction of a-CTD with the CRP or Fis activators where the ?265 determinant? contacts DNA while another surface of the same a-CTD molecule contacts the activator. These results reveal a new versatility for a-CTD in transcriptional activation. This work was carried out in collaboration principally with Dr. J.L. Rosner and with Drs. Bindi Dangi and Angela Groneneborn.
Showing the most recent 10 out of 14 publications
