No abstract

No abstract

No abstract

Can a transcriptional activator known to bend DNA be functionally replaced by a sequence‐directed bend in Escherichia coli? To investigate this question, a partially truncated promoter was used, deleted of its ‐35 region and of its CRP binding site, leaving only two Pribnow boxes as functional elements. Synthetic and naturally occurring curved DNA sequences introduced upstream from these elements could restore transcription at either one of the two natural starts. Some of these hybrid promoters turned out to be more efficient than the CRP activated wild‐type gal promoter in vivo. Control experiments performed with very similar sequences devoid of any curvature produced weak promoters only. Minimal changes in the location of the centre of curvature or perturbation in the amount of curvature strongly affected the level of expression. No significant stimulation of transcription could be detected in vitro. Furthermore, both gal P1 and P2 starts could be activated in vivo but also in vitro via a properly positioned CRP binding site. This partial analogy suggests that bending induced by the cAMP‐CRP complex upon binding to its site may be biologically relevant to the mechanism of transcriptional activation.

The effects of a number of mutations in the E. coli cyclic AMP receptor protein (CRP) have been determined by monitoring the in vivo expression and in vitro open complex formation at two semi-synthetic promoters that are totally CRP-dependent. At one promoter the CRP-binding site is centered around 41.5 base pairs upstream from the transcription start whilst at the other promoter it is 61.5 base pairs upstream. The CRP mutation E171K reduces expression from both promoters whilst H159L renders CRP totally inactive: neither mutation stops CRP binding at either promoter. The mutations K52N and K52Q reverse the effect of H159L and 'reeducate' CRP to activate transcription. CRP carrying both H159L and K52N activates transcription from the promoter with the CRP site at -41.5 better than wild type CRP. In sharp contrast, this doubly changed CRP is totally inactive with respect to the activation of transcription from the promoter carrying the CRP site at -61.5. Our results suggest that CRP can use different contacts and/or conformations during transcription activation at promoters with different architectures.

We have made a systematic study of how the activity of an Escherichia coli promoter is affected by the base sequence immediately upstream of the -10 hexamer. Starting with an activator-independent promoter, with a 17 bp spacing between the -10 and -35 hexamer elements, we constructed derivatives with all possible combinations of bases at positions -15 and -14. Promoter activity is greatest when the 'non-template' strand carries T and G at positions -15 and -14, respectively. Promoter activity can be further enhanced by a second T and G at positions -17 and -16, respectively, immediately upstream of the first 'TG motif'. Our results show that the base sequence of the DNA segment upstream of the -10 hexamer can make a significant contribution to promoter strength. Using published collections of characterised E.coli promoters, we have studied the frequency of occurrence of 'TG motifs' upstream of the promoters' -10 elements. We conclude that correctly placed 'TG motifs' are found at over 20% of E.coli promoters.

No abstract

Little is known about the molecular mechanism by which histone-like nucleoid-structuring (H-NS) protein and cyclic AMP-catabolite activator protein (CAP) complex control bacterial motility. In the present paper, we show that crp and hns mutants are nonmotile due to a complete lack of flagellin accumulation. This results from a reduced expression in vivo of fliA and fliC, which encode the specific flagellar sigma factor and flagellin, respectively. Overexpression of the flhDC master operon restored, at least in part, motility in crp and hns mutant strains, suggesting that this operon is the main target for both regulators. Binding of H-NS and CAP to the regulatory region of the master operon was demonstrated by gel retardation experiments, and their DNA binding sites were identified by DNase I footprinting assays. In vitro transcription experiments showed that CAP activates flhDC expression while H-NS represses it. In agreement with this observation, the activity of a transcriptional fusion carrying the flhDC promoter was decreased in the crp strain and increased in the hns mutant. In contrast, the activity of a transcriptional fusion encompassing the entire flhDC regulatory region extending to the ATG translational start codon was strongly reduced in both hns and crp mutants. These results suggest that the region downstream of the ؉1 transcriptional start site plays a crucial role in the positive control by H-NS of flagellum biosynthesis in vivo. Finally, the lack of complementation of the nonmotile phenotype in a crp mutant by activation-deficient CAP mutated proteins and characterization of cfs, a mutation resulting in a CAP-independent motility behavior, demonstrate that CAP activates flhDC transcription by binding to its promoter and interacting with RNA polymerase.