Genomics of the marA/soxS/rob regulon of Escherichia coli: identification of directly activated promoters by application of molecular genetics and informatics to microarray data

Summary Microarray analyses are providing a plethora of data concerning transcriptional responses to specific gene regulators and their inducers but do not distinguish between direct and indirect responses. Here, we identify directly activated promoters of the overlapping marA, soxS and rob regulon(s) of Escherichia coli by applying informatics, genomics and molecular genetics to microarray data obtained by others. Those studies found that overexpression of marA, or the treatment of cells with salicylate to derepress marA, or treatment with paraquat to induce soxS, resulted in elevated transcription of 153 genes. However, only 27 out of the promoters showed increased transcription under at least two of the aforementioned conditions and eight of those were previously known to be directly activated. A computer algorithm was used to identify potential activator binding sites located upstream of the remaining 19 promoters of this subset, and conventional genetic and biochemical approaches were applied to test whether these sites are critical for activation by the homologous MarA, SoxS and Rob transcriptional activators. Only seven out of the 19 promoters were found to be activated when fused to lacZ and tested as single lysogens. All seven contained an essential activator binding site. The remaining promoters were insensitive to stimulation by the inducers suggesting that the great majority of elevated microarray transcripts either were misidentified or resulted from indirect effects requiring sequences outside of the promoter region. We estimate that the total number of directly activated promoters in the regulon is less than 40.