| Short conserved DNA sequence elements located upstream of the transcriptional start site are often binding sites for transcription factors that regulate a group of genes involved in a similar cellular function. Additional genes with these motifs in their upstream regions are good candidates for involvement in the same cellular function. Thus, upstream regulatory motifs can provide powerful hypotheses about links in the genetic regulatory network.;If a DNA motif shared by a coregulated group of genes is known, a recognition matrix constructed from the base frequencies in this motif can be used to predict additional regulon members. Chapter 2 describes the construction of a library of regulatory motifs for 55 known Escherichia coli transcription factors, and the use of recognition matrices for these motifs to search for additional motif instances in the complete E. coli K12 genome. Chapter 3 presents in more detail the construction of the regulatory motif for one of these 55 transcription factors, ArcA, which mediates response to redox conditions. In Chapter 4 we use our library of regulatory motifs to make a set of ranked predictions about gene regulation based on the presence of over-represented clusters of predicted motif instances in the upstream regions of genes. Several of these predictions are currently being tested experimentally.;New regulatory motifs can be discovered computationally by local alignment of upstream regions from coregulated sets of genes. In Chapters 5 and 6, we predict coregulated sets of genes in 22 complete prokaryotic genomes by using comparative genomics and discover new DNA motifs shared by the members of these predicted regulons. By combining three recently published methods for predicting functional interactions between genes based on comparative genomics (methods based on conserved operons, protein fusions, and con-elated evolution) with our motif discovery methods, we have constructed a set of predicted bacteria] regulons that share significant upstream regulatory motifs. We have also analyzed the conservation of motifs between bacterial species. By aligning upstream regions of homologs to E. coli regulon members in other genomes, we found that at least 27 percent of the known E. coli DNA regulatory motifs are conserved in H. influenzae or B. subtilis, but substantial divergence in binding motifs is found in more distantly related genomes. |
