Integrative comparative genomics approaches were used to infer transcriptional regulatory networks (TRNs) in 14 Shewanella species and a set of other g-proteobacteria with sequenced genomes. To accomplish this goal, we combined the identification of transcription factors (TFs), TF-binding sites (TFBSs) and cross-genome comparison of regulons with the analysis of the genomic and functional context inferred by metabolic reconstruction. The reconstructed TRNs for the key pathways involved in central metabolism, production of energy and biomass, metal ion homeostasis and stress response provide a framework for the interpretation of gene expression data. This analysis also helps to improve functional annotations and identify previously uncharacterized genes in metabolic pathways. Finally, we attempted to reconstruct possible evolutionary scenarios of these TRNs.

Using this approach, we identified candidate TFBSs for more then 20 TFs of known specificity, including global regulators (Crp, Fnr, ArcA, Fur, LexA) and specialized regulators of the metabolism of nitrogen (NarP, IscR, NsrR, DNR, NorR), amino acids (BirA, ArgR, MetJ, TrpR, TyrR, HutC), fatty acids (FadR, FabR), carbohydrates (PdhR, HexR, GntR), and cofactors (BirA, IscR). Two novel highly conserved regulons, named LiuR (SO1898) and PsrA (SO2493), were tentatively predicted for the sets of genes involved in the degradation of branch chain amino acids, and fatty acids, respectively. We also identified candidate TFBSs for previously uncharacterized sugar catabolic TFs, termed NagR, SdaR, ScrR, AraR, and BglR, tentatively implicated in the control of the utilization of N-acetylglucosamine, glycerate, sucrose, arabinose and b-glucosides, respectively. Finally, we have mapped the genes and operons controlled by five types of metabolite-binding riboswitches (B12, LYS, RFN, THI, GLY), and six translational attenuators of amino acid biosynthesis pathways (ilv, leu, his, thr, trp, phe operons).

Although some diversity of the predicted regulons is observed within the collection of Shewanella spp., the most striking difference in the overall regulatory strategy is revealed by comparison with E. coli and other g-proteobacteria. Multiple interesting trends in diversification and adaptive evolution of TRNs between lineages were detected including regulon "shrinking", "expansion", "mergers", and "split-ups", as well as multiple cases of using nonorthologous regulators to control equivalent pathways or orthologous regulators to control distinct pathways.

Within the Shewanella lineage, the two major diversification strategies are: constrained ("all or none"), when the regulon is either present or absent in its entirety with tightly conserved regulation of all genes (e.g. for local regulons), and permissive ("loose"), when most genes of a regulon are conserved between genomes, whereas the conservation of respective regulatory sites is much weaker and sometimes not mandatory (e.g. for global regulons). At that, the presence or absence of the constrained regulons is correlated with the pathway essentiality: the biosynthetic NrdR, BirA, and MetJ regulons are always present, whereas sugar catabolism regulons are either present or absent completely. Large regulons seem to be very flexible. Multiple gene and site gains/losses are observed in the LexA, s32, Fnr, ArcA, Crp, Fur and ArgR regulons, although in most cases a conserved core of the regulon can be defined.

Many aspects of metabolic regulation in Shewanella species are substantially different from TRN models that were largely derived from studies in E. coli. Among the most notable are the differences in TRNs for the central carbohydrate pathways. In enterobacteria the central carbon metabolism is controlled by catabolic regulators FruR and Crp, whereas Shewanella species use two other TFs, HexR and PdhR, for this control. The content and functional role of the Crp regulon is significantly different in these two lineages: the catabolism of carbohydrates and amino acids in enterobacteria, and the anaerobic respiration in Shewanella species.