The two faces of short-range evolutionary dynamics of regulatory modes in bacterial transcriptional regulatory networks S. Balaji* and L. Aravind Summary Studies on the conservation of the inferred transcrip- tional regulatory network of prokaryotes have suggested that specific transcription factors are less-widely con- served in comparison to their target genes. This observa- tion implied that, at large evolutionary distances, the turnover of specific transcription factors through loss and non-orthologous displacement might be a major factor in the adaptive radiation of prokaryotes. However, the recent work of Hershberg and Margalit (1) suggests that, at shorter phylogenetic scales, the evolutionary dynamics of the bacterial transcriptional regulatory net- work might exhibit distinct patterns. The authors find previously unnoticed relationships between the regula- tory mode (activation or repression), the number of regulatory interactions and their conservation patterns in g-proteobacteria. These relationships might be shaped by the differences in the adaptive value and mode of operation of different regulatory interactions. BioEssays 29:625–629, 2007. ß 2007 Wiley Periodicals, Inc. Introduction Transcription regulation is mediated by specific transcription factors (TFs), which regulate a particular set of target genes (TGs), by specifically recognizing and binding their promoters. Regulation by specific TFs can either cause ‘‘activation’’ or ‘‘repression’’, which respectively corresponds to increase or decrease of mRNA expression levels with respect to the base line. Over the years, individual studies as well as high- throughput methods have generated an enormous wealth of information on the regulatory inputs provided by specific TFs to their target genes. This has allowed the assembly of transcriptional regulatory interactions and their modes on the genome scale for the prokaryotic model organism, E.coli K12. These data have been made publicly available as RegulonDB (URL: http://regulondb.ccg.unam.mx/index. html), (2) and is being widely used as a base for genomic studies on the structure and evolution of transcription regulation. Typically this information is represented as a network or ordered graph, termed the transcriptional regulatory network (TRN), with two kinds of nodes, namely the TFs and TGs. (3,4) Distribution of regulatory interactions of TFs has been shown to be approximated by a power-law decay. (5,6) This implies that the E. coli TRN has a scale-free topology with a few TFs (hubs) regulatory a large number of TGs, while the rest of the TFs have a limited number of TGs. Some earlier studies on the evolution of the inferred TRN across a phylogenetically wide range of completely sequenced prokaryotic genomes have suggested that TFs and TGs are retained or lost independently of each other. (7,8) In general TGs were found to be maintained to a greater extent than their upstream TFs. It was also observed that hubs were not preferentially retained over TFs with a small number of TGs. These observations on bacterial TRNs suggested that they might be highly flexible, with a notable turnover in the course of evolution of the specific TFs via loss and non- orthologous displacement. This provided a possible model for the adaptive radiations of bacteria, wherein the target genes are maintained across lineages but their regulatory inputs are drastically altered by the turnover of TFs. However, the details of this process at close phylogenetic distances remained unclear. Hershberg and Margalit attempted to understand this by focusing on the g-proteobacteria using a wealth of recently available genomic data for these organisms. (1) Interestingly, the authors found that, within enterobacterial lineage of g-proteobacteria, co-evolution of TFs and their TGs is related to the mode of regulatory interactions between them, i.e. activation or repression. Specifically, repressors, unlike activators, tend to co-evolve tightly with their TGs. Repressors National Center for Biotechnology Information, National Library of Medicine, National Institutes of Health, Bethesda, MD. *Correspondence to: S. Balaji, National Center for Biotechnology Information, National Library of Medicine, National Institutes of Health, Bethesda, MD 20894. E-mail: sbalaji@ncbi.nlm.nih.gov DOI 10.1002/bies.20600 Published online in Wiley InterScience (www.interscience.wiley.com). BioEssays 29:625–629, ß 2007 Wiley Periodicals, Inc. BioEssays 29.7 625 What the papers say