In the bacterial RNA polymerase transcription complex, the sigma factors play the role of guide, specifying where in the genome the complex binds – and thereby which genes are turned on at any one time.
Lab favourite E. coli has seven such sigma factors, one of which has a ‘housekeeping’ role in standard conditions, with the other six swapping into the complex under particular individual stresses – perhaps the most important (at least from a human perspective) role of these alternative sigma factors being the regulation of virulence genes.
A new study from Bernhard Palsson and colleagues in BMC Biology gives a genome-wide view of transcriptional regulation by these factors, through a combination of transcriptome analysis, sequencing of transcriptional start sites, and chromatin immunoprecipitation of the different sigma factors.
What emerges looks nothing like a traditional textbook model where each protein has a separate, neatly-defined function. Instead, there is a complex network of connections, with various sigma factors regulating both each other and themselves, and extensive overlap between their sets of target promoters; sigma factors both compete with each other for binding, and cover for each other when missing.
This won’t be a complete surprise to those who’ve been following the biology of bacterial transcription: similar individual results have been noted before in small-scale experiments. However, this study recapitulates essentially a few decades’ work in a single set of high-throughput experiments, representing a powerful tool for studying transcriptional networks.
The authors raise the possibility of “developing a comprehensive reconstruction of the entire transcriptional regulatory network in E. coli”, and this is a tempting idea: being able to predict the network-wide effects of perturbing a particular gene can offer insights into how these changes might affect the organism as a whole. Genes with many network connections can be the “capacitors” of evolutionary change discussed by Joanna Masel in a recent Q&A for BMC Biology and touched on by Marc Kirschner in his discussion of evolvability, and the sigma factors have many connections indeed.