Evolution of transcriptional regulation in "Escherichia coli"

During gene expression, transcription initiation marks the first step towards synthesis of functional proteins. Expression levels of specific types of RNA molecules in the cell depend on the underlying genotype of the promoter sequence. Prediction of expression levels from the promoter sequence alone can have important implications for the design of artificial promoters. In this work, we explored promoter determinants that cause differences in expression levels and tracked how a certain level can be reached by a directed evolution experiment in E.coli. Promoter sequences were evolved from a million random sequences with selection on expression level and high mutation rate. Mapping of expression phenotypes to the underlying promoter genotypes revealed what sequence features determine the rate of transcription. If no differential expression is required, incorporation of Sigma70 binding sites allows expression. However, predicted affinity of Sigma70 to bind to a promoter sequence in different promoter contexts is not explanatory in terms of expression levels, suggesting that other sequence features determine the rate of transcription. Furthermore, separation of functional promoter sequences to non-regulatory sequences is promoted by high AT content as well as preference of generally longer promoter sequences. Recovery of an essential missing gene function can also be obtained by overexpression of other genes present in the genome by changing the strength of Sigma70 binding to the promoter sequence. Small changes in the expression level were shown to have a severe impact on the fitness of the organism. The amount of deviation away from the optimal expression level in clonal promoter populations has been shown to depend on the promoter’s genotype. We are presenting an evolutionary model to explain under which regulatory settings selection favors high variance in expression levels between cells.