Structure and Evolution of Constitutive Bacterial Promoters

Predicting gene expression levels from any DNA sequence is a major challenge in biology. Using libraries with >25,000 random mutants, we developed a biophysical model that accounts for major features of {sigma}70-binding bacterial promoters to accurately predict constitutive gene expression levels of any sequence. We experimentally and theoretically estimated that 10-20% of random sequences lead to expression and 82% of non-expressing sequences are one point mutation away from a functional promoter. Generating expression from random sequences is pervasive, such that selection acts against {sigma}70-RNA polymerase binding sites even within inter-genic, promoter-containing regions. The pervasiveness of {sigma}70- binding sites, which arises from the structural features of promoters captured by our biophysical model, implies that their emergence is unlikely the limiting step in gene regulatory evolution.