Sequence-dependent model of genes with dual σ factor preference

Abstract

Escherichia coli uses {sigma} factors to quickly control large gene cohorts during stress conditions. While most of its genes respond to a single {sigma} factor, approximately 5% of them have dual {sigma} factor preference. The most common are those responsive to both {sigma}70, which controls housekeeping genes, and {sigma}38, which activates genes during stationary growth and stresses. Using RNA-seq and flow-cytometry measurements, we show that {sigma}70+38 genes are nearly as upregulated in stationary growth as {sigma}38 genes. Moreover, we find a clear quantitative relationship between their promoter sequence and their response strength to changes in {sigma}38 levels. We then propose and validate a sequence dependent model of {sigma}70+38 genes, with dual sensitivity to {sigma}38 and {sigma}70, that is applicable in the exponential and stationary growth phases, as well in the transient period in between. We further propose a general model, applicable to other stresses and {sigma} factor combinations. Given this, promoters controlling {sigma}70+38 genes (and variants) could become important building blocks of synthetic circuits with predictable, sequence-dependent sensitivity to transitions between the exponential and stationary growth phases.