Distinct transcription factor network dynamics underlie early and long-term adaptation to rifampicin in Escherichia coli.

When subject to sublethal rifampicin stress, susceptible Escherichia coli cells quickly adapt to increase survival. We investigated whether the transcription factor network contributes to this adaptation. First, a responsive cohort of 709 genes diverged from the control, suggesting strong disturbance. While this was largely a direct effect of rifampicin, we show evidence that RNAP, σ70, σ38, Fnr, and (p)ppGpp were influential, suggesting early adaptations. We further show evidence that these responses can be promoter sequence dependent and are partially decoupled from changes in single-cell variability. Moreover, the responses of interacting genes were correlated, suggesting coordination. Later, the transcriptome partially realigned with the control, but 272 genes behaved consistently with long-term adaptation. Several exhibited correlated dynamics that can be explained by interactions, which form three independent, star-like modules of adaptive genes controlled by gcvB, cdaR, and lldR, respectively. Finally, we found that orthologous genes of the evolutionarily distant pathogen Mycobacterium tuberculosis have correlated response strengths to rifampicin, suggesting that our observations may be common across bacterial species. Overall, the results suggest that the initial effects of sublethal rifampicin concentrations on E. coli's genome-wide transcriptional levels are dampened over time and then followed by a distinct state influenced by the gene network that may contribute to adaptation. These findings may assist in developing new strategies to disrupt bacterial adaptation to rifampicin.In natural environments, exposure to sublethal antibiotic (AB) stress is a common phenomenon that enhances the emergence of AB resistance. We dissected the genome-wide transcriptional program of Escherichia coli responsible for its initial response and subsequent adaptation to rifampicin. We show that the transcription factor network plays a major role in this program and provide evidence that the transcriptome response patterns are conserved in the evolutionarily distant pathogen Mycobacterium tuberculosis. Our findings may assist in developing new strategies to disrupt bacterial adaptation to rifampicin.