Investigating Systems-Level Determinants of Bacterial Physiology and Fitness

Abstract

Bacterial physiology and fitness emerge from molecular networks spanning multiple levels between genome sequence and organismal phenotype. Two of these molecular phenotypes include DNA modifications and response to dynamic, stressful environments. In Escherichia coli, Dam methyltransferase creates N6-methyladenine (6mA) at GATC sites, producing heritable epigenetic states that regulate transcription, replication, and mismatch repair (MMR). In feast-famine environments, the extended growth curve imposes successive selective pressures across growth, nutrient depletion, and prolonged starvation that differentially engage antagonistic physiological programs favoring growth and survival. The organizational and temporal principles of each molecular phenotype have not been characterized at genome scale. Here I show that Dam methylation across the E. coli genome is depleted at sites of protein-DNA interaction, and that methylome variability increases substantially in the absence of MMR. Levels of 6mA are lower at core promoter elements and transcription factor binding sites, consistent with physical exclusion of Dam by bound proteins. After 2,400 generations of experimental evolution, MMR- clones accumulated significantly more epimutations at these loci than wild-type clones, with methylomes diverging independently across MMR- clones. I further show that a Bayesian multilevel framework for longitudinal transposon-sequencing, applied to a ten-day feast-famine regime, resolves selection rates that vary across time; a one-dimensional Fisher's geometric seascape model recovers a latent trade-off axis opposing early-growth fitness and starvation survival. This axis is consistent with the self-preservation and nutritional competence (SPANC) framework for growth-survival antagonistic pleiotropy and predicts transposon mutant enrichment timing in an independent repeated feast-famine experiment. Taken together, protein-DNA interactions and MMR organize the E. coli methylome, while temporal fitness structure encodes time-specific constraints. These frameworks provide a foundation for understanding how molecular organization between genotype and fitness shapes bacterial adaptation.