Experimental Test of the Influence of Evolutionary History on the Effect of Beneficial and Regulatory Mutations in Escherichia coli

Selection in fluctuating environments can lead to novel adaptations that may limit or facilitate evolution. I used Escherichia coli populations experimentally evolved in limited-glucose and -lactose environments to examine the effect of historical contingency and test for compensation of the cost normally produced by an evolved non-functional lac operon repressor (lacI-) in glucose. I surveyed 23 diverse strains for natural variation in regulatory function of the E. coli lac operon, which integrates transcriptional control of the lac operon with environmental signals from glucose and lactose availability. Although lac operon regulation has been extensively studied, a considerable amount of research is based on a small number of closely related strains. I found that populations initially evolved in a limited-glucose environment, and clones randomly selected from these populations, were more evolvable than the common ancestor during 1k generations of evolution in a lactose-limited environment, indicating specific genetic changes throughout glucose selection led to increased evolvability. Genome sequencing of the pre-evolved glucose clones revealed mutations in the gene iclR as candidates for increasing evolvability. Next, I tested for compensation of the glucose cost of lacI- in experimentally evolved populations after 8k generations of selection in minimal glucose and lactose fluctuating environments. Competitions measuring the lacI- fitness effect indicated that compensation rarely alleviated the cost in glucose, but epistasis commonly increased the benefit in lactose compared to the ancestor. A reporter introduced into these evolved clones indicated lac operon expression changed but had different fitness effects. Finally, using the same reporter but placed into 28 distinct strains, I measured E. coli lac operon expression in inducers combined at different concentrations. The results encompassed regulatory functions of a vast range, and aspects that were associated with genetic relatedness were the most effective at predicting initial lactose growth. Hybrid reference strains containing lacI and the lac operon of five different natural isolates indicated regulatory elements that had more control over expression could be either global or local. My results demonstrate selection can lead to diverse adaptations that can depend on selective history, mutation interactions, or idiosyncrasies.