Functional Epistasis and Evolutionary Dynamics

Abstract

The adaptive mutations that drive evolution do not affect the fitness of the organism directly, instead, their effects are mediated by the rest of the genome in the complex process that transforms genotype to phenotype. This means that the fixation of a single mutation can forbid a class of adaptive mutations whose effects depend on its absence and make accessible class of adaptive mutations whose effects depend on its presence. This simple fact, known as epistasis, underlies a wide range of evolutionary phenomena. Here, I investigate two of these phenomena: second order selection and historical contingency. In rapidly adapting asexual populations, the fate of a new mutant lineage is determined both by its intrinsic capacity to leave more offspring, its fitness, and by its offspring’s potential to acquire further beneficial mutations, its evolvability. Second order selection refers to the possibility that a lineage might be selected for its evolvability. Unfortunately, due to the transient nature of evolutionary dynamics, it is difficult to detect the effects of second order selection. In chapter~\ref{ilt}, we use high-resolution, sequential, DNA barcoding technology to reveal the complex lineage and sub-lineage dynamics in rapidly evolving laboratory populations. These dynamics can be compared to population genetic models to test for instances of second order selection. Historical contingency refers to the fact that the initial genotype of a population can influence its evolution. We can quantify the effect of historical contingency in the lab by quantifying systematic differences in evolutionary outcomes between different initial genotypes. Although historical contingency is widespread, the functional determinants of the systematic differences in evolutionary outcomes are unknown. In chapter~\ref{ka}, we use a hierarchically structured evolution experiment, evolving 20 replicate populations of each of 37 gene deletion mutants nested within 11 functional modules, to test the hypothesis that a gene's function may be predictive of evolutionary outcomes in lineages deficient for that gene. We find that functional relatedness can predict the rate and genetic basis of adaptation in gene deletion mutants: mutants with functionally related deletions tend to adapt more similarly, on average, than those with unrelated deletions.