Mutation bias is a key predictor of adaptation rate

Abstract

A growing body of work suggests that mutation bias - whereby some types of mutations occur more often than others - can influence the genetic basis of adaptation and the distribution of fitness effects of new mutations. For instance, the loss of specific DNA repair genes shifts the mutation spectrum from a transition-biased to a transversion-biased state in Escherichia coli. This reversal of the ancestral mutation bias enhances access to under-sampled variants, increasing the beneficial mutation supply. But are these effects strong enough to alter adaptive outcomes? We show that mutation bias reversal indeed increases adaptation rate in E. coli by about 5%, and halves the variation in adaptation rates across replicate populations. Even though replicate populations of a strain diverge and adaptation slows down over time, the effect of beneficial supply on adaptation rate gets stronger. The advantage of bias reversal accrues due to greater beneficial mutation supply, with ~2-3 times as many beneficial mutations as strains with reinforced bias. Overall, mutation bias strongly influences the genetic basis of adaptation and alters the types of accumulated genomic mutations that may shape subsequent evolution. Our results show that mutation bias is important for predicting evolution, and highlight multiple ways in which mutational processes can shape the genetic and phenotypic outcomes of adaptation.