Abstract

The research reported in this thesis builds on an evolutionary engineering experiment (Pinel 2011) that yielded strains of Saccharomyces cerevisiae tolerant to a lignocellulosic hydrolysate. A highly tolerant strain was later characterized by whole genome and transcriptome sequencing (Pinel 2015). The evolutionary trajectories of mutations identified by sequencing were probed by whole population amplicon sequencing, while their significance to the phenotype was assessed by genotyping of additional mutants. Results of this work suggested that our survey of mutations selected during evolutionary engineering was partial. I therefore hypothesized that a complete survey of mutational diversity by whole population genome sequencing would further refine our understanding of lignocellulosic hydrolysate tolerance in S. cerevisiae. I further conjectured that extending this survey to several time points would reveal some of the fundamental evolutionary mechanisms that shape the outcomes of genome shuffling experiments. In parallel, I hypothesized that phenotypic testing of reverse engineered point mutants would identify mutations responsible for lignocellulosic hydrolysate tolerance in our strains of S. cerevisiae. My data revealed that a stong founder effect and prevalent genetic hitchhiking during genome shuffling lead to the domination of compensatory patterns during evolution. Bias introduced by historical contingency lead to the selection of few genuinely beneficial mutations. In the specific context of lignocellulosic hydrolysate tolerance, mutations in genes NRG1 and GSH1, conferring tolerance to acetic acid, oxidative, and potentially other stresses most prominently enhanced the phenotype.