UC San Diego

Rapid microbial evolution leads to the emergence of multiple drug resistance and transmission of infectious disease among distinct species. Therefore, to successfully design efficient antibiotic treatment and prevent the spread of infections, it is essential to understand the evolution process. In particular, we must be able to predict the effects of mutations that can arise in the adapting populations. The fitness effects of many mutations are context-dependent. Interactions between mutations, or epistasis, make the effects of mutations dependent on their genetic background, while gene-environment interactions modulate the effects of mutations in different environments. Thus, knowing the fitness effect of a mutation alone in one genotype and environment is not enough to predict its effect in different conditions. Recent advances in molecular biology, genetics, and sequencing enabled studying the fitness effects of many mutations across a wide range of different genotypes. While the effects of some mutations are idiosyncratic, the effects of other mutations show a consistent pattern: they become less positive or more negative in genetic backgrounds with higher fitness. The consistency of epistasis for some mutations at least partially resolves the problem of the predictability of the effects of mutations in different genotypes. However, it remains unclear how epistasis changes in different environments. In this thesis, we use directed mutagenesis and short-term experimental evolution to characterize epistasis in yeast Saccharomyces cerevisiae in different environments. Chapter 1 reviews recent findings regarding epistatic and gene-environment interactions. Chapter 2 describes the development of the RB-TnSeq method to generate and measure the effects of many identical mutations in different yeast S. cerevisiae genotypes. This chapter also presents a custom bioinformatics analysis pipeline designed to infer the effects of mutations from the next-generation sequencing data. Chapter 3 reports the application of the RB-TnSeq method to characterize epistasis for 100 random mutations across 42 yeast S. cerevisiae genotypes in six environments, showing that the negative relationship between the effects of mutations and fitness of their genetic background remains largely consistent across environments. Finally, this chapter discusses our findings in the context of the existing literature in this field and describes the potential future research direction.