Quantifying and Understanding Intragenic and Intergenic Epistasis in Yeast

Abstract

Epistasis describes a broad range of interactions within and between molecules. However, limited empirical knowledge is currently available for epistasis at large scale. This dissertation focuses on quantifying intragenic and intergenic epistasis on a large scale. For intragenic epistasis, by combining precise gene replacement and next-generation sequencing, I measured fitness for over 65,000 yeast strains each carrying a unique variant of the tR(CCU)J gene. I managed to quantify epistasis for 61% of all possible combinations of mutations. Almost half of all mutation pairs exhibit significant epistasis, which has a strong negative bias except when the mutations occur at Watson-Crick paired sites. The strong negative bias is also observed for epistasis on the genetic background with one or multiple existing mutations. To study how the fitness landscape and epistasis vary among environments, I measured fitness landscapes in four environments and found that the same mutation almost always has different fitness effects in different environments, indicating pervasive genotype by environment interactions (G×E). Nevertheless, the observed G×E follows a simple piecewise linear relationship. Given the prediction of fitness, an epistasis prediction is also calculated, and the predictive power is comparatively high. Apart from intragenic epistasis, I also studied genetic incompatibility, a form of intergenic epistatic interactions between otherwise functional genes in their conspecific genetic background, which is commonly considered as the major cause of postzygotic isolation and speciation. Despite repeated efforts, Bateson-Dobzhansky-Muller (BDM) incompatibilities between nuclear genes have never been identified between S. cerevisiae and its sister species S. paradoxus. Such negative results have led to the belief that simple nuclear BDM incompatibilities do not exist in yeast. I explored an alternative explanation that such incompatibilities exist but were undetectable due to limited statistical power, and discovered that previously employed statistical methods were not ideal and that a redesigned method improves the statistical power. I also determined, under various sample sizes, the probabilities of identifying BDM incompatibilities that cause F1 spore inviability with incomplete penetrance, and confirm that the previously used samples were too small to detect such incompatibilities, calling for an expanded experimental search for yeast BDM incompatibilities. In summary, this dissertation shows that understanding epistasis at large scale is important and can be achieved through several powerful approaches for elucidating the underlying mechanisms governing evolution, such as available evolutionary trajectories and repeatability of evolution.