Linking protein function to complex traits

Abstract

In this dissertation, I aim to establish the utility of linking large-scale protein mutagenesis to complex trait selection in order to better understand how subtle alterations to protein function lead to shifts in trait values. I demonstrate this approach using three different complex traits in yeas and finally present a novel method to generate mutant phenotypes with native polypeptides. In the second chapter of this dissertation, I focus on the interaction between two complex traits in yeast: mating and invasion. Using exhaustive mutagenesis of Ste12, a key transcription factor that contributes to both traits, I uncover an unexpected inverse relationship between mating and invasion, and find that an altered DNA-binding mode underlies this phenomenon. In the third chapter, I investigate the relationship between cell growth rate and the heat-sensitive trimerization of a conserved transcriptional regulator of the heat-shock response, Hsf1. With a set over 400,000 variants of Hsf1 with mutations in the trimerization domain, I examine growth rates of each variant under basal and heat-shock conditions to find key features that contribute to temperature-sensitive growth regulation. I find patterns of a trimeric helical geometry in variants that reduce function, and a role for two outer helical positions in conferring a fitness benefit under heat-shock. In the fourth chapter, I present a novel method for the identification of dominant negative polypeptides, and demonstrate its utility in yeast. I confirm the sensitivity of the method by identifying hundreds of dominant negative inhibitors derived from the yeast URA3 and HSF1 genes. Lastly, I show that the method can scale to large, complex libraries for unbiased selection of novel inhibitors of cell growth. In the fifth chapter, I describe a method for the identification of DNA sequences that act as enhancers of gene expression adapted for use in live plant tissues. The sensitivity of the method is confirmed using a known plant enhancer derived from a virus, and individual subsets of the enhancer are tested to find the smallest sequences with maximum enhancer activity. In the final chapter, I discuss the goals of the field of genomics-enabled protein science, and challenges that remain in translating these studies to better predict effects of genetic variants in the clinic and more broadly. I end the dissertation with a brief prospective analysis of new methods to understand protein function in light of emerging technologies.