COMPARATIVE GENOMIC APPROACHES FOR CHARACTERIZING PROTEIN-LIGAND INTERFACES ACROSS THE TREE OF LIFE

Abstract

Many of the biological functions that serve as the foundation of life occur at the interfaces of proteins and their ligands. The fundamental importance of protein-ligand interfaces suggests that they may play a role in the phenotypic variation observed across species. Recent advances in the speed with which genomes are sequenced and protein structures are solved enables the high-throughput study of these interfaces at a scale that has not previously been possible; this dissertation leverages these data to answer two questions about protein-ligand interfaces. First, to what degree can protein residues that mediate interactions with viral proteins evolve in mammals that are engaging in an evolutionary arms race with more rapidly evolving viral pathogens and does this reflect the variable susceptibility of different hosts to viral infections? Second, how do amino acids that comprise protein-ligand interfaces vary over evolutionary time, and do we see differential patterns of variation when considering residues that bind different ligands? We tackle the first of these two questions by analyzing the family of mammalian proteins that can be bound by the spike protein of the SARS-CoV-2 coronavirus, and the second of these questions by analyzing more than 1000 one-to-one orthologous gene groups across 18 primate species. Overall, the findings herein show that protein-ligand interfaces are capable of evolving in response to viral infections over evolutionary timescales, and that while ligand-binding sites are generally strongly conserved in primates, ligand interfaces that do vary are enriched in proteins responsible for critical biological processes that are associated with strong phenotypic effects.