Proteome-wide mapping of sequence-function relationships using mistranslation

Abstract

Amino acid substitutions fuel molecular innovations across the tree of life, yet they also underpin the majority of molecular, cellular, and organismal dysfunction. Delineating between such disparate mutational outcomes is critical in fulfilling the promise of precision medicine and harnessing the power of proteomics for understanding and engineering protein biology. In this dissertation, I present experiments establishing modular and high-throughput proteomic methods to characterize the effects of amino acid substitutions on protein structure and function en masse. Specifically, I showcase Miro (Chapter 2), a proteomics platform that expands mutational scans from single proteins to entire proteomes. I helped establish Miro in Saccharomyces cerevisiae and, once established, I applied this technology to systematically probe the effects of non-canonical amino acid (ncAA) substitutions on protein thermal stability (Chapter 3). Specifically, I first developed a high-throughput thermal stability assay inspired by Thermal Proteome Profiling and Proteome Integral Solubility Alteration. Using this streamlined method, I then coupled it with eight mistranslated proteomes and quantified the effects of ~9000 ncAA substitutions on the stability of >700 proteins. I computationally mapped substitutions back to protein structure to reveal a significant role of local sequence contexts in shaping the impact of a ncAA substitution. I also expanded my analysis to generate protein-specific mutational sensitivity maps, which uncovered clusters of deleterious mutations close in both sequence and three-dimensional space. Many of these clusters also overlapped with regions of known function, highlighting how positional ncAA sensitivity can illuminate functional protein regions. I then coupled this high-throughput stability assay with small molecules to map ATP binding sites across the yeast proteome (Chapter 4). Lastly, I used TPP to identify substrates of the SARS-CoV-2 protease, NSP5, in human cell lines (Chapter 5), by looking for changes in protein stability that arise due to the expression of different NSP5 constructs (wildtype, catalytically dead, or GFP only).