UCLA

A central challenge of biochemistry is determining the consequences of modifications to proteinaceous amino acids as they pertain to protein function and broader cellular function. Mass spectrometry (MS)-based proteomics enables proteome-wide identification and quantification of modified peptides either through chemical modifications to amino acids such as covalent probe addition or even genetic modifications such as single amino acid variants (SAAVs). Chemical modifications are often assayed through chemoproteomic workflows in order to identify specific ligandable sites by covalent probes designed to be reactive towards certain amino acids, such as cysteine residues in the case of cysteine chemoproteomics. While chemoproteomic workflows are great tools to discover targetable sites by many different types of covalent probes, the scope for which the direct and indirect effects of liganding specific cysteines contribute to reported modes-of-action remains under-characterized. Knowledge of such effects are especially important for prioritizing compounds for follow up studies when performing high throughput screens with multiple scout fragments. Similarly, when using MS-based proteomics for the study of genetic modifications to amino acids, proteogenomic workflows often only address the identification of modifications such as SAAVs and their relative abundance within cells but fail to stratify or establish functional significance of each of these variants. In order to determine functional significance, follow up biochemical studies are often needed which is a time-intensive and low-throughput process. In this work, we first address improving the throughput of cysteine chemoproteomics through the adaptation of the single-pot solid-phase-enhanced sample preparation (SP3) workflow paired with our isotopic capture reagents for cysteine chemoproteomic profiling. We scaled the workflow to accommodate a 96-well plate sample preparation and in-plate peptide enrichment, along with identifying low-cost magnetic beads capable of comparable cysteine coverage, which allowed us to profile 12,265 and identify 2633 ligandable cysteines. We also demonstrated its use for screening multiple compounds, including atropisomers, at several concentrations to better determine unique cysteine and protein targets of such molecules. Next, we investigated a panel of cysteine reactive fragments in order to delineate observed degrader effects on our protein of interest, the SARS-CoV-2 non-structural protein 14 (nsp14), from proteome-wide proteostasis dysregulation. We found that direct covalent modification, ubiquitylation of nsp14, as well as the inactivation of protein disulfide isomerases (PDIs) 3 and 6 are necessary for nsp14 depletion. In addition, we implicated cysteine-reactive electrophiles, used at screening concentrations, in generalized ubiquitylation, proteasome activation, widespread aggregation, and the formation of stress granules upon cysteine-reactive electrophile screening. These unexpected effects on cellular proteostasis led us to develop a framework for identifying wide ranging cellular effects that should be monitored alongside screens which use protein depletion as a readout for identifying cysteine-reactive degrader compounds. Turning to SAAVs, we sought to use thermal proteome profiling (TPP) as a high-throughput proteomic methodology to stratify variants of functional significance as read out by shifts in thermal stability. We first established the use of TPP to characterize alterations in thermal stability between wild-type (WT) and variant proteoforms using B-Raf and its more well-known SAAVs as a model system for this platform. We validated the use of TPP both at the protein and peptide level for assaying variant thermal stability first within an overexpression system, then within a pooled B-Raf variant cell line. Finally, we extended our variant TPP methodology to quantify 77 endogenous variants in Molt-4 leukemia cells through proteogenomic workflows and identify stabilizing and destabilizing SAAVs.