Multiplex labeling chemistry methods for protein footprinting

Abstract Protein footprinting (PF) is a powerful medium resolution structural biology technique for assessing protein structure and dynamics that relies on “bottom-up” mass spectrometry (MS) to detect, identify, and quantitatively analyze the small (5-15 residue) peptides that are generated in protease-based workflows. Early on hydrogen deuterium exchange (HDX) led the way, later

advances in irreversible reagent development, such as hydroxyl radical footprinting (HRF) mediated by radiolysis; or methods utilizing photolysis of peroxide, and recently plasma, have been introduced and refined. As a result, PF is routinely applied in understanding the effects of protein-ligand binding on higher order interactions in solution, even for large macromolecules like

antibodies, large molecular complexes, and membrane proteins; all of which are important drug targets and biological machines. However, state of the art HRF-based PF studies typically report data from only <20% of the possible sites within the above peptides, limiting the overall impact. In part this is due to the high reactivity of sulfur containing (Met in particular) and aromatic residues relative to many others, creating dynamic range issues for simultaneously detecting high and low abundant species in the same experiment. In this proposal, responsive to PAR-19-253, we propose a range of novel labeling, biophysics, and mass spectrometry methods, based on radical activated trifluoromethylation based chemistries, to provide multiplex, high- resolution labeling and mass spectrometry analysis workflows to maximize value and impact of PF studies, with readouts from 50-100% of accessible side chains within the relevant peptides. The project leverages our advanced synchrotron radiolysis platform and our validation and collaboration strategies will extend the results to major PF platforms such as: radiolysis, photolysis, and plasma based HRF. In Aim 1 (months 0-20), we will benchmark the chemistry of hydroxyl radical induced TFM, understanding its side chain-based reactivity compared to OH radical by examining amino acid and peptide based reactivity for a variety of available TFM reagents. Initial developments of these TFM labeling approaches leverage our advanced synchrotron radiolysis platform for HRF, but we will validate the method using both radiolysis and photolysis, and finding optimum conditions for expanding PF coverage while optimizing dynamic range of labeling. Milestones: Modify ~16/20 residues in a one-pot reaction (vs ~12/20 today). Dynamic range of ~100 or less in separate +16 (-O) and +68 (-CF3) channels (vs. 1000-today in one +16 (-O) channel). In Aim 2 (months 20- 40), we will extend our findings that the Langlois reagent is a promising candidate for hydroxyl radical induced TFM and probing protein structure in a quantitative way using structurally understood and PF tractable calmodulin and estrogen receptor as benchmarked targets. Milestone: Optimize and validate a simple, easy to implement workflow for quantitative structure assessment of proteins based on TFM labeling (radiolysis and photolysis) at structural resolution of 50-100% of residues at the peptide level of accessible residues. Demonstrate method reflects structure at an accuracy of +/- 30Å2. In Aim 3 (months 40-60) we will explore Langlois and other reagents as chemical labeling approaches for footprinting driven small molecule structure activity analysis, to expand structural coverage of protein pockets for optimizing ligand-protein interaction analysis. Milestone: 18/20 residues potentially labelled through multi-pot reactions achieving 80- 100% coverage. In addition, the technology developments will be tested and disseminated in beamline to benchtop protocols with collaborations and through direct training of a wide PF community through publications and workshops.