Proteome-Wide Deconvolution of Drug Targets and Binding Sites by Lysine Reactivity Profiling

Abstract

Recently, numerous efforts have been devoted to identifying drug targets and binding sites in complex proteomes, which is of great importance in modern drug discovery. In this study, we developed a robust lysine reactivity profiling method to systematically study drug-binding targets and binding sites at the proteome level. This method is based on the principle that binding of a drug to a specific region of target proteins will change the reactivity of lysine residues that are located at this region, and these changes can be detected with an enrichable and lysine reactive probe. Coupled with data-independent acquisition (DIA), the known target proteins and corresponding binding sites were successfully revealed from K562 cell lysates for three model drugs: geldanamycin, staurosporine, and dasatinib. In addition, the druginduced conformational changes of certain targets were also revealed by our method during the screening of staurosporine. The screening sensitivity of our method revealed from the screening of stuarosporine and dasatinib was comparable with that of thermal proteome profiling (TPP) or machine learning-based limited proteolysis (LiP-Quant). Overall, 21 and 4 kinase targets, including adenosine 5′-triphosphate (ATP)-binding targets, were identified for staurosporine and dasatinib in K562 cell lysates, respectively. We found that target proteins identified by TPP, LiP-Quant, and our method were complementary, emphasizing that the development of new methods that probe different properties of proteins is of great importance in drug target deconvolution. We also envision further applications of our method in proteome-wide probing multiple events that involve lysine reactivity changes.

Cellular protein painting for structural and binding sites analysis via lysine reactivity profiling with o-phthalaldehyde

“…The abundance decrease of native peptide identified could potentially arise from direct interference of trypsin by OPA, hence steps are incorporated in our workflow to remove excess OPA prior enzymatic digestion of proteins (27). Nevertheless, to further validate that the decreased abundance of native peptides arises from increase in OPA-labelled peptides, purified MAP kinase-activated protein kinase 2 (MAPKAPK2), which contains 30 lysine residues, was incubated with OPA followed by bottom-up MS analysis where OPA is included as variable lysine modification during MS spectral search.…”

“…5A-B), all which were located in the known binding regions in the target proteins. 26 Specifically, Lys 58 was located directly in the interaction region, while Lys 69 and Lys 84 are located near the interaction region (Fig. 5C).…”

“…Specifically, the peptides with significant lysine residues shifts in reactivity were [ [E] and for HSP90AB1 (Fig. 5A-B), all which were located within or near the known binding regions of geldanamycin in these target proteins (27). For HSP90AA1, Lys 58 was located directly in the interaction region, while Lys 69 was located near the interaction region (Fig.…”

Ligand Modification-Free Methods for the Profiling of Protein-Environmental Chemical Interactions

“…Target proteins could be determined by the abundance changes on the probe labeled sites or protease cleavage sites in the presence and absence of chemical ligands under native conditions. Representative examples include limited proteolysis-coupled mass spectrometry (LiP-MS) and lysine reactivity profiling methods …”

“…Representative approaches include proteolysis-based drug affinity responsive target stability (DARTS) approach 74 and limited proteolysis-coupled mass spectrometry (LiP-MS), 73 chemical labeling-based hydrogen−deuterium exchange (HDX) 78 and lysine reactivity profiling methods (like target responsive accessibility profiling (TRAP) approach). 76,79 In these methods, quantifying the conformotypic proteolytic or chemically labeled peptides/sites enables proteome-wide deconvolution of binding proteins, sequence regions, and sites (Figure 5). As an illustrative example, LiP-MS identified 1678 metabolite-protein interactions and 7345 putative binding sites for adenosine 5′-triphosphate (ATP), L-phenylalanine (L-Phe), and phosphoenolpyruvate (PEP) in prokaryotic E. coli.…”