Česká společnost pro hmotnostní spektrometrii

(13. ročník České konference hmotnostní spektrometrie a 11. Neformální proteomické setkání - PL-02)
Proteomics advancing pharmacology

Bernhard Küster ^1 *

1. Technical University of Munich

Abstrakt

This talk will focus on recent chemical proteomics work from my laboratory aiming to advance mechanism-centric pharmacology by merging unbiased target engagement with global proteome response measurements, all in a full dose-dependent manner. Building on dose-dependency as a unifying concept, we established strategies that decode compound action beyond single targets, pathways and transcriptional programs. The decrypt, decryptM and decryptE frameworks align concentration series with proteome-wide responses to infer direct binding, pathway engagement, and adaptive rewiring, enabling mechanism-of-action inference and biomarker nomination in a single experiment. Applying these paradigms to a pharmacologically rich space, we mapped the target landscape of ~1,000 kinase inhibitors, revealing pervasive polypharmacology, unexpected lipid and metabolic kinase engagement, and chemotypes with privileged selectivity patterns. These resources can rationalize phenotypic effects and guide repurposing and lead optimization campaigns. Complementary studies extended the approach to epigenetic modulators, exemplified by quantitative proteomic dissection of lysine deacetylase inhibitors that connected target engagement to dose-dependent proteome remodeling and functional outcomes. Together, these advances operationalize chemical proteomics as an engine for systems pharmacology: they provide scalable, label-free measurements of drug–protein interactions, capture on- and off-targets alongside downstream consequences, and yield actionable maps that bridge medicinal chemistry to cellular physiology. Open data and software from these studies furnish community benchmarks for selectivity modeling, target deconvolution, and network-aware drug design. Looking ahead, integration with high-throughput sample preparation, ultra-fast LC.MS/MS and AI-driven analysis promise routine profiling large compound libraries and anticipatory modeling of drug responses across biological contexts including cancer patients.

* Korespondující autor: kuster@tum.de