NMR²-Based Drug Discovery Pipeline Presented on the Oncogenic Protein KRAS

Fragment-based drug discovery has emerged as a powerful approach for developing therapeutics against challenging targets, including the GTPase KRAS. Here, we report an NMR-based screening campaign employing state-of-the-art techniques to evaluate a library of 890 fragments against the oncogenic KRAS G12V mutant bound to GMP-PNP. Further HSQC titration experiments identified hits with low millimolar affinities binding within the SI/SII switch region, which forms the binding interface for the effector proteins. To elucidate the binding modes, we applied NMR molecular replacement (NMR²) structure calculations, bypassing the need for a conventional protein resonance assignment. Traditionally, NMR² relies on isotope-filtered nuclear Overhauser effect spectroscopy experiments requiring double-labeled [¹³C,¹⁵N]-protein. We introduce a cost-efficient alternative using a relaxation-based filter that eliminates isotope labeling while preserving structural accuracy. Validation against standard isotopically labeled workflows confirmed the equivalence of the derived protein-ligand structures. This approach enabled the determination of 12 NMR² KRAS-fragment complex structures, providing critical insights into structure-activity relationships to guide ligand optimization. These results demonstrate the streamlined integration of NMR² into a fragment-based drug discovery pipeline composed of screening, binding characterization, and rapid structural elucidation with or without isotopic labeling.