A chemoproteomics-based drug discovery strategy is presented that utilizes a highly parallel screening platform, encompassing more than 1000 targets, with a focused chemical library prior to target selection. This chemoproteomics-based process enables a data-driven selection of both the biological target and chemical hit after the screen is complete. The methodology has been exemplified for the purine binding proteome (proteins utilizing ATP, NAD, FAD). Screening of an 8000 member library yielded over 1500 unique protein-ligand interactions, which included novel hits for the oncology target Hsp90. The approach, which also provides broad target selectivity information, was used to drive the identification of a potent and orally active Hsp90 inhibitor, SNX-5422, which is currently in phase 1 clinical studies.
2010 Elsevier Ltd. All rights reserved.