Discovery of covalent ligands on the proteome-wide scale

There may be an untapped reservoir of “druggable” protein targets, according to a new study. To help find potential drug targets, Benjamin F. Cravatt, Keriann M. Backus, and coworkers at Scripps Research Institute in La Jolla, Calif., have extended a method called fragment-based ligand discovery to a proteome-wide scale in native biological systems and used it to identify many proteins that weren’t previously known to bind ligands (Nature 2016, DOI: 10.1038/nature18002). The researchers exposed intact cells or cell lysates to protein-binding molecular fragments attached to electrophilic groups that covalently react with cysteine. Then they treated the proteins with an alkyne-labeled probe known to react with many cysteines in the proteome. If a fragment is already bound to a cysteine, that cysteine is described as “liganded,” and the fragment blocks the probe from binding. In 637 proteins, the researchers identified 758 liganded cysteines, which they defined as ones for which probe binding was reduced by at least 75% relative to samples exposed to dimethyl sulfoxide (DMSO) instead of the fragments. More than 85% of the proteins thus identified were previously not known to bind ligands. For example, the researchers identified ligands that target and distinguish between the inactive precursor forms of two caspases, a family of enzymes that are involved in cell death.