High-throughput diversification of protein-ligand surfaces to discover chemical inducers of proximity

Chemical inducers of proximity (CIPs) stabilize biomolecular interactions, often causing a privileged rewiring of cellular biochemistry. While rational design strategies can expedite the discovery of heterobifunctional CIPs, molecular glues have predominantly been discovered by serendipity. Envisioning a prospective approach to discover molecular glues for a pre-selected target, we hypothesized that pre-existing ligands could be systematically decorated with chemical modifications to discover protein-ligand surfaces that are tuned to cooperatively engage another protein interface. Using high-throughput chemical synthesis to diversify a ligand for the transcriptional coactivator ENL with 3,163 structurally diverse chemical building blocks, we discovered a compound dHTC1 that elicits potent, selective, and stereochemistry-dependent degradation of ENL by induced binding to CRL4CRBN. Unlike prior CRBN-based degraders, dHTC1 binds the ligase with high affinity only after forming the ENL:dHTC1 complex, relying on a hybrid interface of protein-protein and protein-ligand contacts. Altogether, this study points toward an expanded chemical space for co-opting the therapeutically important substrate receptor, CRBN, and a second proof-of-concept extending this approach to BRD4 further validates high-throughput chemistry as a facile tool to discover new degraders To compare gene expression changes in mouse-cell depleted bone marrow cells from an MV4;11 xenograft, 4 mice per group received 3 doses of vehicle, (R)-dHTC1 or (S)-dHTC1 50 mg/kg i.p. BID. Samples were collected 1 hour after the last dosing. RNA sequencing was performed with samples from 3 individual mice per group.