Large Library Docking for Polypharmacology

Polypharmacological molecules are attractive for complex illnesses. Here, we explored large library docking for joint activity against target pairs. Retrospectively, as libraries grew, so too did the number of likely dual-activity molecules. In prospective docking of a 900-million molecule library against three target pairs (α_2A/SERT, MOR/SERT, and α_2A/MOR), we sought analgesic compounds. Both the α_2A/SERT and SERT/MOR campaigns led to dual binders with low μM to high nM activities with high hit rates; tetrahydropyridines from the α_2A/SERT campaign were also active against 5-HT_2A. However, even though cryo-EM structures confirmed the docking-predicted poses, optimization struggled to improve potency. Still, in mouse behavioral assays, the most potent α_2A/SERT compound ('z7149) was effective against pain without inducing conditioned place preference, and the molecule had potent antidepression and anxiolytic drug-like behavior, consistent with its SERT/5-HT_2A activities. This study reveals both advantages and challenges of docking for polypharmacology.