We describe a computer-aided drug-discovery (CADD) pipeline that accounts for protein flexibility. Small-molecule ligands in solution do not encounter static protein targets but rather highly dynamic molecular machines. Many CADD efforts nevertheless consider only a single crystallographic conformation. Such single-structure approaches are sometimes successful because crystallographic conformations are among the many conformations that proteins adopt. Still, these methods may miss additional ligands that bind alternate but equally druggable non-crystallographic conformations. To identify these alternate druggable conformations, we begin with a protein-only molecular dynamics simulation (no ligand) to account for apo-protein flexibility. To assess the impact of small-molecule binding on the conformational ensemble, we apply secondary mixed-solvent simulations (with small chemical probes) to representative conformations extracted from the initial apo simulation. The probes congregate at druggable “hot spots” on the protein surface. Pharmacophores constructed from these hotspots enable subsequent compound searching, and docking serves as a secondary, validating assessment of binding (consensus scoring). To illustrate the utility of the pipeline, we apply it to the RNA-binding protein La-related protein 1 (LARP1), an mTORC1-associated protein that regulates ribosome biogenesis. Elevated LARP1 levels have been implicated in several cancers, so understanding LARP1 dynamics provides critical pharmacological insights. We leverage these insights to identify a bioactive compound with pharmacological potential.

Simulation-based drug-discovery pipeline applied to LARP1, a cancer-implicated drug target

ACS Spring 2022

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.