Understanding drug-protein binding thermodynamics, together with their binding kinetic behavior, is critical for rational drug discovery. Binding kinetics can also be a critical differentiator and predictor of drug efficacy and safety. This presentation will introduce a newly developed strategy that applies milestoning theory to extract kinetics and thermodynamics information from running short classical molecular dynamics (MD) simulations for frames from a given dissociation path. High-dimensional ligand-protein motions (3N-6 degrees of freedom) during ligand dissociation were reduced by use of principal component modes for assigning more than 100 milestones, and classical MD runs were allowed to travel multiple milestones to efficiently obtain ensemble distribution of initial structures for MD simulations and estimate the transition time and rate during ligand traveling between milestones. We used five pyrazolourea ligands and cyclin-dependent kinase 8 with cyclin C (CDK8/CycC) as our model system. Based on the unbinding free energy berries, we modified a ligand to increase its residence time and validated our prediction experimentally. We also established a program termed Binding Kinetics Tool (BKiT) to make the calculations easier. This strategy allows us to investigate dissociation kinetics of large and flexible systems for designing small-molecule drugs with desired binding kinetics.

Tuning ligand binding kinetics using transient states and unbinding free energy profile

ACS Spring Meeting 2021

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