It has been reported on a set of applications that the binding kinetics is more predictive than the binding affinity of the efficacy of a drug molecule in vivo. Although the thermodynamics depends only on the endpoints of the binding pathway, the kinetics depends on details of the binding process, particularly the ligand binding transition state. Details of the ligand binding transition state have proved difficult to study both experimentally – due to its transient nature and short lifetime – and by simulation – due to the long timescales involved in ligand binding and release. Fundamental outstanding questions remain: How broad is the transition state? How robust is the transition state with respect to small changes in the ligand? Can we use models of the transition state to rationally design ligands with faster or slower binding kinetics? To address these questions, we have developed WExplore, an algorithm based on concurrent loosely-coupled trajectories. By measuring the trajectory flux into the unbound state, we directly compute ligand residence times without using a Markovian assumption that show excellent agreement with those determined experimentally. We report simulations of the release of ligands from soluble epoxide hydrolase (sEH), which have an experimentally determined residence times extending to tens of minutes. We obtain broad sampling of ligand exit pathways, involving distinct exit channels and significant coupled motions between the ligand and the bound receptor. These simulations allow us to quantify the breadth of the transition state ensemble, and identify the specific protein-ligand interactions that govern binding kinetics.

What are the molecular interactions that govern ligand residence time? Insights from molecular dynamics

254th National Meeting (2017)

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