Speaker
Description
High-performance computing has turned atomistic Molecular Dynamics (MD) into a practical tool for resolving biomolecular processes with atomic resolution. For protein-ligand systems, these simulations expose the interactions governing binding and unbinding, a key issue in drug design. Drug efficacy, however, depends not only on binding affinity but also on residence time (RT), the kinetic parameter describing how long a ligand stays bound. Experiments can measure RT but cannot resolve its molecular basis, a gap atomistic MD can fill by mapping the free-energy landscape and dissociation pathway together.
That gap becomes acute when mutations in a target alter RT while leaving affinity essentially unchanged, implying that the transition-state (TS) ensemble along the unbinding pathway is selectively perturbed while the bound state is preserved. We examine this in the human adenosine A2A receptor bound to the antagonist ZM241385, a GPCR-ligand system relevant to Parkinson's disease and immuno-oncology. Using metadynamics-based enhanced sampling approaches, we reconstruct both the binding free-energy landscape and the dissociation kinetics of the wild-type receptor and mutants that reduce RT while preserving affinity. Our simulations reproduce the experimental RT differences and show that mutations act primarily by reshaping the TS region, providing a structural basis for rational, RT-optimized ligand design.