Calculation and prediction of binding free energies for the matrix metalloproteinases

The zinc-dependent matrix metalloproteinases are drug targets of interest for diseases ranging from arthritis to cancer. Unfortunately, the use of computational rational drug design has been limited by the challenges introduced by the zinc center. We present an extension of the MM/PB/SA methodology which allows us to calculate the relative binding energies of six known nanomolar carboxylate ligands of MMP-1. We are able to rank the neutral and charged ligands correctly. We further illustrate the utility of our approach by modifying the best-binding ligand of our set and predicting a better binding ligand.