Protein-Ligand Binding Simulation with the Martini Coarse-Grained Force Field

Abstract

Clarifying the mechanism of protein-ligand interactions is one of the most important research subjects in the field of biophysics. However, most of the research efforts have been devoted to predicting the docking structures. The process of ligand binding remains to be clarified. Molecular dynamics simulation is a straightforward way to study the ligand-binding process. However, reproducing the ligand-binding process in an all-atom simulation is quite difficult because it requires a very long time simulation. For this reason, we have explored the possibility of coarse-grained simulation. In this study, we used the MARTINI force field, in which four non-hydrogen atoms are mapped to one particle on average. We performed ligand-binding simulations for two protein-ligand pairs, the levansucrase-glucose and LinB-1,2-dichloroethane, that differ in the shape of the ligand-binding pocket and in the physicochemical properties of the pocket and the ligand. Each simulation system was composed of one protein molecule, randomly placed ligands, and explicit water solvents. One-microsecond simulations were repeated 100 times with different initial placements of the ligands. For each protein-ligand system, we observed that ligand molecules entered into the correct ligand binding pocket. To obtain further details, we calculated the distributions and the flows of ligand molecules on the protein surface. The distributions of ligands revealed that the ligands were stable in the ligand binding pockets. The analyses of the ligand fluxes demonstrated that the CG ligand molecules entered the ligand-binding pockets through specific pathways. These results suggest that coarse-grained simulation is a good approach to studying protein-ligand binding processes. We will discuss the effects of the shape of the ligand-binding pocket and physicochemical properties of the pocket and the ligand on the binding process.