Water is becoming more and more recognized as an active contributor to protein-ligand binding thermodynamics. However, the computational investigation of protein-water interactions is time-consuming and oftentimes remains limited to a few conformations of the protein. Also, little attention is paid to the timescale at which water molecules sample translational and orientational configurational space, especially during ligand binding processes. We propose a novel approach, Local Correlation Analysis (LoCorA), which facilitates temporal analysis of solute-water interactions obtained from MD simulations without any or only little restraining potentials. In LoCorA, we construct small subregions in a flexible frame of reference, centered on the side chain of the amino aicds and analyse the translational and rotational residence times of the water molecules in the first solvent layer of the amino acids (the local water molecules). Thus, the approach facilitates the analysis of water molecules during conformational changes of the protein binding pocket, as well as during ligand binding processes. As a means of normalization, we introduce solvent damping factors, a new dimensionless descriptor for translational and rotational stability of water molecules specific for each amino acid sidechain in a protein binding pocket. The approach is applied to several Thrombin and Trypsin test systems. For the latter, high-resolution neutron scattering structures are available and thus enable validation of the simulated solute-solvent interactions against experimental data. With the LoCorA approach, we were able to identify large differences for the time scales at which water molecules enter and (re-)orient in the solvation shell of different amino acids in the binding pockets of the two proteins. For some binding site residues, common to the two very similar protein systems, we observed large differences in solvent damping behaviour (one order of magnitude), which indicates that the barriers for the desolvation of these amino acids are important for selective substrate recognition. Our approach can be used for enhancing the understanding of the underlying mechanism of solvation and desolvation of protein binding pockets. The methodology is integrated into the python package LoCorA and will be available to the scientific community from the GitHub page of the presenting author (https://github.com/wutobias).

Approaching temporal structure and stability of protein-water interactions by using local correlations: The local correlation analysis method

2019 Spring National Meeting

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