Simultaneous Identification of Multiple Binding Sites in Proteins: A Statistical Mechanics Approach

Area of Science:

• Computational chemistry
• Biophysics
• Molecular modeling

Background:

• Understanding molecular interactions is crucial for drug discovery and protein function analysis.
• Existing models often struggle to simultaneously account for electrostatic, hydrophobic, and steric effects.
• Accurate prediction of binding sites requires a comprehensive approach to molecular interactions.

Purpose of the Study:

• To introduce the Hydrophobic Dipolar Poisson-Boltzmann Langevin (HDPBL) model, an extension of the Poisson-Boltzmann model.
• To simulate the behavior of water dipoles, ions, and hydrophobic molecules around solutes.
• To investigate the self-consistent organization of solvent, ions, and cosolvents and identify protein binding sites.

Main Methods:

• Extension of the Poisson-Boltzmann model incorporating self-orienting Langevin water dipoles, ions, and hydrophobic molecules.
• Use of a Yukawa potential for hydrophobic interactions and a cubic lattice for steric constraints and incompressibility.
• Derivation of a two-equation system to self-consistently calculate densities of water, salt, and hydrophobic molecules.

Main Results:

• The HDPBL model successfully predicts the organization of ions, cosolvent, and solvent molecules around proteins.
• Density peaks generated by the model indicate the presence of compatible binding sites for different types of ligands.
• Validation showed the model's ability to detect hydrophobic and polar ligand binding pockets and characterize lipid-binding sites on membrane proteins.

Conclusions:

• The HDPBL model provides a unified framework for studying complex molecular interactions around solutes.
• It enables the simultaneous prediction of binding sites for diverse molecules, offering insights into protein-ligand interactions.
• This model has significant potential for applications in drug discovery and understanding biological systems.