Binding free energy calculations in the polar and nonpolar T4 lysozyme mutants

COMP 403

Yuqing Deng, Biosciences Division, Argonne National Laboratory, B236 Building 221, 9300 S Cass Avenue, Argonne, IL 60439 and Benoit Roux, Dept of Biochemistry and Molecular Biology, University of Chicago, Chicago, IL 60637.
Calculations of the binding free energies of aromatic ligands to the polar (L99A/M102Q) and nonpolar (L99A) cavity in engineered T4 lysozyme mutants using free energy perturbation and molecular dynamics (MD) are presented. Biasing potentials are used to restrain the ligand orientation and center-of-mass movement relative to the protein. MD simulations are run with explicit water molecules inside a sphere of 15 Angstrom diameter centered on the binding pocket. The influence from the rest of the system is incorporated using the generalized solvent boundary potential (GSBP) method. The total binding free energy is separated into repulsive, dispersive, electrostatic contributions, addition and removal of the restraints. The computed binding free energies are generally in good agreement with experimental values well. In particular, benzene, toluene as well as phenol bind to the T4 lysozyme mutant with the engineered polar cavity, whereas only benzene and toluene bind to the mutant with the engineered nonpolar cavity. The free energy decomposition shows that phenol does not bind due to the greatly unfavorable electrostatic interaction. The decomposition of the nonbond contributions shows that the the binding cavity is pre-formed in the apo protein, thereby reducing the cost of steric clash upon ligand binding. For similar-sized ligands The contributions from nonpolar interactions is similar for both and polar and nonpolar cavities.