Computational studies of protein-ligand and protein-solvent interactions

This thesis demonstrates the use of novel computational methods based on classical mechanics and semiempirical quantum mechanics (QM) to understand biophysical phenomena such as protein-ligand interaction and protein-solvent interaction. Protein-ligand interaction is considered with the aim to calculate the free energy of binding between a protein and a ligand during complexation. Such methods are of immense interest because of their practical applications in structure based drug design. Chapter 2 reviews the theory and computational methods for calculating free energy of binding.; The physically based approach presented in this thesis has the ability to capture binding affinity trends in a diverse range of protein-ligand complexes. The power of this approach to predict binding affinity within protein families and its ability to score ligand poses docked to a protein target is demonstrated in Chapter 3. Its ability to discriminate between native and decoy poses that highlight the crucial role played by electrostatic interactions in molecular recognition is also shown. In particular, the case of metal-ion mediated ligand binding is investigated and obvious advantages of using a QM based method is demonstrated. The performance of this scoring function with respect to other available scoring functions in the literature is discussed.; Pairwise decomposition of the interaction energy between molecules is a powerful tool that can increase our understanding of macromolecular recognition processes. In Chapter 4, the pairwise decomposition of the interaction energy between the enzyme human carbonic anhydrase II (HCAII) and fluorine substituted ligand N-(4-sulfamylbenzoyl)benzylamine (SBB) using semiempirical QM methods is demonstrated. The interaction between the ligand and the protein is dissected by dividing the ligand and the protein into subsystems with the goal of understanding the structure-activity relationships as a result of fluorine substitution. The analysis reveals that the fluorine-substituted benzylamine group of SBB does not directly affect the binding free energy. Rather, the strength of the interaction between Thr199 of HCAII and the sulfamylbenzoyl group of SBB affects the binding affinity between the protein and the ligand. These observations underscore the importance of the sulfonamide group in determining binding affinity as shown by previous experiments.; In Chapter 5, dielectric permittivity of proteins is calculated from MD simulations and semiempirical QM based methods. Multi-nanosecond MD simulations are performed on staphylococcal nuclease, T4 lysozyme, HIV-1 protease and BPTI in a water box. Linear scaling divide and conquer (D&C) method is applied to calculate atomic charges that account for polarization due to the environment. Frohlich-Kirkwood model is then used to calculate the dielectric permittivity of four proteins from their computed dipole moment fluctuations. It is shown that dielectric permittivity of proteins is variable and depends on their chemical and structural properties. Polarization due to an implicit solvent leads to significant increase in the permittivity of the proteins under study. The important role of polarization in determining the dielectric permittivity of a protein is demonstrated.