Computational Studies Of Structure And Interactions Between Biological Molecules

In this work, a series of theoretical methods were employed to study the structures andinteractions of some small biomolecules (including cocaine, nicotine and neurotransmitters ctc) andbiomacromolecules (PDE5). The thesis consists of three chapters: The first charpter is a briefintroduction of some recent progress of the development of ab initio electronic structure theory formolecules in solution and a summary of the active site structure of phosphodiesterase-5 and itsproblems to be solved. The second charpter is the computational determinations of the absolute pK_avalues of some amine compounds through first-principles electronic structure calculations usingfour different solvation models, along with an analysis of the most stable conformation of protonatedhistamine in solution based on the calculated Gibbs free energies and vibrational frequencies. Thethird charpter is computational studies of the dynamic structures of phosphodiesterase-5 (PDE5)active site by combined molecular dynamics simulations and hybrid quantum mechanical/molecularmechanical calculations and the structure-activity and structure-selectivity correlation of cyclicguanine derivatives as PDE5 inhibitors by molecular docking, CoMFA and CoMSIA analyses.In the first part of the second chapter, the absolute pK_a valucs of 24 representative aminecompounds in aqueous solution, including cocaine, nicotine, 10 neurotransmitters, and 12 anilines,were calculated by performing first-principles electronic structure calculations. Four differentsolvation models, i.e., the surface and volume polarization for electrostatic interaction (SVPE) model,the standard polarizable continuum model (PCM), the integral equation formalism for the polarizablecontinuum model (IEFPCM), and the conductorlike screening solvation model (COSMO) wereemployed to account for the solvent effect. Within the examined computational methods, thecalculations using the SVPE model lead to the absolute pK_a values with the smallestroot-mean-square-deviation (rmsd) value (1.18). When the SVPE model was replaced by the PCM,IEFPCM, and COSMO, the rmsd value of the calculated absolute pK_a values became 3.21, 2.72, and3.08, respectively. With the empirical corrections using the linear correlation relationships, thetheoretical pK_a values are much closer to the corresponding experimental data and the rmsd valuesbecome 0.51-0.83. The smallest rmsd value (0.51) is also associated with the SVPE model. All of theresults suggest that the first-principles electronic structure calculations using the SVPE model are areliable approach to the pK_a prediction for the amine compounds. In the second part of the secondchapter, first-principles electronic structure calculations were performed on a variety of possible molecular structures of side-chain protonated histamine and different conformers of thefully-protonated state (dication). The calculated results demonstrate that the solvent effectssignificantly affect the relative Gibbs free energies of different molecular structures and, therefore,change their relative concentrations. We also calculated the vibrational frequencies of monocationhistamine in aqueous solution. The calculated scaled theoretical wavenumbers are in good agreementwith the corresponding experimental values for both the natural and the N-deuterated g3I-I conformer.The average deviation was only 10 cm^-1. All the calculated results indecate that the most stableconformation is the g3H conformer rather that the t₃H conformer. The detailed structural informationobtained from the present work might be a valuable reference for future computational studies ofhistamine binding with various biomacromolecular systems.In the first part of the third chapter, various quantum mechanical/molecular mechanical(QM/MM) geometry optimizations starting from an x-ray crystal structure and from the snapshotstructures of constrained molecular dynamics (MD) simulations have been performed to characterizetwo dynamically stable active site structures of phosphodiesterase-5 (PDE5) in solution. The onlydifference between the two PDE5 structures exists in the catalytic, second bridging ligand (BL2)which is HO^- or H₂O. It has been shown that, whereas BL2 (i.e. HO^-) in the PDE5(BL2=HO^-)structure can really bridge the two positively charged metal ions (Zn²⁺ and Mg²⁺), BL2 (/.e. 1-120) inthe PDE5(BL2=H₂O) structure can only coordinate Mg²⁺. It has been demonstrated that the results ofthe QM/MM geometry optimizations are remarkably affected by the solvent water molecules, thedynamics of the protein environment, and the embedded charges of the MM region in the QM part ofthe QMM/MM calculation. The PDE5(BL2=H₂O) geometries optimized by using the QM/MMmethod in different ways show strong couplings between these important factors. ThePDE5(BL2=H₂O) geometries determined by the QM/MM calculations neglecting these three factorsare all consistent with the corresponding geometries determined by the QM/MM calculations thataccount for all of these three factors. These results suggest the overall effects of these three importantfactors on the optimized geometries can be roughly canceled out. However, the QM/MM calculationsthat only account for some of these factors could lead to considerably different geometries. Theseresults might be useful also in guiding future QM/MM geometry optimizations on other enzymes. Inthe second part of the third chapter, molecular docking and 3D-QSAR analyses were performed tounderstand how PDE5 and PDE6 interact with a series of cyclic guanine derivatives. Using theconformations of the compounds revealed by molecular docking, CoMFA and CoMSIA analysesresulted in the first quantitative structure-activity relationship (QSAR) and first quantitativestructure-selectivity relationship (QSSR) models (with high cross-validated correlation coefficient q² and conventional correlation coefficient r² values) for predicting the inhibitory activity against PDE5and the selectivity against PDE6. The high q² and r² values, along with further testing, indicate thatthe obtained 3D-QSAR and 3D-QSSR models will be valuable in predicting both the inhibitoryactivity and selectivity of cyclic guanine derivatives for these protein targets.