Microsecond Simulations of the PmHMGR Second Hydride Transfer Transition State and Investigations into Molecule and Enzyme Dynamics

In this thesis, I will primarily discuss how molecular dynamics simulations of the transition state of PmHMGR compared to ground state simulations facilitated the prediction of allosteric residues that help stabilize the transition state of the enzyme. Three residues were chosen for experimental mutagenesis based on the computations, L407S, E399A, and R396A. The residues were located on the flap domain of the enzyme and at least 15 Å away from the transferring hydride. Mutagenesis on PmHMGR produced the L407S, R396A and E399A mutants. Kinetics were run and the L407S had a 57% reduction in activity, E399A had a 69% reduction in activity, and R396A had 10% in comparison to the wild-type enzyme. While the R396A mutant was more resistant to mutation, the other two mutants caused disruption in the normal function of PmHMGR, confirming the prediction from the computational analysis. This is one of the first instances of computational chemistry predicting allosteric residues in regard to catalysis.

In addition, several collaborative projects will be discussed. First, density functional theory was used to analyze the triplet state density of the visible-light-activated [2+2] photocycloaddition through a Lewis Acid rhodium catalyst. This aided in the determination of triplet excited state mechanism. Molecular dynamics was used to help drug discovery efforts for the epigenetic HDAC and P300/CBP systems, which, in the case of the latter, shed light on the role of protein dynamics on the ligand-protein and protein-protein interface. Finally, Monte Carlo sampling techniques were employed to study how conformational preference of natural products and natural product analogues can influence their biological activity and was utilized in the design of new analogues to probe the conformation-activity relationship.