UC San Diego

The techniques of Molecular Dynamics (MD) as well as Accelerated MD (AMD) are established computational methods for investigating the motions of biomolecules that can be successfully applied to large systems to achieve enhanced conformational sampling. Chapter 1 also introduces Brownian dynamics simulations for studying the diffusional encounter of ions or small molecules with binding sites. Additionally, the use of coupled ordinary differential equations (ODEs) enables mathematical modeling of complex biological systems at a variety of scales. In Chapter 2, AMD simulations are applied to the transmembrane Sarcoplasmic Reticulum Ca²⁺ ATPase (SERCA), a calcium pump that utilizes energy from ATP hydrolysis to drive calcium ions across a concentration gradient. The enhanced conformational sampling achieved with AMD allows identification of collective motions that partition SERCA crystallographic structures into several catalytically unique states and also supports the role of Glu309 gating in Ca²⁺ binding. Brownian dynamics simulations demonstrate the important contribution of surface-exposed, polar residues in the diffusional encounter of Ca²⁺ with SERCA. In chapter 3, Brownian dynamics simulations as well as continuum models of the bifunctional Plasmodium falciparum dihydrofolate reductase-thymidylate synthase (P. falciparum DHFR-TS) are used to explore electrostatic- mediated substrate channeling. The results indicate that electrostatic-mediated channeling in this system is (̃15% - 25%) at physiological pH and ionic strength and that removing the electric charges from key basic residues significantly reduces the electrostatic-mediated channeling efficiency of P. falciparum DHFR-TS. Subtle differences in structure, active-site geometry, and charge distribution in protozoan DHFR-TS enzymes appears to influence both electrostatic-mediated as well as proximity -mediated substrate channeling. In chapter 4, multi-scale mathematical models are used to study the dynamics of a hypothetical anti-HIV genetic therapy at the molecular, within-cell, within-host, and epidemiological scales. This multi-scale approach allows us to make predictions about ideal design parameters for a hypothetical anti-HIV genetic therapy. Our models show that a specific genetic therapy could autonomously target infectious superspreaders at the epidemiological scale, possibly lowering HIV/AIDS prevalence significantly. In chapter 5, insights into the systems detailed in the previous chapters are briefly discussed with an emphasis on the importance of both molecular simulations as well as multi- scale mathematical modeling