Deciphering Chemomechanical Couplings in Proteins Using Microsecond-level Molecular Dynamics Simulations
Date of Graduation
Doctor of Philosophy in Cell & Molecular Biology (PhD)
Second Committee Member
Third Committee Member
sampling, simulations, spike protein, human fibroblast growth factor (hFGF)
All-atom molecular dynamics (MD) simulations combine the high temporal resolution of experimental methods like smFRET and spatial resolution of methods like x-ray crystallography, to provide a detailed dynamic picture of biomolecular processes. Here, microsecond-level atomistic MD simulations have been used to characterize chemomechanical couplings in human fibroblast growth factor 1 (hFGF1) and the spike proteins of SARS CoV-1 and SARS-CoV-2. hFGF1 is a globular signaling protein that is involved in several physiological processes ranging from cell proliferation to wound healing. Experimental studies have previously described the low proteolytic and thermal stability of hFGF1, in addition to the stabilizing role of heparin. Here, a conformational change in the hFGF1 heparin-binding pocket that occurs only when heparin is absent, is described for the first time. Comparisons with experimental data indicate that this conformational transition is implicated in the low thermal stability of hFGF1. Unique electrostatic interactions that contribute to heparin-mediated stabilization are also described. This work also describes a novel binding affinity estimation approach involving restrained umbrella sampling simulations. The absolute binding affinity for the hFGF1-heparin interaction determined using this approach is in very good agreement with data from isothermal titration calorimetry (ITC) experiments. This binding affinity study revealed that restraining ligand orientation is essential for effective sampling along a protein-ligand distance collective variable.The differential dynamic behavior of the SARS-CoV-1 and CoV-2 spike proteins is also described in this work. Spike protein activation is the first step in the “effective binding” process leading to interaction with the human ACE2 receptor. This study shows that the active form of the CoV-1 spike protein is less stable than that of the CoV-2 spike protein and that the energy barriers associated with activation and inactivation are higher in CoV-2. A “pseudo-inactive” state of the CoV-1 spike protein is described for the first time, wherein the N-terminal domain (NTD) interacts with the receptor-binding domain (RBD). This highlights the potential role of the NTD in spike protein inactivation. The relatively slower kinetics of spike protein activation and inactivation in CoV-2 indicate that it might spend more time bound to the ACE2 receptor than CoV-1, which in turn might provide an explanation for the higher transmissibility of CoV-2.
Govind Kumar, V. (2021). Deciphering Chemomechanical Couplings in Proteins Using Microsecond-level Molecular Dynamics Simulations. Graduate Theses and Dissertations Retrieved from https://scholarworks.uark.edu/etd/4375