Characterizing heterogeneous solvation environments and their influence in protein stability and binding through molecular dynamics simulations.

摘要

How interfacial solvation environments affect the structure and dynamics of biomolecules is an underlying thread of my thesis. I have used molecular dynamics as a tool for understanding the complicated interactions of these solvation environments in modulating both protein stability and ligand binding. One of the most prevalent heterogeneous environments in biology is the lipid-water interface. I have focused my investigation of the lipid-water interface on its role in the stability of both the secondary and tertiary structural elements of the Glycophorin A (GpA) transmembrane dimer. While the tertiary dimer structure is modulated by the hydrophobic lipid bilayer environment, we show that interactions of GpA with ordered interfacial water are commensurate to intra-helical forces. Water reordering in the presence of the GpA transmembrane domains is observed from simulations and is possibly measurable by experiment. Significant water ordering is also observed in a different system, the binding cavity of cellular retinol binding protein II (CRBPII). The hydrophobic retinol ligand compartmentalizes the binding cavity into a set of smaller disconnected pockets, significantly restricting water motion. In the unbound state water rotation in the binding cavity is restricted compared to translational freedom. We interpret this behavior with a "sliding puzzle pieces" model of water dynamics within the cavity. This model allows for water mobility while conserving hydrogen bonding networks. Analysis of co-solute exclusion around the hydrophobic retinol provides the number of waters in the hydration shell of the ligand as well as the number of waters released upon association to CRBPII. A final goal of my thesis is developing computational methods for efficiently calculating the free energy of solvation for different molecular species in an aqueous and solution. Our investigation has been focused on effective applications of the Jarzynski relationship, which relates non-equilibrium work events to the free energy of a process. We have shown that employing a generalized probability density functional form fits our data well and can be used to obtain accurate free energy estimates. Development of this method for pure polar and non-polar solvation environments will allow for eventually extension into more complex heterogeneous environments.