Energetic and conformational studies of nonspecific adsorption of simple protein-like chain molecules using dynamic Monte Carlo simulations

Dynamic Monte Carlo simulations of short HP (hydrophobic-polar) protein-like chains to solid-liquid surfaces are used to probe thermodynamic and dynamic aspects of protein adsorption. The HP model enables the enumeration of all chain conformations, thereby aiding understanding of the relation between adsorption thermodynamics and changes in accessible chain conformations resulting from the sorption process. Simulation results indicate that HP chains having a single conformation at their lowest energy in solution adsorb such that the new lowest energy state of the system is conformationally degenerate. As a result, adsorption can lead to an increase in chain entropy. Entropically-driven adsorption is found to be likely when the interaction energy between the hydrophobic segments of the chain and the sorbent is weak and equals the contact energy between two hydrophobic units within the chain. Chain sequence and sorbent properties are shown to profoundly influence adsorption thermodynamics. Simulations are carried out where intra- and intermolecular hydrophobic interaction energies are varied to examine the influence of the stability of the native-state conformation on adsorption thermodynamics over a range of sorbent hydrophobicities. Lower stability chains tend to adsorb more readily on hydrophilic sorbents and experience greater average changes in conformation, usually accompanied by a loss in entropy. Adsorption to more hydrophobic sorbents leads to a loss in chain conformational entropy, irrespective of the stability of the native state. Lateral confinement on the sorbent surface is shown to greatly reduce the degrees of freedom in the chain, thereby resulting in a strong stabilization of the native-state conformation of the chain in its adsorbed state. This effect is compared to experimental data for nonspecific adsorption of hen egg-white lysozyme to silica to explain the increase in adsorbed enzyme activity as a function of surface loading and geometry. Studies of run-averaged energy trajectories for chain adsorption indicate that the process follows a basic energy path characterized by well-defined energy levels, suggesting the presence of natural kinetic barriers. This thesis demonstrates the value of simple mesoscopic protein-like chain models and dynamic Monte Carlo simulations of their adsorption behavior in understanding better the mechanisms and forces driving nonspecific protein adsorption.Applied Science, Faculty ofChemical and Biological Engineering, Department ofGraduat