Coarse-grained modeling of concentrated protein solutions

In this dissertation, we develop a coarse-grained model to study protein stability in concentrated protein solutions. Our approach uses information from a heteropolymer collapse theory to determine the thermodynamics and physical characteristics of the native and denatured protein states. This information is incorporated into state-dependent inter-protein potentials. The magnitudes of the protein-protein attractions are based on mean-field calculations, which are related to the properties of the native and denatured states. This approach connects protein sequence information with protein-protein interactions and ultimately, through computer simulations and analytical theories, the effects of protein concentration on protein stability. Our results show that proteins will denature in solution if the driving force for unfolding, which is associated with the strength of the hydrophobic attractions between nonpolar amino acid residues, outweighs stabilizing “crowding” effects. However, we predict that sequence hydrophobicity plays a strong, nontrivial role in determining whether a protein is stabilized or destabilized as a function of concentration in solution. Protein sequences with a higher fraction of hydrophobic residues show non-monotonic stability behavior as protein concentration is increased, while protein sequences with a lower fraction of hydrophobic residues tend to be stabilized as a function of protein concentration. Our results agree with the available experimental data. We also extend the original heteropolymer collapse theory to account for both the effects of pressure on protein stability and the presence of hydrophobic patches on the native protein surface. By accounting for the effects of pressure on hydrophobic hydration, we are able to reproduce the characteristic “closed-loop” stability diagrams typically observed for globular proteins as a function of temperature and pressure. We find that strong directional attractions, as a result of the hydrophobic patches on the native protein surface, can stabilize the native state by forming highly organized chains, similar to the experimentally observed behavior for sickle cell hemoglobin. Finally, we combine the above ideas into an analytical theory to study the effects of temperature, pressure, and concentration on protein stability. Our results are qualitatively consistent with both experimental literature and our results from the computer simulations. We also suggest some future directions for this project.