Protein solution thermodynamics: a quasichemical perspective of solvent effects

The unfolded(U) folded(F) transition of a polypeptide chain is only marginally stable, with the net free energy change favoring F being typically about 5 kcal/mol (≈ the strength of couple of hydrogen bonds). This small stability arises from a delicate balance of large compensating contributions from hydration effects and intramolecular interactions within the polypeptide chain. Understanding this balance is of principal interest in understanding the molecular basis of forces stabilizing the protein and computer simulations can, in principle, aid in this quest. Computer simulations have advanced to a stage where in silico folding of small polypeptides is now feasible, yet characterizing the hydration thermodynamics of polypeptides larger than a few amino acids remains a daunting challenge. Indeed, over the long history of molecular simulations there have been no calculations, till recently, of the hydration thermodynamics of proteins within an all-atom description of the solvent. Building on the regularization approach to free energy calculations developed in our group, we solve the problem of calculating the hydration thermodynamics of pro- teins in all-atom simulations. These calculations are at a level of resolution that rival what is normally possible for simple solutes such as methane. Our framework also has the virtue of directly quantifying the hydrophobic and hydrophilic contributions to hydration, contributions that are of fundamental interest in understanding the ther- modynamics of protein folding. Using the regularization approach, we have studied the coil to helix transition of model deca-peptides and the response of this transition to thermal and chemical stresses. A major finding of our analysis of the temperature dependent hydration of the pep- tides is that signatures attributed to hydrophobicity have a hydrophilic basis instead. Interestingly, these hydrophilic effects are dominated by the hydration of the peptide- backbone. Further, the textbook picture of rationalizing hydrophobic hydration in terms of specific water structures is shown to be implausible. On balance, in our model systems intramolecular interactions are as important, if not more important, than hydration effects in both the primary-to-secondary and secondary-to-tertiary structure transition. Examining the role of chemical stresses also provides surprises, while also broadly supporting the importance of hydrophilic effects indicated by the temperature depen- dent hydration behavior. The denaturants urea and GdHCl are found to strengthen the hydrophobic effect for the primitive hydrophobe, a cavity that repels water, and yet they denature proteins. Within the helix-coil model studied here, urea stabilizes the coil over the helix by promiscuous hydrophilic interactions primarily mediated by dispersion forces. GdHCl, on the other hand, unfolds the helix by destabilizing it more than the coil, again by tilting the balance of hydrophilic contributions in favor of the coil. TMAO alleviates primitive hydrophobic effects and yet it drives the coil-to-helix transition, primarily by weakening the favorable hydrophilic hydration of the peptide backbone. Tracing the reasons for the current accepted dogmas on dominant forces leads us to consider group-additive models, and these turn out to be Achilles’ heel in modeling protein hydration thermodynamics: such models are fundamentally flawed in ignoring solvent-mediated correlations between different residues on a polypeptide. These correlations can lead to ascribing greater or lesser hydrophobicity/hydrophilicity to the defined chemical group depending on the molecular context in which the group is placed. We suggest abundant caution in relying on such group-additive approaches for a many-body system with energy scales that can span many orders of magnitude. This thesis, together with accumulating experimental evidence, calls for a rig- orous reconsideration of the currently accepted dogmas regarding “dominant” forces driving protein folding and hence also the mechanism of folding based on such dogmas