Cosolvents play a large role in shaping protein structure and stability. While the effects of some cosolvents on protein stability and dynamics have been studied to great lengths, many cosolvents have yet to be explored and different cosolvents can have similar effects on protein behavior through unique mechanisms. This demonstrated sensitivity to the chemical details of cosolvents necessitates individual study of the physical mechanisms through which cosolvents alter protein thermodynamics and kinetics. Several experimental techniques, such as NMR spectroscopy and circular dichroism spectroscopy, have proved indispensable in the mechanistic study of cosolvent effects of proteins. However, molecular dynamics simulations can provide unparalleled insight into the specific interactions between protein and cosolvent and can estimate energetic changes for distinct protein conformation<br/>In this talk, we describe a general, efficient method for determining the effects of a cosolvent on the free energy landscape and kinetics of a protein given preexisting sampling from the cosolvent-free conformational ensemble. Our approach involves building a baseline Markov state model (MSM) from the existing simulations, running short simulations for a representative structure of each MSM state in the presence of the desired cosolvent in order to calculate preferential interaction coefficients and subsequently transfer free energies, and finally using the principle of maximum caliber to infer perturbed transition probabilities. The resulting model provides detailed physical insight into the causes of shifts in the conformational ensemble with cosolvent concentration in addition to predicting change in free energy of folding and changes to the folding rate with cosolvent concentration. While this approach relies on extensive a priori simulations, which are more commonly attainable in recent years, it provides a means to study the ensemble-level impact of cosolvents on protein thermodynamics and kinetics at comparatively little additional computational cost. We apply our method to the denaturation of the N-terminal domain of the L9 protein (NTL9) by urea.

Markov state model framework for exploring the effects of cosolvents on protein thermodynamics and kinetics

2019 Spring National Meeting

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.