(384e) Preferential Interaction Parameters of Proteins In Mixed Solvents | AIChE

(384e) Preferential Interaction Parameters of Proteins In Mixed Solvents

Authors 

Shukla, D. - Presenter, Massachusetts Institute of Technology
Trout, B. L. - Presenter, Massachusetts Institute of Technology


Co-solvents effect protein in mixed solvents through changes in surface tension and preferential interaction with surface residues and backbone. Preferential binding of co-solvents to the protein surface is a critical factor in understanding protein stability in solutions as well as for the design of new additives to inhibit protein aggregation. Preferential interaction of a co-solvent with protein is quantified in terms of preferential binding parameter which measures the excess number of co-solvent molecules in the vicinity of a protein molecule as compared to the bulk solvent. Statistical mechanical approach for quantifying thermodynamic properties of protein in mixed solvents was outlined by Baynes and Trout.(1) They used a predictive, molecular level approach for the study of preferential binding of proteins to co-solvents based on all atom, statistical mechanical models. For the computation of the preferential binding coefficient they used only 2ns of dynamics. Their results for preferential binding coefficient computed from MD simulations and experimental data for proteins RNase A & T1, were in good agreement.

In this work, extensive study is done on model systems comprising of co-solvents urea, glycerol, arginine hydrochloride, guanidinium hydrochloride and glucose with proteins RNaseT1, Lysozyme and α-Chymotripsinogen A. Trajectories in range 10-20 nanoseconds are analyzed in order to validate this method. Simulation results reported by Baynes and Trout(1) based on 2ns dynamics differ from the results obtained from extended runs. Large protein structure fluctuations were observed for RNase T1 which had an influence on the preferential binding parameter values. It is also observed that preferential binding for glycerol is over-predicted for all proteins. Therefore, a comparison of simulated and experimental data is performed using two different force field parameters for glycerol. This work highlights the effect of length of simulation, force field parameters and protein structure fluctuations on estimation of the preferential binding coefficient of proteins in mixed solvents.

References:

1. B. M. Baynes, B. L. Trout, Proteins in Mixed Solvents: A Molecular Level Perspective. J. Phys. Chem. B, 107, 2003, 14058-14067.