Excluded volume contribution to cosolvent-mediated modulation of macromolecular folding and binding reactions

Abstract

Water-miscible cosolvents may stabilize or destabilize proteins, nucleic acids, and their complexes or may exert no influence. The mode of action of a specific cosolvent is determined by the interplay between the excluded volume effect and direct solute–cosolvent interactions. Excluded volume refers to the steric exclusion of water and cosolvent molecules from the space occupied by solute, an event accompanied by a decrease in translational entropy. In thermodynamic terms, the excluded volume effect is modeled by creating a cavity which is sufficiently large to accommodate the solute and which is inaccessible to surrounding molecules of water and cosolvent(s). An understanding of the relationship between the energetic contributions of cavity formation and direct solute–cosolvent interactions is required for elucidating the molecular origins of the stabilizing or destabilizing influence of specific cosolvents. In this work, we employed the concepts of scaled particle theory to compute changes in free energy of cavity formation, ∆∆GC, accompanying the ligand–protein binding, protein dimerization, protein folding, and DNA duplex formation events. The computations were performed as a function of the concentration of methanol, urea, ethanol, ethylene glycol, and glycine betaine. Resulting data were used in conjunction with a previously developed statistical thermodynamic algorithm to estimate the excluded volume contribution to changes in preferential hydration, ∆Γ21, and interaction, ∆Γ23, parameters and m-values associated with the reactions under study. The excluded volume contributions to ∆Γ21, ∆Γ23, and m-values are very significant ranging from 30 to 70% correlating with the size of the cosolvent molecule. Our results suggest that a pair of “fully excluded cosolvents” with negligible solute–solvent interactions may differ significantly with respect to their excluded volume contributions to ∆Γ21, ∆Γ23, and m-values thereby differently influencing the equilibrium of the reaction being sampled. This notion has implications for understanding the long-standing observation that, in osmotic stress studies, various osmolytes may produce significantly distinct estimates of hydration/dehydration for the same reaction.