Water-protein interaction

Interactions with water are the most extensively studied, yet least understood contributions to the stability of globular proteins. A new initiative in our laboratory uses novel NMR and osmometry experiments to probe protein-water interactions. The motivation for this effort is a curious property we observed for Prolactin that appears to involve hydration: Prolactin becomes more stable in the presence of denaturants such as GdnHCl and LiBr, but is not strongly affected by NaCl. Our hypothesis is that Prolactin experiences hydrophobic strain in the native state that can be relieved by diminishing the hydrophobic effect. While denaturant stabilization is pronounced in Prolactin, it may be a property of other proteins, and could explain some anomalous kinetic refolding and stability data. We are nearing completion of the structure of Prolactin under ration with Dr. Birthe Kragelund (University of Copenhagen) and Prof. Joseph Martial (University of Liege). Intermolecular NOEs have been used to study the location and lifetime of relatively long-lived protein-water interactions. These experiments are complicated by chemical exchange and spin diffusion, which can obscure the intermolecular dipole-dipole cross relaxation rates. We showed that neglect of these effects leads to overestimation of the interaction lifetimes for large proteins. Carefully designed pulse sequences can suppress these competing mechanisms of magnetization exchange, and yield more accurate lifetime estimates. We have shown that cross relaxation rates are influenced by cosolutes, and can be used to distinguish cosolutes that are preferentially excluded from the protein surface from those that are accumulated.

Aims: NMR studies can identify regions of the protein in contact with water, but cross-relaxation rates depend on both the spatial distribution and the interaction lifetime; separating the two is underdetermined. In addition, short-lived water-protein interactions that are not detectable by NMR may nonetheless be thermodynamically significant. Osmometry is a technique for measuring water activity, a direct probe of free energy. Any solute added to water lowers the activity, but the extent depends on the nature of the solute-water interaction. In principle, adding a protein to water lowers the activity. Proteins are only sparingly soluble, however, and at the highest attainable concentration the effect should be on the order of one part on 10^5. This is well outside the range of commercial osmometers. We are constructing an osmometer with the aim of achieving 1 part in 10^6 precision. In addition to enabling a thermodynamic "count" of the water molecules interacting with a protein, this precision will enable detection of changes in water-protein interactions resulting from folding/unfolding, binding, and aggregation. By directly measuring the free energy change of water, we aim to separately measure the hydrophobic contributions to thermodynamic parameters obtained by more conventional means, such as calorimetry or van't Hoff analysis.