Key points are not available for this paper at this time.
Frozen protein and nucleic acid solutions at -35 degrees C show relatively narrow (100 milligauss) proton nuclear magnetic resonance signals which are assignable to water that is sufficiently mobile to reduce the dipolar broadening normally associated with solids. Hydration was found to be 0.3 to 0.5 gram of water per gram of protein. Nucleic acids are three to five times as hydrated as proteins. Conformational changes in solution produce detectable changes in linewidth or amount of "bound" water, or both. The very fact that the water signals can be observed by high resolution nuclear magnetic resonance suggests that it is not "ice-like" in any literal sense, although it is clearly less mobile than liquid water at the same temperature. A simple model is described which considers both surface hydration and trapped water.
Kuntz et al. (Fri,) studied this question.