There has been some debate as to the sites of water sorption and the extent of coverage, particularly with regard to the participation of sites along the polypeptide backbone. Even the amount of "bound" water has been difficult to define unambiguously. The physical properties of water interacting with proteins have been studied with a variety of techniques, including dielectric relaxation methods (see Chapter 4), infrared and NMR spectroscopy, Rayleigh scattering of Mossbauer radiation (RSMR) (see Chapter 5), hydrodynamic methods as well as thermodynamic measurements (see reviews by Kuntz and Kauzmann , Edsall and McKenzie , Rupley and Careri , Luscher-Mattli , and the volume of "Methods in Enzymology" edited by Hirs and Timasheff ). Most methods provide some estimate of the amount of water that interacts with the protein and that has structural or dynamic properties, as measured by the particular technique, that are different from those of bulk water. This water is termed either hydration water or "bound" water. Attempts by Finney and Poole  to rationalize the literature on this subject led them to the following conclusion: "The amount of bound water is a function of the method used to examine it and probably also of the form of the sample used."
In this section, we examine the picture of sequential hydration developed by Rupley and Careri . Although the general picture that has emerged is reasonably consistent and many correlations between the results from different techniques are found, several issues remain to be fully resolved. Recent infrared results  cast some doubt on the interpretation of changes in the amide I band on hydration and the identification of some of the sorption sites. Finney and Poole  also interpret the heat capacity data somewhat differently from Rupley, which leads to a different estimate of the water necessary to cover apolar regions of the protein surface.
There are several critical hydration levels in the adsorption isotherm that mark the onset of important processes. Recovery of surface and internal mobility is obviously of particular importance, and we discuss this later in a separate section. Here, we group together three important and interesting hydration-dependent phenomena—recovery of enzymatic activity, proton percolation, and "nonfreez-ing water."
A. Infrared Spectroscopic Studies of Protein Hydration
Because of the sensitivity of vibrational modes to hydrogen bonding, Fouriertransform infrared spectroscopy can, in suitable cases, provide important information on the hydration process and has been used to identify the sites of water sorption. The infrared spectrum (i.e., the amide I band) is also sensitive to changes in protein conformation, however, and it can be difficult, in the absence of other evidence, to distinguish changes in protein conformation from perturbations in the environment and vibrational properties caused by removal of water. Careri et al.  have conducted a detailed study of the infrared spectrum of lysozyme films as a function of hydration. The spectral changes they observed were attributed to removal of water without conformational change, an assignment that may not be entirely correct in light of recent infrared  and NMR [26,27] spectroscopic studies (see Sec. IV). Films were cast on IR transparent windows from aqueous solutions of the protein and were exchanged with D2O vapor in a dry box. The films were dried in a stream of dry nitrogen. Subsequent hydration of the protein films was achieved by passage of a stream of dry nitrogen and D2O vapor over the sample. Hydration levels were monitored gravimetrically and from the integrated intensity of the 1210 cm-1 band, which is assigned to the OD bending mode. Difference spectra determined as a function of hydration for the amide I band at 1660 cm-1 and the carboxylate band at 1580 cm-1 established that the first D20 molecules added to the dry protein hydrated carboxylic acid groups producing carboxylate groups, which accommodated about 40 D2O molecules, consistent with the number of strongly bound water molecules determined from parameters of the D'Arcy-Watt isotherm. Weaker binding sites, primarily main-chain carbonyl groups, accounted for a further 100 water molecules. These results correlate well with regions identified in the sorption isotherm by heat capacity measurements  (see below), namely, that coverage of charged groups in lysozyme is complete at a hydration level of 0.07 h and coverage of polar groups is complete at 0.25 h. Slightly different estimates of these values have been given by Finney and Poole  based on their infrared studies with wedge-shaped protein films. Hydration of carboxylates appears to be complete at 0.1-0.12 h, while NH hydration is completed at about 0.26 h and carbonyl group hydration is complete at about 0.32 h.
As noted earlier, Careri et al.  assumed that no water-induced conformational changes occurred in their infrared studies of lysozyme hydration and at tributed changes in the amide I band entirely to the removal of water. There is now evidence, however, from solid state 13C NMR spectroscopic studies of lysozyme for a dehydration-induced increase in the distribution of protein conformations (static disorder) [26,27]. In addition, recent infrared spectroscopic studies  suggest that conformational change rather than removal of water is responsible for much of the spectral perturbations in the amide I band observed on dehydration. A number of observations suggest that such conformational changes that do occur when a protein is dehydrated are confined to the hydration range below about 0.2 h (see Sec. V). The largest changes in the amide I difference spectrum are associated with this hydration range. Do these changes reflect the addition of water to carbonyl groups and thus allow identification of the region in the sorption isotherm where coverage of carbonyl groups occurs, or do they reflect conformational changes, not necessarily related to the coverage of carbonyl groups? The preservation of efficient dipolar coupling in solid state 13C cross-polarization magic-angle spinning (CP/MAS) NMR spectra on hydration of lysozyme suggests that hydration-induced conformational changes are relatively small in this protein [26,27]. The infrared studies described in Sec. VI, however, indicate that the extent of dehydration-induced conformational changes are highly protein dependent.
The heat capacity of lysozyme has been measured as a function of hydration by Yang and Rupley  and by Finney and Poole , Their work provides a wealth of information and defines a number of distinct regions in the sorption isotherms. The apparent specific heat capacity of lysozyme (component 2) is shown as a function of hydration in Fig. 4 and is defined by
Was this article helpful?