Partial specific volumes and interactions with solvent components of proteins in guanidine hydrochloride

Abstract

The partial specific volumes of twelve proteins were determined by density measurements. For these proteins in their native state, the determined values of partial specific volumes are generally in good agreement with the accepted literature values except in the case of bovine a-lactalbumin. The determined value of 0.704 is similar to that found for lysozyme. Preferential interaction parameters of 6 M guanidine hydrochloride with these proteins were also measured. For the twelve proteins studied, the preferential interaction with solvent components vanes between 0 and 0.17 g of guanidine A lthough an exact knowledge of the partial specific volume, 6, of a protein is essential for the determination of molecular weights from ultracentrifuge data and small-angle X-ray scattering, in the past this parameter was seldom measured. The molecular weights were calculated from values of 6 assumed, or calculated from amino acid composition. And yet the importance of accurate measurements of the partial specific volume was generally recognized, since a small error in that parameter is multiplied several fold in the calcula- tion of the molecular weight, in particular when measurements are carried out in concentrated solutions of denaturant, such as 6 M Gdn.HCl,I a frequent practice in studies of subunit systems. Uncertainties in estimates of molecular weight can lead to wrong conclusions about the number of subunits in the native macromolecular assembly and to serious errors in the calculation of the thermodynamic parameters of associating systems. A classical example is the uncertainty which prevailed for several years about the exact .number of polypeptide chains in rabbit muscle aldolase (Kawahara and Tanford, 1966; Schachman and Edelstein, 1966; Castellino and Barker, 1968; Reisler and Eisenberg, 1969 ; Meighen and Schachman, 1970). Until recently, the methods available for the measurement of the partial specific volume either required frequently prohibi- tive amounts of material, as in conventional pycnometry, or involved long complicated procedures, as in the density gradient column technique (Linderstrgm-Lang and Lanz, 1935; Hvidt ef al., 1954; Reithel and Sakura, 1963). Recently, several new and elegant approaches to the measurement of 8, requiring small amounts of material, have been described. These include the HD-D20 method of Edelstein and Schach- man (1967), the magnetic float method of Ulrich et af. (1964), and the precision densimeter, based on the frequency of vibra- tion of a tuning fork, consisting of a sample-filled quartz hydrochloride per g of protein. In no case is interaction preferential with water. The total binding of denaturant to protein is calculated and the correlation between the observed and expected number of denaturant molecules bound is found to be good, if peptide bonds and aromatic side chains are taken as the binding sites. The changes in volume upon transfer from dilute salt to 6 M guanidine hydrochloride of the proteins studied are calculated and compared with theoretical values reported in the literature.