SynopsisA computer program, based upon the method of Trautman, has been developed to calculate density-gradient proportionality constants, /3O, as a function of density for any salt for which density and activity coefficient data are available. Results are given for twenty 1: 1 electrolytes at 25°C. These salts are the chlorides, bromides, iodides, and nitrates of the five alkali cations. The program is available for use for any other salt and for a variety of temperatures if the necessary data can be found or measured. Between six and eleven centrifuge runs at different densities were performed for each of seven of these salts experimentally to determine Po. In general, the experimental and theoretical values agree within 3% except a t the extremes of density. This demonstrates the validity of the computer program and is the first extensive demonstration that the thermodynamic calculation of density gradients is correct. Five new quadratic relationships between refractive index and density are given.

SynopsisThe buoyant density titrations of five ionizable copolypeptides in concentrated CsCl solutions have been determined. The results are used to formulate models for predicting the buoyant density titration behavior of copolypeptides and proteins using the previously reported homopolypeptide buoyant density titratiop curves. It was determined for these copolypeptides that the best predictive model must include not only the buoyant densities of the constituent amino acid residues and the relative composition, but also hydration and salt binding.Hydrations determined for the homopolypeptides are used in the copolypeptide predictive model. The hydrations of the neutral homopolypeptides were readily calculable since their buoyant densities are thermodynamically defined in terms of their partial specific volumes and hydrations. For the case of a charged macromolecule, an expression for the buoyant density as a function of the number and nature of the bound ions, its partial specific volume, and its relative hydration has also been available for some time. This heretofore intuitive relationship is now derived from thermodynamic principles and allows calculations of hydrations to charged macromolecules which bind either cations, anions, or both.The potentiometric titrations of three of the five copolypeptides in concentrated CsCl solutions were determined in order to study the effect of residue interaction and solvation effects on their ionization behavior. The potentiometric results are also combined directly with the buoyant density titration results to determine the correlation of the buoyant density with the degree of ionization. As in the cases of poly(Glu) and poly(His), the buoyant density of the copolypeptides changed linearily with the degree of ionization.The buoyant density titrations of two nonionizahle homopolypeptides, poly(G1y) and poly(Ala), were determined in concentrated CsCl solutions. The buoyant density was * Part I: Almassy et al.'SHARP ET AL. found to increase with increasing pH, despite the fact that side chains do not contain ionizable groups. This is the first evidence from homopolypeptide or copolypeptide data that buoyant density changes can be observed from effects other than side-chain ionizations.

[article in Biopolymers 17,817-836 (1978)] Previously reported' standard pressure coefficients (Go) and standard partial specific volumes of the solvated component (i&) for egg albumin (EA), pooled human y-immunoglobulin (IgG), and bovine serum mercaptalbumin (BMA) are in error. These errors are due to the incorrect use of the formula2 to calculate the quantity ( r f -r:). The effect of using the correct expression for (r: -r:) is to substantially decrease in magnitude Ap-especially a t high angular velocities (w2). This effect substantially changes the slopes of the Ap-vs-PO plots for the aforementioned proteins. The new values for Us,o for EA at pHs 10.14 and 7.14 and IgG at pH 7.00 are 0.766,0.762, and 0.772 ml/g, respectively. The new values of $O are 35.9 f 1.3,34.6 f 1.1, and 9.6 f 5.1 X atm-*, respectively. Note that Go is positive for IgG in contrast to the negative value reported previously, but is not significantly different from zero at p = 0.05. The Ap-vs-PO plot for BMA at pH 5.34 using the correct expression for ( r f -r;) retains its curvilinear characteristics. No values of Uz,O and $0 are reported for BMA because of this curvilinear character and the lack of data at low pressures. The discussions regarding the physical significance of these values for Us,O and $O in the previous report' should be reevaluated accordingly.

The compositional buoyant densities, ρ 00;, of human γ‐immunoglobulin, bovine serum mercaptalbumin, and egg albumin have been measured in CsCl solutions in the analytical ultracentrifuge as a function or pressure. Standard pressure coefficients, ψ0, and standard partial specific volumes of the solvated proteins, υ 0italicS,0, have been computed from these data. The ψ0 values obtained are strikingly different from each other and from the only other pressure coefficients which have been measured, those values obtained for nucleic acids and nucleoproteins. The ψ value for γ‐immunoglobulin is negative, the first nonpositive value obtained, and suggests an unusual internal structure for this protein. The pressure coefficient of mercaptalbumin is not constant. A second‐order relation is derived and utilized to interpret these data. The slope of the ρ 00(P) plot for egg albumin was constant and negative and yielded values of ψ0 which are about 20% as large as those reported for DNA. Evaluation of published isopiestic data for egg albumin in CsCl solutions provided the dependence of preferential hydration on water activity. This quantity, (dΓ′/da 00) as well as α, were found to be negative. The values of ψ0 and α were used to compute the effective density gradient from which the correct molecular weight of egg albumin was obtained. The apparent specific volume of egg albumin in a buoyant CsCl solution was measured using the Mettler‐Paar densimeter.

The buoyant titrations of poly-L-lysine and poly-L-histidine in CsCI, RbCI, CsBr, RbBr and KBr were measured. Large differences in buoyant densities measured at low pH were observed for both polymers. Densities in the alkali chloride solutions are lower than for the bromides and the buoyant density increases as the size of the cation decreases in the halide series. All buoyant densities converge to a common value for each polymer at high pH. These data are interpreted in terms of salt-pair formation at low pH and preferential hydrations were computed both for this charged species at low pH and for the neutral polymers at high pH. The resulting data were correlated with the water activity in each buoyant solution. The variation of hydration with water activity for the charged species is found to be similar to that previously reported for bovine serum mercaptalbumin while the hydrations of the neutral polymers are found to be nearly independent of water activity. Approximate observed ionization constants were determined for each polymer in four salt solutions.