Static light scattering from concentrated protein solutions II: experimental test of theory for protein mixtures and weakly self-associating proteins

Abstract

Using an experimental technique recently developed in this laboratory (Fernández C. and A. P. Minton. 2008. Anal. Biochem. 381:254-257), the Rayleigh light scattering of solutions of bovine serum albumin, hen egg white ovalbumin, hen egg white ovomucoid, and binary mixtures of these three proteins was measured as a function of concentration at concentrations up to 125 g/L. The measured concentration dependence of scattering of both pure proteins and binary mixtures is accounted for nearly quantitatively by an effective hard particle model (Minton A. P. 2007. Biophys. J. 93:1321-1328) in which each protein species is represented by an equivalent hard sphere, the size of which is determined by the nature of repulsive interactions between like molecules under a given set of experimental conditions. The light scattering of solutions of chymotrypsin A was measured as a function of concentration at concentrations up to 70 g/L at pH 4.1, 5.4, and 7.2. At each pH, the measured concentration dependence is accounted for quantitatively by an effective hard particle model, according to which monomeric protein may self-associate to form an equilibrium dimer and, depending upon pH, an equilibrium pentamer or hexamer.

Figure 1

Dependence of normalized scattering intensity upon total molar concentration of BSA, ovalbumin, and binary mixtures. Experimental data: circles, squares, triangles, pentagrams and diamonds are results from solutions containing BSA mole fraction 1 (pure BSA), 0.8, 0.5, 0.2, and 0 (pure ovalbumin), respectively. Solid curves for the pure proteins were calculated using the effective hard sphere model for a single species with best-fit parameter values given in Table 1. Dashed curves for the mixtures were calculated using the effective hard sphere model for two species and mole fractions constrained to the values given above. Solid curves for the mixtures were calculated using the effective hard sphere model for two species with adjustable mole fraction of BSA using the following best fit values: 0.84, 0.55, and 0.24 respectively.

Figure 2

Dependence of normalized scattering intensity upon total molar concentration of BSA, ovomucoid, and an equimolar mixture. Experimental data: circles, squares, and triangles, are results from solutions containing BSA mole fraction 1, 0.5, and 0, respectively. Solid curves for the pure proteins were calculated as described above. Dashed curve for the equimolar mixture was calculated using the effective hard sphere model for two species and BSA mole fraction constrained to 0.5. The solid curve for the equimolar mixture was calculated using the effective hard sphere model for two species with adjustable mole fraction of BSA using a best fit value of 0.58.

Figure 3

Dependence of normalized scattering intensity upon total molar concentration of ovalbumin, ovomucoid, and an equimolar mixture. Experimental data: circles, squares, and triangles are results from solutions containing ovalbumin mole fraction 1, 0.5, and 0, respectively. Solid curves for the pure proteins were calculated as described above. The dashed curve for the equimolar mixture was calculated using the effective hard sphere model for two species and ovalbumin mole fraction constrained to 0.5. The solid curve for the equimolar mixture was calculated using the effective hard sphere model for two species with adjustable mole fraction of ovalbumin using a best fit value of 0.55.

Figure 4

Concentration dependence of normalized scattering intensity of chymotrypsin A solutions at pH 4.1 (A), pH 5.4 (B), and pH 7.2 (C). Circles: Experimental data; thick solid curve: best fit of the three species effective hard sphere model with allowance for formation of equilibrium dimer and one higher oligomer, using the best-fit parameter values indicated in Table 2. Best fit curves calculated from the monomer-dimer-pentamer and monomer-dimer-hexamer models at pH 5.4 are indistinguishable. Also plotted in each panel are reference curves calculated according to the one-species effective hard sphere model for hypothetical nonassociating monomer (thin solid line), dimer (dashed), pentamer (dotted) and hexamer (dot-dashed).

Figure 5

Fractional abundance of species plotted as a function of the logarithm of total protein concentration at pH 4.1 (A), 5.4 (B), and 7.2 (C), calculated as described in Minton et al (5) using the best-fit parameter values given in Table 2. Species abundance at pH 5.4 is calculated according to the best-fit monomer-dimer-hexamer model (solid curves) and the best-fit monomer-dimer-pentamer model (dashed curves).

Figure 6

Comparison of values of the pH-dependent monomer-dimer equilibrium constant presented in Table 2 (+ symbols) with values previously reported for chymotrypsin A in the literature (all other symbols as indicated in the caption to Fig. 4B of reference 15).