Determination of protein-protein interactions at high co-solvent concentrations using static and dynamic light scattering

Protein-protein interactions are commonly measured in terms of the second osmotic virial coefficient, B22from static light scattering (SLS) or the interaction parameter, kDfrom dynamic light scattering (DLS). Often these measurements are carried out at high co-solvent compositions, where correction factors are required for the light scattering analysis. For lysozyme in aqueous solutions containing the co-solvents NaCl, arginine chloride, urea, sucrose or guanidine chloride, we show that B22determination requires using in the light scattering equation the refractive index increment of the protein measured at constant solvent chemical potential. Because the increment decreases with increasing co-solvent composition, using a constant value can lead to mis-interpretation of protein-protein interaction trends deduced from the B22measurements. Furthermore, there is a contribution to the intensity auto-correlation function measured by dynamic light scattering due to co-solvents. This effect is removed by including longer delay times when fitting the cumulant analysis to determine the diffusion coefficients. We show that an experimentally observed correlation between B22and kDis recovered once these correction factors have been applied. The findings are particularly relevant to biopharmaceutical industry, where B22and kDmeasurements are used for screening excipient effects in liquid formulations.