Method qualification and application of diffusion interaction parameter and virial coefficient

This research focused on evaluation and application of two methods in studying weak protein–protein interactions, i.e. diffusion interaction parameter (KD) and second virial coefficient (B22), both of which are first-order coefficients of protein interactions. Although the plate-based KD method successfully distinguished KD values with relatively large difference in a pH ranging study, it failed to make a consistent statistical decision to determine close interactions as shown by the comprehensive ANOVA analysis. We also validated the DLS-based B22 method by using a model protein lysozyme. The dramatic change of solution appearance for lysozyme as a function of NaCl concentration highlighted the importance of B22 in understanding protein interactions. Moreover, B22 measurement for a MAb fragment suggested a more repulsive protein interaction in histidine buffer than in citrate buffer. The coefficient of variation was <10% when B22 was on an order of magnitude of 10−4 L mmol/g2 in contrast to >30% when it approached 10−5 L mmol/g2. In this research, we also made an attempt to study protein–protein interactions in concentrated MAb fragment solutions (e.g. >50 mg/mL). Our data suggested that such interactions could be empirically modeled by high-order virial expansions.