DSF method optimization and its application in predicting protein thermal aggregation kinetics

Differential scanning fluorimetry (DSF) has gained wide acceptance in the therapeutic protein development. However, the effects of dyes and surfactants that may affect structural transitions have not been studied thoroughly to date. We therefore first optimized the DSF method by studying surfactant-containing formulations and found that the presence of surfactants generally required medium-to-high protein concentrations and that high SYPRO® Orange concentration in a DSF experiment may lower protein thermal transitions. We also benchmarked DSF against differential scanning calorimetry (DSC) and evaluated the capability of thermal parameters (from DSF/DSC) to predict real-time thermal aggregation kinetics monitored by size exclusion chromatography (SEC) and analytical ultracentrifugation (AUC) in different scenarios. For monoclonal antibody (MAb) fragment, both DSF and DSC were predictive of thermal aggregation rate. For MAb3, a good correlation was observed between DSF and DSC, none of which was, however, indicative of protein aggregation kinetics. In a surfactant ranging study, DSF did not agree with DSC and was not predictive of the aggregation kinetics of the MAb fragment. The concentration-dependent thermal behavior was also studied by DSF. Although higher concentration, in general, tends to lower protein transition temperature, case where it was independent of protein concentration was also presented.