Applicability of Using Physical Stability Data and Advanced Visualization Methods in Protein Comparability Studies

Abstract

Comparability assessments are frequently performed during biopharmaceutical drug development to evaluate the effects of manufacturing process changes on the structural integrity, safety and efficacy of candidate protein drug products. Physiochemical analytical evaluations (an essential element of any comparability exercise) involve multiple assessment techniques mapping the different aspects of protein structural integrity (primary and higher-order structures). Although significant advances in regards to primary structure evaluations have been made including peptide and oligosaccharide mapping techniques combined with mass spectrometry, the need for more sensitive approaches for the evaluation of higher order structures of proteins, especially in their final pharmaceutical dosage forms, remains an important challenge. In this dissertation, the ability of using multiple protein conformational stability studies (combined with novel data visualization approaches) as surrogates for evaluating and comparing different proteins's higher-order structure are explored in the context of comparability. Empirical phase diagrams (EPDs) and radar charts, constructed using large data sets from high throughput, lower-resolution biophysical techniques, are used to detect major (point mutations of FGF-1) and minor (different glycosylation patterns of an IgG1 mAb and Fc molecules) structural differences between proteins. In addition, for the IgG1 Fc glycoforms expressed in yeast, the effect of N-linked glycosylation site occupancy, as well as the effect of different charge states in various nonglycosylated forms, within the CH2 domain was examined. Using this approach, differences in conformational stability were detected under stress conditions that did not necessarily detect structural integrity differences in these proteins in the absence of these stresses. Thus, an evaluation of conformational stability differences may serve as an effective surrogate to monitor more subtle differences in higher-order structure between protein samples as part comparability assessments.