Hierarchical clustering of therapeutic proteins based on agitation-induced aggregation propensity and its relation to physicochemical parameters

Abstract

Physical stresses such as agitation induce protein aggregation, which causes adverse effects on the immune system of patients, leading to challenges in drug development. Aggregation induced by physical stresses can be minimized by formulation optimization. In this study, 120 combinations of 10 therapeutic proteins and 12 different formulations (4 pH conditions and 3 salt concentrations) were prepared. Subsequently, the agitation-induced aggregation propensity of each protein was investigated by evaluating its monomer recovery (%) using size exclusion chromatography. Hierarchical clustering was applied to categorize each protein according to its aggregation propensity, resulting in two groups of proteins: group A and B. The aggregation propensity of proteins in group A was insensitive to changes in formulation conditions because conformational, colloidal, and interfacial stabilities were minimally affected by changes in the pH and salt concentration and a compensation mechanism existed between conformational and colloidal stabilities. Thus, proteins in group A can be formulated with a relatively high degree of freedom. In contrast, the aggregation propensity of proteins in group B was sensitive to changes in formulation conditions. Multiple regression analysis of the physicochemical parameters and monomer recovery of proteins in group B clarified that changes in conformational stability in response to changes in formulations primarily contributed to the sensitivity of the monomer recovery to changes in formulation conditions. For all antibodies, there was a positive correlation between the monomer recovery after agitation and that after quiescent storage at 40 °C for 1 month, suggesting that a stable formulation can be obtained without the quiescent testing. Therefore, a proposed formulation optimization strategy based on the agitation-induced monomer recovery can improve the efficiency of formulating selected therapeutic proteins. This strategic approach is expected to accelerate the development of therapeutic proteins while reflecting the importance of aggregation factors and quiescent stability in the optimization of therapeutic protein formulations.