Modeling and mitigation of high-concentration antibody viscosity through structure-based computer-aided protein design

• DOI: 10.1371/journal.pone.0232713
• Publication Date: 2020-05-07T17:28:47Z
• Authors: James R. Apgar; Amy S. P. Tam; Rhady Sorm; Sybille Moesta; Amy C. King; Han Yang; Kerry Kelleher; Denise Murphy; Aaron M. D’Antona; Guoying Yan; Xiaotian Zhong; Linette Rodriguez; Weijun Ma; Darren E. Ferguson; Gregory J. Carven; Eric M. Bennett; Laura Lin
• Keywords: Models, Molecular; 0301 basic medicine; Protein Conformation; Surface Properties; Viscosity; Science; Q; R; Becaplermin; Antibodies, Monoclonal; 03 medical and health sciences; Immunoglobulin G; Mutation; Medicine; Computer-Aided Design; Humans; Research Article

For an antibody to be a successful therapeutic many competing factors require optimization, including binding affinity, biophysical characteristics, and immunogenicity risk. Additional constraints may arise from the need formulate antibodies at high concentrations (>150 mg/ml) enable subcutaneous dosing with reasonable volume (ideally <1.0 mL). Unfortunately, exhibit viscosities that place impractical (such as multiple injections or large needle diameters) on delivery impede efficient manufacturing. Here we describe optimization of anti-PDGF-BB reduce viscosity, enabling increase in formulated concentration 80 mg/ml greater than 160 mg/ml, while maintaining affinity. We performed two rounds structure guided rational design optimize surface electrostatic properties. Analysis this set demonstrated net-positive charge change, disruption negative patches were associated decreased but effect was greatly dependent local environment. Our work here provides comprehensive study exploring wide sampling charge-changes Fv CDR regions along targeting patches. In total, generated viscosity measurements for 40 unique variants full sequence information which significantly larger more complete dataset has previously been reported.