In designing and manufacturing therapeutic antibodies, a top objective is precise binding to an intended target. Sometimes, the antibodies can bind to each other through self-association. In a recent article in mAbs, Peter Tessier, PhD, the Albert M. Mattocks professor in the departments of pharmaceutical sciences, chemical engineering, and biomedical engineering at the University of Michigan in Ann Arbor, and colleagues noted: “Antibody self-association is a particularly important property because high self-association is linked to increased risks for high viscosity, aggregation, and/or opalescence, especially for high-concentration formulations used for subcutaneous delivery.”
Tessier’s team developed what they describe as “a high-throughput protein engineering method for rapidly identifying antibody candidates with both low self-association and high affinity.” This work included screening antibodies for variants with lower levels of self-association and then analyzing them with deep sequencing and analysis based on machine learning.
When asked about the key finding from this work, Tessier says that it was “our demonstration that bococizumab—a clinical-stage antibody with notoriously poor biophysical properties, including high self-association and nonspecific binding—can be dramatically improved using only a few mutations in its binding loops while maintaining high binding affinity.” He added, “Surprisingly, these mutations even significantly increase the folding stability of the antibody.”
This is just a start in applying this approach to therapeutic antibodies. “Our strategy now needs to be tested for other antibodies during the discovery and optimization stages of drug development to evaluate its generality, explains Tessier, whose team eventually could turn this method into a tool that helps bioprocessors make antibody-based therapies that maintain the desired properties.”
