(681f) Engineering Humanized Monoclonal Antibodies for Improved Expression in Chinese Hamster Ovary Cells

Monoclonal antibodies represent a significant fraction of recently approved and pipeline biopharmaceuticals, providing critical new therapies for diseases such as breast cancer and rheumatoid arthritis. As new strategies have emerged for antibody engineering to improve clinical efficiency by improving affinity, potency, pharmacokinetics and pharmacodynamics, it has become apparent that modification to the nucleic acid and protein sequence can also substantially impact manufacturing considerations, including expression levels, purification, aggregation, and stability.

We have previously identified a humanized monoclonal antibody in which a single amino-acid change (from alanine to glycine) in the Kabat position 49 (the heavy chain variable region one residue before the CDR2) led to a four-fold decrease in productivity when expressed in CHO cells. To identify the mechanism for this change in productivity, we created a panel of antibody variants containing all 20 amino acids except glutamine in position 49 and screened for expression level in CHO cells. Small, uncharged residues, particularly with hydroxyl groups (i.e. threonine and serine) gave the highest levels of expression, while bulky substituents and charged groups gave very low levels of expression.

Current studies are focused on understanding the regulatory mechanisms for these differences. We hypothesize that the primary effects of sequence on productivity are on the rates of protein folding and the stability of the folded proteins with ?less stable? variants more subject to intracellular degradation, leading to reduced secretion of properly folded proteins.