Human-engineered monoclonal antibodies retain full specific binding activity by preserving non-CDR...

Area of Science:

• Immunology
• Biotechnology
• Structural Biology

Background:

• Conventional antibody humanization via complementarity-determining region (CDR) grafting can reduce binding affinity.
• Framework residues significantly influence CDR structure and antibody-ligand affinity.
• A need exists for improved methods to humanize antibodies while preserving or enhancing function.

Purpose of the Study:

• To present a general model for antibody variable region design based on structure-function relationships.
• To classify amino acid positions by risk (low, moderate, high) for human substitutions.
• To experimentally validate the model using a humanized anti-CD5 antibody (H65).

Main Methods:

• Developed a model classifying amino acid positions in antibody variable regions based on risk to binding affinity and immunogenicity.
• Classified substitutions as low-risk (solvent-exposed, non-critical), moderate-risk, or high-risk (antigen-binding, structural).
• Applied the model to engineer the H65 murine monoclonal antibody, testing low-risk and moderate-risk substitutions.

Main Results:

• The 'human-engineered' H65 antibody with 20 low-risk substitutions retained full binding avidity.
• An engineered H65 variant with 14 additional moderate-risk substitutions showed enhanced avidity (3- to 7-fold).
• The model successfully preserved and even improved antibody binding characteristics after humanization.

Conclusions:

• The presented model offers a strategy for designing human-engineered antibodies with reduced immunogenicity and preserved or enhanced binding affinity.
• This approach is broadly applicable for developing therapeutic antibodies with improved safety and efficacy profiles.
• The classification of amino acid positions provides a rational framework for antibody engineering.