Antibody humanization is used for reducing the immunogenicity of animal monoclonal antibodies (mAbs) and for improving their activities in the human immune system. The development of humanized monoclonal antibodies for use as human therapeutics represents one of the fastest growing segments of the biopharmaceutical industry. More and more humanized mAbs are in clinical trials and marketed as therapeutical drugs.
Monoclonal antibodies are usually first produced in rodents, such as mice and rats, with specific antigens. Most of these mAbs are immunogenic when administered to humans because of the differences in antibody backbone sequences between rodents and humans. Therefore, in many cases, it is necessary to convert non-human antibodies into ones with human like sequences. The antibody humanization process can result in reduced antibody immunogenicity and improved activities in the human immune system, significantly increasing their success rate in clinical studies. In the meantime, it is essential to retain the antigen affinity and specificity of the monoclonal antibodies. The Antibody humanization process is very challenging for many scientists working on antibody drug discovery. Our scientists and scientific advisors involved in the antibody humanization project had working experiences in large biotech and biopharma companies studying, manufacturing and marketing therapeutical antibodies, such as Biogen, Amgen, and Genentech. We have developed novel and validated process for antibody humanization and obtained humanized antibodies with affinities comparable to those of the parental rodent monoclonal antibodies in as short as 2 months.
Our antibody humanization platform combines the advantages of both rational and empirical approaches. After a thorough analysis of sequence and structural information related to the parent monoclonal antibody to get humanized, a framework region (FR) is selected from the most homologous human mature antibody and germline sequences, and three FR choices are used in the design to avoid potential limitation posed by a specific framework. This is followed by sequence homology-based CDR (complementarity-determining region) and SDR (specificity-determining residues) analysis, as well as transfer of parent CDR/SDR into the selected framework. A proprietary 3-D structure modeling based approach is used to identify positions in the human FR, Vernier zone, and canonical structure residues that need to be back mutated to restore CDR conformation and optimal antigen binding. The structural algorithm generates multiple possibilities for each position to allow maximal flexibility at this design stage, and all these residue combinations is put through a molecular evolution process to identify the best antibody sequence to achieve the highest activity and specificity.
Antibody humanization may result in decreased antigen-antibody binding affinity. To optimize the antigen-antibody binding affinity, an antibody library containing all combinations of potential residue choices is constructed. The antibody library is presented on the surface of phage or yeast, and antibodies with the highest affinity toward the desired antigen are identified through multiple screening/amplification steps with increasing stringency. These humanized antibodies are further evaluated for their binding affinity, specificity, and production yield in transiently transfected CHO cells (which is a good indicator of antibody stability), and several antibody molecules with balanced favorable properties are selected for further sequence analysis and characterization. These optimized antibodies can be further tested for their suitability in large-scale production if needed.
By combining sequence homology/3-D structure analysis and molecular evolution approaches, we are able to generate humanized antibodies with affinities comparable to the parent antibodies in a short time frame.