AI Assists in Therapeutic Monoclonal Antibodies Development

Posted by Lisa George on December 25th, 2020

Since 1985, about 100 independent monoclonal antibodies (mAb) have been developed for use as drugs. mAb is a homogeneous preparation of antibodies (or antibody fragments). The protein sequence of each antibody in the product is the same. Therefore, it is expected that the antigen recognition site, affinity, biological interaction and downstream biological effects of each antibody are the same. This distinguishes mAbs from polyclonal antibodies, which have heterogeneous protein sequences and recognize different epitopes on the same antigen.

Some basic facts about therapeutic monoclonal antibodies

Technology is used to modify and mass-produce mAbs, so that they can be used as therapeutic drugs in patients. Initial antibody selection-the key to determining the effectiveness of mAbs is the quality of the interaction between the hypervariable region and the target antigen, also known as complementarity-determining region (CDR). The selection of target antigens is usually based on a scientific understanding of disease mechanisms and/or disease-specific antibody effects observed in preclinical models or individual patients.

The downstream effect of antibody-antigen binding is also the key to ensure clinical efficacy and low toxicity. Although the immunogenicity of mAbs is more complicated (that is to say, it is not just a matter of the number of amino acid residues), these effects can be reduced if antibodies that do not contain certain epitopes in the heterogeneous (namely, non-human species) are used.

The following methods can be used to construct antibodies that react with the intended target:

Immunize animals

Use target antigens to immunize animals, usually mice or rats. This is the most common and only technically feasible method in the early stage of mAb production. Animal spleen cells are used to obtain candidate B cells for the production of therapeutic mAbs specific to the target. Orthoclone OKT3 was obtained in this way.

Serious risks may accompany with this method. Some patients will develop an immune response to mouse antibody sequences after exposure to mouse antibodies. Once a patient develops human anti-mouse antibodies, they usually can no longer continue to receive the original mAb or other therapeutic mAbs with similar mouse sequences. Therefore, some technologies have been developed to modify immunoglobulin molecules, such as humanizing antibodies or constructing chimeric antibodies. Most of the mAbs initially selected in animals will use these technologies, namely, use human immunoglobulin gene locus instead of murine sequence to transform mice, thereby producing human antibodies in mice.

Obtain existing antibodies

Isolate existing antibodies against the target antigen from the patient. This method is particularly suitable for cancer treatment, because its conventional treatment usually includes removal of tumors and/or regional lymph nodes, and these tissues can be used to obtain tumor infiltrating lymphocytes. Existing antibodies can also be isolated from peripheral blood, bone marrow or other lymphoid tissues, such as the spleen or tonsils. Examples of application of this method include various mAbs under development that target viruses such as HIV and hepatitis C virus (HCV).

Screening of antibody libraries

Antibody libraries are constructed using molecular technology or purchased. The antibody library can be screened in vitro to obtain antibodies that can bind to the target antigen. The size and diversity of each antibody library vary greatly. The antibody library can be constructed using phage display technology or other combinations. Phage is a virus that can infect bacteria. When using phage display technology to construct an antibody library, a large number of sequences are inserted into phage in a ratio so that each phage clone generates a single antibody or antibody fragment. Researchers can adjust the size and diversity of the antibody library. A larger and more diverse antibody library is more likely to obtain therapeutic mAbs or mAb fragments with the highest affinity and specificity for the target antigen. Therapeutic mAbs derived from the phage display antibody library include adalimumab, raxibacumab and belimumab.

AI helps to select therapeutic monoclonal antibodies

By using AI to analyze a large amount of experimental data to make up for the lack of automatic data analysis by computers will help researchers find antibodies that can be used under any unique experimental conditions. Another way is to index all antibody details by browsing a large amount of scientific literature and using intelligent computer programs based on machine learning. This can effectively standardize the experimental program and improve the research efficiency of researchers. In order to promote new scientific discoveries at a faster speed, the choice of reagents has the following advantages:

1. Improve the efficiency in the antibody selection process and reduce the probability of errors;

2. Can quickly select reagents within 30 seconds;

3. Reduce the hard cost of consumables up to US million per year;

4.  Shorten the research time of scientists.

MedAI is an expert that newly arises in the AI-assisted drug development field. It recently launched AI systems that can help suggest a suitable approach of antibody development, thus reducing the risks of failure and the duration of pre-clinical steps, as well as strengthening intellectual protection.