Therapeutic mAb Engineering Considerations

Monoclonal antibodies (mAbs) are classic examples of personalized diagnostics and therapeutics (Malik and Ghatol, 2021). They bind specifically to nearly any therapeutically pertinent biomolecule and can be produced by a wide range of engineered organisms (Eidenberger et al., 2023). A common use of mAbs is to help the immune system target cancer (Zinn et al., 2023) and inflammatory disease (Zhu et al., 2023), such as for past emergency use against SARS-CoV-2 (Ravi et al., 2023). Well over 100 mAbs have been approved for use in human therapeutics (Lyu et al., 2022). The pertinence of therapeutic mAbs is perhaps best indicated by its global market size: estimated at over $162 billion in 2021, and predicted to reach over $390 billion by 2030 (Straits Research, 2022).

It is challenging to identify and engineer the mAb characteristics that are most likely to result in successful clinical trials (Jhajj et al., 2023). Using computational tools to optimize mAbs can help in such efforts (Makowski et al., 2023). Here, we provide an overview of pertinent issues and industry solutions to mAb engineering, and computational approaches that will speed up research and development.

Main considerations for mAb engineering

Immunoglobulin G (IgG) is the only class of antibodies that is currently used as therapeutic mAbs (Malik and Ghatol, 2021). IgG consists of two arms and a leg, forming a Y shape. The arms are the antigen-binding fragments. The complementarity-determining regions within the arms are highly variable and are essential to the specificity of the mAb. The leg is the crystallizable fragment, which activates the immune system against the antigen, such as an essential biomolecule of tumor biology. Understanding any conformational changes that might occur upon mAb binding to the antigen, as well as the comprehensive kinetics of binding rather than simply mAb binding affinity, is essential to optimizing the pharmacology of engineered mAbs (Chiu et al., 2019).

There are various methods for mAb engineering; the traditional experimental approach is to generate and screen large mAb libraries. “We recommend screening antibodies produced with hybridoma, yeast display, and mammalian cell display technologies by ELISA and fluorescence-activated cell sorting,” says Qian Gao, Senior Product Manager at ACROBiosystems. “For antibodies produced with phage display, we recommend biopanning, which has been quite helpful for finding antibodies against SARS-CoV-2 and other deadly viruses.”

Feng Hao, Director at KYinno Biotechnology, explains some of the advantages and limitations of various mAb engineering approaches: “The hybridoma approach is more traditional, and we have two transgenic mouse models for antibody discovery. However, there is a lack of stable fusion partners for many lab animals except mice and rats. Alternatively, whereas phage display and single B cell screening are more high-throughput, the former has low developability and the latter has high instrumental and reagent costs.”

Hao further recommends that mAb researchers “undertake comparisons with on-shelf, well-validated benchmark antibodies where possible; and screen for nonspecific binding, as we do with our AB5000 platform.” In this context, also consider that mAb specificity in clinical applications might be hindered by trogocytosis (biomolecular extraction from the cell surface; Ochs et al., 2023).

The two main bottlenecks of mAb engineering

Minimizing mAb structural variants and optimizing product formulation are especially challenging. A single mAb product contains many variants, attributable largely to errors in post-translational modifications—such as glycosylation—during manufacturing (Carrara et al., 2021). Such errors can hinder the efficacy, activity, and stability of the mAb. One can minimize variants in the mAb product by applying protein engineering (such as amino acid exchange) to the mAb and optimizing the processing parameters (such as continuous cultivation). Even after optimizing production steps, raw material costs are high: 20× to 100× higher than for conventional pharmaceuticals. In one case study, removing use of protein A in the affinity chromatography workup reduced raw material costs by 39% (Bansode et al., 2022).

Product formulation is laborious because of the complex architecture of mAbs (Cararra et al. 2021). The details of formulation depend on the route of administration: intravenous is the most common route, excluding the intramuscular route for vaccines. For example, preparing a lyophilized powder for subsequent reconstitution can damage mAbs. However, sugar excipients—appropriate for both solid and solution state formulations—can minimize mAb aggregation and degradation.

Computation to the rescue

Experimental engineering of thousands to millions of mAb variants for screening can be burdensome and might lead to inappropriate biophysical parameters such as low solubility and limited large-scale manufacturability (Kim et al., 2023). Thus, computational approaches that minimize subsequent experimental work can be useful. In this context, Hie et al. (2023) trained general protein language models on millions of natural protein sequences. Their work was based on the mAb wild-type sequence, and did not require initial task-specific training data. They devised seven human IgG antibodies that bound to various viral antigens. Screening no more than 20 mAb variants for each mAb over two rounds of experimental evolution substantially improved mAb affinity, such as by two orders of magnitude against an Ebola pseudovirus. A plausible further application of this work is to directed evolution against enzymatic activity and antibiotic resistance.

mAbs have unique utility in treating cancer, inflammatory, and other diseases. However, mAbs are not conventional drugs and cannot be easily produced as homogenous, readily formulated products. As such, mAb engineering is a complex task that imparts substantial manufacturing, screening, and formulation challenges. Computational tools can help minimize the experimental work that is necessary for delivering a mAb product to market.

References

Bansode V, et al. (2022). Contribution of protein A step towards cost of goods for continuous production of monoclonal antibody therapeutics. J. Chem. Technol. Biotechnol. 97(9):2420–2433.

Carrara SC, et al. (2021). From cell line development to the formulated drug product: The art of manufacturing therapeutic monoclonal antibodies. Int. J. Pharm. 594:120164.

Chiu ML, et al. (2019). Antibody structure and function: The basis of for engineering therapeutics. Antibodies 8(4):55.

Eidenberger L, et al. (2023). Plant-based biopharmaceutical engineering. Nat. Rev. Bioeng. 1:426–439.

Hie BL, et al. (2023). Efficient evolution of human antibodies from general protein language models. Nat. Biotechnol. (published ahead of print, Apr. 24, 2023).

Jhajj HS, et al. (2023). Unlocking the potential of agonist antibodies for treating cancer using antibody engineering. Trends Mol. Med. 29(1):P48–P60.

Kim J, et al. (2023). Computational and artificial intelligence-based methods for antibody development. Trends Pharmacol. Sci. 44(3):P175–P189.

Lyu X, et al. (2022). The global landscape of approved antibody therapies. Antib. Ther. 5(4):233–257.

Makowski EK, et al. (2023). Simplifying complex antibody engineering using machine learning. Cell Syst. 14(8):P667–P675.

Malik B and Ghatol A. (2021). Understanding how monoclonal antibodies work. StatPearls Publishing, Treasure Island, FL (last accessed Sep 21, 2023). PMID: 34283484.

Ochs J, et al. (2023). Trogocytosis challenges the cellular specificity of lineage markers and monoclonal antibodies. Nat. Rev. Immunol. 23:539–540.

Ravi G, et al. (2023). Efficacy and safety of anti-SARS-CoV-2 monoclonal antibodies: An updated review. Monoclon. Antib. Immunodiagn. Immunother. 42(2):77–94.

Straits Research (2022). Monoclonal antibodies market: Information by source (murine, chimeric, humanized, human), indication (cancer, autoimmune diseases), end-user (hospitals), and region—Forecast till 2030. Pune, India.

Zhu CS, et al. (2023). Identification of procathepsin L (pCTS-L)–neutralizing monoclonal antibodies to treat potentially lethal sepsis. Sci. Adv. 9(5):eadf4313.

Zinn S, et al. (2023). Advances in antibody-based therapy in oncology. Nat. Cancer 4(2):165–180.