Written and maintained by CASRAI Editorial Board
Published · Last updated
On this page
Monoclonal antibodies are made in two stages: a discovery route (hybridoma, phage display, transgenic mice or single B cell cloning) generates the antibody, and for therapeutics it is then expressed in fed-batch Chinese hamster ovary (CHO) culture and purified by Protein A capture, polishing chromatography, virus filtration and ultrafiltration/diafiltration.
Short answer: discovery first, then expression and purification
The word “produced” covers two different jobs that are easy to blur. The first is finding or generating one antibody clone that binds the target. The second is manufacturing that clone as a consistent protein product. A 2021 review of antibody discovery technology notes that the hybridoma method (Köhler and Milstein, 1975) generated more than 90% of the antibodies approved for therapeutic use, most of them in chimeric or humanized versions, which shows how the two stages connect: a hybridoma-derived antibody is usually re-engineered, then manufactured recombinantly. A 2024 history of antibody manufacturing describes the industry platform that most products now follow after discovery: fed-batch CHO culture, Protein A capture, one or two polishing chromatography steps, virus filtration, and concentration and formulation by ultrafiltration and diafiltration.
This page is an overview for researchers, lab staff and supply and procurement readers who want the whole chain in one place. It is general information about methods, not a bench protocol, and it is not medical advice. The step-by-step hybridoma procedure has its own page, linked below.
Definition and scope
A monoclonal antibody (mAb) is an antibody population made by a single clone of cells, so every molecule has the same sequence and binds the same epitope. That is the contrast drawn on the polyclonal vs monoclonal antibodies comparison, where polyclonal preparations are mixtures from many B cell clones. Monoclonal antibodies serve as research reagents (for example as the primary antibodies described in primary and secondary antibodies) and as biologic drugs. The production chain is shared in outline, but a reagent antibody and a drug substance are held to very different manufacturing and regulatory expectations, so this page keeps the two stages separate.
How monoclonal antibodies are produced in practice
Stage 1: generating the antibody (discovery routes)
Four routes are described in the antibody-technology literature. The comparison table at the end of this page (see the data table below) sets them side by side. In short:
- Hybridoma. The method is based on immunizing animals with the desired antigen and then fusing specific B lymphocytes with immortal myeloma cells. The hybrids are cloned to obtain stable monoclonal cell lines, and selected clones are then moved to large-scale culture. The reviewers list low efficiency of cell fusion and hybridoma isolation, and genetic instability of hybridoma cell lines, among the limits. The full immunization, fusion, selection and cloning sequence is covered in the hybridoma guide.
- Phage display. First reported in 1990, this route does not require an animal host. Screening a large number of clones increases the chance of finding a good antibody. The repertoire is limited to roughly 1010 to 1011 variants, and phage-selected antibodies are produced in E. coli and are therefore not glycosylated unless moved to a eukaryotic platform. Ribosome display offers a larger library, around 1014 variants.
- Transgenic mice with human immunoglobulin loci. Immunizing these animals makes possible murine hybridomas that secrete fully human antibodies. The first hybridoma-derived human antibody isolated from transgenic animals to reach approval was panitumumab, approved in 2006.
- Single B cell methods. Antibody genes are recovered directly from individual B cells, from vaccinated or naturally immunized human subjects, without culturing or immortalizing the cells. In experimental studies B cells can be isolated from multiple samples without euthanizing animals. The antibody genes are cloned into expression cassettes and expressed in cell-based or cell-free systems. The reviewers record that single-cell sorting devices are expensive.
The same review reports that hybridoma-derived antibodies intended as drugs generally “need to be converted into human mAbs” by chimerization, humanization or use of transgenic animals. That is the practical bridge between Stage 1 and Stage 2.
Stage 2: expression in culture
For manufactured therapeutic antibodies the platform described in the 2024 history is fed-batch CHO cell culture using chemically defined media. The history traces the pattern back to early products: the first approved antibody, Orthoclone OKT3 (1986), was made in hybridoma cells grown in the mouse peritoneum, before recombinant CHO production took over, and rituximab (launched 1997) is described as a chimeric antibody manufactured in a CHO cell line in a facility with a 12,000 L bioreactor. For the cell system itself see Chinese hamster ovary cells, and for the vessel see what is a bioreactor.
Scale has grown. The history reports that facilities built in the 1999 to 2019 period had bioreactors ranging from 12 to 25 kL, frequently six to twelve of them feeding two downstream purification trains. Cell line development and production media are credited for the rise in antibody titer, the concentration of antibody in the culture at harvest. The table below gives the representative titers the author lists for full-length IgG1 antibodies.
| Era | Years | Representative titer (g/L) |
|---|---|---|
| 1 | 1975 to 1998 | Less than 1 to 2 |
| 2 | 1999 to 2019 | 2 to 5 |
| 3 | 2020 to 2023 | 5 to 8 |
| 4 | 2024 onward | 8 and above |
The author’s table footnote states that these titers are representative of full-length IgG1 antibodies, so they should not be read as a figure for every antibody format or every product.
Stage 3: capture, polishing and formulation
After culture, the cells are removed by clarification, which the history says is done by centrifugation. The industry then converged on a downstream platform of Protein A capture followed by one or two polishing chromatography steps and virus filtration. Protein A capture is described as having become the dominant capture step in nearly all antibody purification processes. Ultrafiltration and diafiltration then concentrate the antibody and exchange it into the formulation, producing the drug substance. The author notes that the virus filtration and ultrafiltration/diafiltration steps contribute to assurance of the viral safety of the product.
Viral safety is also the subject of a regulatory guideline. ICH Q5A(R2), “Viral Safety Evaluation of Biotechnology Products Derived from Cell Lines of Human or Animal Origin”, was issued as final FDA guidance in January 2024 and describes risk-based principles and mitigation strategies to assure the viral safety of biotechnology products, including the data to include in a marketing application. It updates the 1998 ICH Q5A guidance.
How to produce monoclonal antibodies at research scale versus manufacturing scale
People who search for how to produce monoclonal antibodies are often asking about the bench rather than the plant. The sources read for this page support a scale-neutral description: the discovery stage is the same whatever the end use, and the manufacturing stage described above (CHO fed-batch, Protein A, polishing, virus filtration, UF/DF) is the platform for therapeutic drug substance. Reported commercial scale runs from 12,000 L launch facilities to companies reporting 100 kg drug substance batches, with some of the largest facilities able to produce 150 kg batches. A laboratory producing an antibody for research use does not need that scale, and the sources read here do not give a research-scale procedure, so working laboratories should follow the validated protocol and instructions from their own institution, supplier or core facility rather than adapt the manufacturing description.
Types and related concepts
The data table below compares the four discovery routes on host animal, expression host and glycosylation, library size and stated limits. Related formats and terms are covered elsewhere on the site: single-chain variable fragments (scFv), which are commonly associated with display technology, and the Fc domain, the part of the molecule that differs between full-length antibodies and fragments. Single B cell antibodies are typically expressed as Fab fragments in the cited review, which is one reason a discovery-stage fragment and a manufactured full-length antibody are not interchangeable.
Where it matters for labs, clinics and supply decisions
- Know which stage a product description refers to. “Hybridoma-derived” describes how the clone was generated. “Recombinant” or “CHO-expressed” describes how it is made. A supplier datasheet may state either or both; the two are different questions.
- Animal use is a route-level difference. Hybridoma and transgenic mouse routes require immunized animals, whereas phage display does not require an animal host and single B cell methods can use B cells from samples without euthanizing animals.
- Glycosylation depends on the host. Antibodies made in E. coli are not glycosylated, while eukaryotic hosts such as CHO produce glycosylated antibodies. Where glycosylation matters to function, the expression host is therefore part of the specification.
- Genetic stability. The hybridoma literature flags that hybridoma lines may be genetically unstable.
- Viral safety. For biologic drug substance, virus filtration, UF/DF and the ICH Q5A(R2) framework are part of the manufacturing and regulatory picture, not optional add-ons.
Common misunderstandings
- “Monoclonal antibodies are made in mice.” Mice are used in the hybridoma and transgenic-mouse discovery routes, but phage display needs no animal host and the manufacturing platform for therapeutics is CHO cell culture in bioreactors.
- “Hybridoma technology is obsolete.” The 2021 review that asks “is it still useful?” reports that over 90% of therapeutic antibodies approved were generated by it, mostly as chimeric or humanized versions, although it also lists real limits.
- “Higher titer means a better antibody.” Titer describes how much antibody a culture yields, not its binding or quality. The history reports titers rising across eras because of cell line development and media, not because of changes to the antibody itself.
- “One method fits all.” The routes differ on host, glycosylation, library size and cost, as the data table shows.
Frequently asked questions
How are monoclonal antibodies produced?
A clone is first generated by a discovery route such as hybridoma fusion, phage display, transgenic mice or single B cell cloning. For therapeutic products the antibody is then expressed in fed-batch CHO culture and purified by Protein A capture, polishing chromatography, virus filtration and ultrafiltration/diafiltration, as described in a 2024 history of the field.
How to produce monoclonal antibodies?
The overview above describes the stages and the manufacturing platform. It is not a bench protocol. For a laboratory setting, follow the validated procedure of your institution, core facility or supplier, and see the hybridoma guide for the classical route.
Are monoclonal antibodies made in CHO cells?
The standard manufacturing platform for therapeutic antibodies described in the 2024 history is fed-batch CHO culture in chemically defined media. Earlier antibodies were produced in hybridoma cells grown in the mouse peritoneum.
Do all routes need animals?
No. Hybridoma and transgenic mouse routes involve immunized animals. Phage display does not require an animal host, and single B cell methods can use B cells from human subjects or from samples collected without euthanizing animals.
What is Protein A used for?
Protein A capture is the dominant first purification step in nearly all antibody purification processes described in the 2024 history, followed by one or two polishing chromatography steps.
Sources
- Moraes JZ et al. “Hybridoma technology: is it still useful?” Current Research in Immunology 2021;2:32-40. https://pmc.ncbi.nlm.nih.gov/articles/PMC9040095 (checked 2026-11-14).
- Kelley B. “The history and potential future of monoclonal antibody therapeutics development and manufacturing in four eras.” mAbs 2024;16(1):2373330. https://pmc.ncbi.nlm.nih.gov/articles/PMC11218797 (checked 2026-11-14).
- U.S. FDA. “Q5A(R2) Viral Safety Evaluation of Biotechnology Products Derived from Cell Lines of Human or Animal Origin” (final guidance, January 2024). FDA guidance page (checked 2026-11-14).
Data: Four routes for generating monoclonal antibodies compared
Compares hybridoma, phage display, transgenic mouse and single B cell routes on animal requirement, expression host and glycosylation, library size and stated limits. Compiled by the CASRAI editors from the statements in one peer-reviewed 2021 review of antibody discovery technology; blank values are marked as not stated rather than estimated.
| Route | Animal host needed to generate antibodies | Expression host and glycosylation | Library size | Limits stated in the review |
|---|---|---|---|---|
| Hybridoma | Yes, animals are immunized with the antigen | Hybridoma (eukaryotic) cells; hybridoma-derived antibodies for therapy are converted by chimerization or humanization | Not applicable (clones from fused B cells) | Low efficiency of cell fusion and hybridoma isolation; hybridoma lines may be genetically unstable |
| Phage display | No animal host required | Selected antibodies made in E. coli, so not glycosylated | About 10^10 to 10^11 variants (ribosome display about 10^14) | Library size limited by bacterial transformation efficiency; no glycosylation in E. coli |
| Transgenic mice with human Ig loci | Yes, transgenic mice are immunized | Murine hybridomas secreting human antibodies | Not applicable | Shares hybridoma limits; first approved example was panitumumab (2006) |
| Single B cell cloning | No; B cells from vaccinated or naturally immunized humans, or from animal samples collected without euthanasia | Antibody genes cloned into expression cassettes for cell-based or cell-free expression | Not stated | Single-cell sorting devices are expensive; cell-free expression can give incorrect folding of antibody domains |
Sources
- Moraes JZ et al. Hybridoma technology: is it still useful? Curr Res Immunol 2021 (accessed 2026-11-14)
Compiled by CASRAI from the sources above. Free to reuse under CC BY 4.0 with attribution.







