Antibody Production: The Actual Process

Antibodies are produced by B cells — specifically plasma cells, which are the terminally differentiated effector form of B lymphocytes. When an antigen enters the body, B cells that recognize it get activated, usually with help from CD4+ T helper cells. They undergo clonal expansion, somatic hypermutation in the germinal center, affinity maturation, and eventually differentiate into either memory B cells or antibody-secreting plasma cells. That is the biological reality. But if you are actually working in a lab and need to produce antibodies for research or therapeutic use, the picture gets much messier and a lot less elegant.

How Antibodies Are Produced By (and What Nobody Tells You)

The standard pathway people learn in undergrad involves immunizing an animal, collecting serum, and purifying the immunoglobulins. Polyclonal antibodies come from that process — a mixture of billions of clones, each recognizing different epitopes on the antigen. Monoclonal antibodies require hybridoma technology, which is essentially mouse B cells fused with myeloma cells, then screened, cloned, and expanded. The classic Köhler and Milstein method from 1975 still forms the backbone of everything, even though the field has moved on significantly. Here is the thing that trips up people who are new to this: the immunogen itself determines everything downstream. If your antigen is a pure, well-folded protein with a clean epitope, great. Most antibodies aren't produced against clean proteins in real life. They are produced against peptide conjugates, denatured lysates, or recombinant fragments that may not present the native conformation. I spent three months trying to get a usable monoclonal antibody against a phosphorylated epitope on a intrinsically disordered region of a signaling protein. The protein refused to fold consistently, the phospho-modification was labile during purification, and every clone we generated either didn't bind at all or cross-reacted with the unphosphorylated form. We ended up switching to a synthetic phospho-peptide conjugated to keyhole limpet hemocyanin with a spacer arm that kept the phosphate group exposed. That changed everything. The clone we isolated from that immunization worked in Western blot, IP, and flow cytometry. Took another four months to stabilize the clone and scale it up, but at least it bound what we needed it to bind. The hybridoma process itself has several failure points. Fusion efficiency varies wildly between labs and even between batches. My experience is that PEG-mediated fusion of spleen B cells to SP2/0 myeloma cells gives roughly 1 in 10^6 to 10^7 viable hybrid cells per fusion attempt. You need to plate generously — 96-well plates, at least 24 per fusion, ideally more. The HAT selection medium kills unfused myeloma cells but also kills B cells naturally after about 1–2 weeks. Hybridomas survive because they inherit the B cell's salvage pathway enzymes and the myeloma's infinite proliferation capacity. Screening comes next. ELISA is the workhorse, but it only tells you about binding to your coated antigen. You also need to test cross-reactivity, isotype, and stability. I always run a quick Western blot in parallel because an antibody that works in ELISA but not under denaturing conditions is useless for half the applications people buy it for.

The Modern Alternatives to Traditional Hybridoma Production

Single B cell sorting and scFv/Fab cloning from vaccinated humans or animals has largely replaced hybridoma technology for many applications. You isolate individual B cells using FACS, reverse transcribe the mRNA, amplify the heavy and light chain variable regions by PCR, and clone them into expression vectors. This is faster, avoids the need for cell line stabilization, and gives you the exact sequence — which matters if you ever need to regenerate the antibody or engineer it. The downside is that you get scFvs and Fabs, not full-length IgGs, unless you go through an extra step of reconstituting the constant regions. And screening throughput is lower because you are working with individual clones rather than bulk culture supernatant. Phage display libraries are another route. You build a library from naive or immunized B cell repertoires, display the Fab fragments on bacteriophage, and capture binders against your antigen through biopanning. This is entirely in vitro and can yield high-affinity binders in 2–3 rounds of selection, which is roughly 2–3 weeks from library construction to positive clone. The catch is that phage display biases toward certain epitope classes and missing glycosylation patterns can distort the conformational landscape your displayed fragments see. I ran a phage display campaign against a GPCR extracellular domain and got eight strong hits. None of them recognized the receptor on live cells. The protein was refolded from inclusion bodies and the ligand-binding pocket was partially collapsed. Switching to a detergent-solubilized membrane prep for panning rescued two clones that worked in flow cytometry. It cost another month and a bunch of detergent optimization, but it was faster than going back to immunization. For therapeutic antibody development, the game has shifted further. Transgenic mice like VelocImmune produce fully human antibodies natively, eliminating the immunogenicity problem that plagued early therapeutics. Human B cell cloning from convalescent donors has become routine after viral outbreaks — the SARS-CoV-2 response saw hundreds of therapeutic antibodies cloned directly from patient samples within weeks of sequencing the virus. The bottleneck there isn't production anymore; it's clinical development and manufacturing scale-up.

Practical Considerations That Determine Whether Your Antibody Will Actually Work

Antibody affinity maturation in vivo takes weeks to months. If you are producing antibodies for research use and need them fast, you might compromise on affinity and just use whatever comes out of a standard 8-week immunization protocol. Those antibodies work fine at 1–5 µg/mL in Western blot but fall apart in immunoprecipitation or flow cytometry where you need sub-nanomolar affinity. I have seen people spend more time troubleshooting failed IPs than they would have spent optimizing the initial immunization. A booster schedule that includes increasing antigen doses and longer intervals between boosts generally improves affinity more than simply adding more boosts. Isotype matters more than people admit. IgG2a and IgG2b in mice tend to have better Fc-mediated functions — complement activation, ADCC, and Fc receptor binding — than IgG1. If you are producing antibodies for in vivo work or pull-down assays, the isotype can be the difference between a successful experiment and a wasted week. I always check isotype by ELISA before committing to hybridoma cloning or B cell sorting. Purification strategy depends entirely on your end use. Protein A/G affinity chromatography gives you pure IgG in one step, but it co-purifies albumin and other serum proteins if you are working with polyclonals. For monoclonals, ion exchange and size exclusion are standard polishing steps. If you need antibody fragments for imaging or intracellular use, you will need enzymatic digestion with papain or pepsin, followed by additional purification. Papain cleavage above the hinge region produces two Fab fragments and one Fc fragment. Pepsin cleavage below the hinge produces F(ab')2 and small pFc' debris. The choice affects your binding valency and your ability to cross-link antigens. I learned this the hard way when I used a pepsin-generated fragment for a co-IP and got nonspecific pull-downs because the bivalent F(ab')2 was bridging multiple proteins in the lysate. Switching to a papain-derived Fab solved it because monovalent binding doesn't cross-link.

Antibodies Are Produced By Plasma Cells But That Is Only the Beginning

The biological mechanism is straightforward. The practical execution is full of decision points that compound. Antigen design, adjuvant choice, immunization schedule, isotype screening, cloning strategy, expression system, purification method, and application validation all interact. A mistake early in the pipeline is expensive to fix later. I recommend investing real time in antigen characterization and immunogen design before you ever touch a mouse or start a B cell sort. A well-defined antigen saves more headaches than any optimization you can do downstream. The field also has a reproducibility problem that isn't widely discussed. Many commercial antibodies are cloned from the same few hybridoma lines and redistributed under different catalog numbers. lot-to-lot variation in polyclonal preparations is enormous because each immunization produces a different repertoire. If you publish with an antibody and someone else cannot replicate your results, the problem is often the antibody, not the biology. Always validate with knockout controls, peptide pre-absorption, or independent antibodies targeting different epitopes. It adds cost and time but it is the only way to be confident your signal is real. For most people producing antibodies for the first time, the hybridoma route with a well-optimized immunization remains the most reliable path to a full-length IgG. For speed and sequence certainty, single B cell cloning is harder to beat. Phage display is worth considering when your antigen is difficult to produce in native form or when you need to screen thousands of potential binders rapidly. Each approach has trade-offs and none of them guarantee success. The common factor across all of them is that the quality of your antigen and the clarity of your application determine more about the outcome than any of the technical steps themselves.