Antigenicity is the ability of a molecule to be bound by an antibody. A peptide is antigenic when an antibody can bind it, and the patch it binds is the epitope. Short peptides can carry an epitope, but they are too small to raise antibodies on their own. So labs couple them to a large carrier such as KLH.
- An epitope is the region of an antigen that an antibody contacts; the antibody's matching region is the paratope.
- Linear epitopes are neighboring residues; conformational epitopes are distant residues brought together by folding.
- Short peptides usually behave as haptens and are coupled to a carrier protein to raise antibodies.
- KLH is a very large carrier, reported at about 450,000 to 13,000,000 daltons, rich in lysines for coupling.
- Peptides of about 10 to 20 residues are commonly described as workable immunogens.
- Testing on a different carrier than the one used to raise the antibody rules out anti-carrier binding.
What an antibody actually recognizes
What does an antibody see when it meets a protein? It does not read the whole molecule. It touches one small patch of the surface, a few residues across, and either fits that patch or does not.
That patch is the epitope. The molecule carrying it is the antigen, and the matching region on the antibody is the paratope. A large protein can present many epitopes at once, each recognized by a different antibody.
Antigenicity is the ability of a molecule to be recognized and bound by an antibody. It is a property of the fit between an epitope and a paratope. A peptide is antigenic, in this sense, when it carries a stretch of residues an antibody can bind.
This page covers peptides as antigens in laboratory antibody work: how epitopes are described, why short peptides need a carrier, and how binding is checked. It makes no claims about any compound's effect in a living system.
- Pick a surface-exposed stretch of 10 to 20 residues
- Confirm the peptide's identity and purity
- Couple it to a large carrier such as KLH
- Raise the antibody against the conjugate
- Test binding on peptide-BSA and the native protein
What are linear and conformational epitopes?
A linear epitope is a run of neighboring residues, while a conformational epitope is built from residues brought together by folding. Epitopes come in two broad kinds, and the difference decides whether a short peptide can stand in for a protein at all.
A linear epitope, also called a continuous epitope, is made of residues that sit next to each other in the sequence. An antibody that binds a linear epitope can often bind a short synthetic peptide copying that stretch, because the residues it recognizes are all present in one piece.
A conformational epitope, also called discontinuous, is made of residues that are far apart in the sequence but brought together when the protein folds. Cut the protein into pieces and the epitope falls apart. A short peptide copying one of those pieces carries only part of the surface the antibody recognized.
So a synthetic peptide is best suited to linear epitopes. For conformational ones, the protein's folded shape is part of the antigen, and sequence-based prediction tools have to be supplemented by structure-based methods.
Why is a short peptide treated as a hapten?
A short peptide is treated as a hapten because an antibody can bind it, but it is usually too small to raise one alone. Recognition and response are two different things. An antibody that already exists may bind a short peptide well. Getting a host to make that antibody in the first place is harder, because a small peptide on its own is usually too small and too simple to trigger a useful antibody response.
Immunologists call a molecule like that a hapten: something that can be recognized once an antibody exists, but that does not raise antibodies by itself. The authors of a 2020 study producing antibodies against casein peptides put it plainly, describing their peptides as haptens that needed conjugation to a carrier protein to induce a sufficient response.
The practical answer is to attach the peptide to something large and foreign. The carrier supplies the size and complexity the immune system responds to, and the peptide rides along as the epitope of interest.
Carrier proteins
Keyhole limpet hemocyanin, shortened to KLH, is the carrier most often named in antibody work. It is a very large protein, reported at roughly 450,000 to 13,000,000 daltons depending on its assembly, and it comes from a sea snail that is evolutionarily distant from mammals. It also carries many lysine residues, which give the coupling chemistry plenty of places to attach a peptide.
Bovine serum albumin, BSA, is the other common carrier. In the casein study it served a different role: the immunizing conjugate was peptide-KLH, while peptide-BSA was used to coat plates for testing. Switching carriers between the two steps is a deliberate choice.
The reason is that antibodies form against the carrier as well as the peptide. If the same carrier appeared in both steps, a positive test result could come from anti-carrier antibodies. Testing against the peptide on a different carrier shows the binding is to the peptide.
Choosing a peptide antigen
When the goal is an antibody that recognizes a native protein, the peptide is chosen from that protein's sequence with some care. Three considerations come up again and again in published work.
| Factor | Why it matters |
|---|---|
| Length | Peptides of about 10 to 20 residues are commonly described as workable immunogens once coupled to a carrier. |
| Hydrophilicity | Water-loving stretches tend to sit on a protein's surface, where an antibody can reach them. In the casein study, the chosen N-terminal peptide was 44 percent hydrophilic residues against 33 percent for the C-terminal one. |
| Predicted epitope score | Prediction servers such as BepiPred score each residue for its likelihood of sitting in a B-cell epitope. |
Which residues count as hydrophilic depends on their side chains. Our guides to the R group and polar amino acids cover that chemistry.
Finding epitopes: experiment and prediction
How does anyone know where the epitopes on a protein are? The direct answer comes from experiments. A crystal structure or a cryo-electron microscopy structure of an antibody bound to its antigen shows exactly which residues touch. Peptide assays take the opposite route, testing many short overlapping peptides from a sequence and seeing which ones an antibody binds. Both approaches are slow and expensive.
Prediction fills the gap. The Immune Epitope Database, a curated public collection, held roughly 623,000 entries according to a 2026 review, and prediction tools are trained on records like these. Sequence-based tools look for likely linear epitopes; structure-based tools try to find conformational ones.
The same review is candid about the limits. Training data lean heavily toward viral proteins, high-resolution antibody-antigen structures are scarce, and conformational epitopes remain hard to predict. A predicted epitope is a starting point for an experiment, not a result.
How do you check that an antibody binds the peptide?
The usual check is an enzyme-linked immunosorbent assay, ELISA. In the indirect form, the antigen is fixed to a plate, the antibody sample is added, and a second antibody carrying an enzyme reports how much of the first one stuck. In the casein study that second antibody was peroxidase-labeled, the substrate was TMB, and the color was read at 450 nanometers.
Before testing, the antibodies were purified from the serum on protein G, which grips the constant Fc region shared by antibodies of that class. That step separates antibodies from everything else in the serum, though it does not by itself pick out the ones specific to the peptide.
A clean result shows binding to the peptide, ideally presented on a carrier the antibody never saw, and binding to the native protein the peptide came from. The casein study reported both: its purified antibodies bound the peptide-BSA conjugate and native goat alpha-s1 casein more strongly than the unpurified serum did.
Specificity is a separate check. The same study ran a western blot against milk from cow, buffalo, horse and camel, and saw a band only for goat. Binding to the peptide but not the protein usually points back to the epitope question: the stretch may be buried in the folded protein, or part of a conformational epitope a linear peptide cannot reproduce.
Why the peptide's identity matters here
An antibody raised against a peptide is only as specific as the peptide was. If the synthetic material contained deletion sequences or modified residues, some of the antibodies raised may recognize those instead of the intended stretch.
That is why antigen peptides are checked the same way as any research peptide: identity by mass spectrometry and purity by HPLC, reported on a certificate for the lot used. Our guide to LC-MS identity testing explains how the mass result is read.
Purity figures in the published work show why this matters. The two casein peptides in the 2020 study were reported at about 95 and 85 percent purity. A peptide at 85 percent carries a meaningful share of other species into the conjugate, and every one of them is a potential target for the antibodies that come back.
FOR LABORATORY AND IN-VITRO RESEARCH USE ONLY. NOT FOR HUMAN OR ANIMAL CONSUMPTION.
Common questions
What is the difference between an antigen and an epitope?
Why do peptides need a carrier protein?
Why use a different carrier for testing than for raising the antibody?
Can a short peptide mimic any epitope on a protein?
More science guides
Sources
- Mohsin et al., Generation of High Affinity Anti-Peptide Polyclonal Antibodies Recognizing Goat alpha-s1-Casein, Molecules 25(11), 2020. Peptides treated as haptens needing a carrier, KLH size and lysine content, BSA conjugates for ELISA coating, the 10 to 20 residue range, hydrophilicity figures and BepiPred scoring.
- B-cell epitope prediction in the age of machine learning: advancements and challenges, Journal of Translational Medicine, 2026. Definitions of B-cell epitopes, linear versus conformational epitopes, paratopes, and the role of structure-based methods for conformational epitopes.
