Are these excluded thru central tolerance? What if you ingested something with a unique molecular structure that you hadn't ingested before?

  • $\begingroup$ Maybe my answer to your other question answers somehow this one? $\endgroup$
    – ddiez
    Commented Jan 19, 2016 at 5:35

2 Answers 2


B cells containing receptors against food's or drug's molecular structures may exists. However if the naive B cell binds the molecule without a second signal coming simultaneously from CD4 Tfh cells (linked recognition) the B cells recognizing this antigen will become anergic or will die by apoptosis. This occurs with antigens to which we are chronically exposed, like food.

Reference: Janeway's Immunobiology, 8th edition. Chapter 8 (The development and survival of lymphocytes). In particular section 8-26, p320-321.

  • 1
    $\begingroup$ I would add that many food allergy structures are recognized in hypersensitivity reactions because they oddly enough resemble helminthic peptide antigens. Furthermore, lipid/CHO-type antigens which aren't peptide can't be handled in the same way as peptide antigens, hence we have gamma-delta T cells and surface receptors like TLR-4 that have to do this job. Since antibodies get generated generally due to peptide antigen presentation by APCs, there's a lack of robust antibody production to non-peptide antigens. $\endgroup$
    – CKM
    Commented Jan 20, 2016 at 15:18

I’m not an immunologist, but I think it fair to say that antibodies may well be able to bind to certain food or drugs. However there are several reasons that this may not cause their removal or trigger an immune response. These include:

  • Location. If the food is in the stomach I imagine that the digestive environment favours breakdown to sugars, amino acids etc and exposure to antibodies is minimal.
  • Chemistry. Certain substances such as fat, nucleic acids, are weakly antigenic.
  • Size. Small molecules are not immunogenic unless attached to a larger one.

Why are small molecules not immunogenic?

I think this is probably the key point here, so I'll expand on it a little in general terms.

If an antibody binds a foreign protein, this is large enough that other antibody molecules can bind it too. And the antibody is bivalent, so it may bind a second foreign protein. This leads to a cross-linked matrix that can be recognized by macrophages and engulfed. If a small molecule binds to an antibody this cannot happen and the molecule may well diffuse off again.

As well as binding those soluble antibodies (IgG) already present, antigens cause the antibody-producing cells (B-cells) to divide to produce more of the particular antibody. This happens because the B-cells have a membrane bound variant (IgM) with the same specificity on their surface. When a molecule binds to two of these membrane antibodies a signal is triggered to the cell. This can happen with a large molecule, but a small molecule can bind only one IgM on the surface and this expansion of antibody-producing cells will not happen.

This size requirement for immunogenicity can be understood in the context of the size of the bacteria and viruses which one presumes the immune system evolved to combat.

More Information

As I mentioned, I deliberately answering the question in general terms. If you wish to know more detail (and about cellular immunity) the following on-line book chapters (although a little out of date) may be of interest.

  • $\begingroup$ hypersensitivity (allergy) is an autoimmune response, and it is caused by ingestion (peanuts) or inhalation (pollen) $\endgroup$ Commented Aug 17, 2016 at 23:23
  • $\begingroup$ @FoldedChromatin — You should critcize the question, not my answer, as you are criticizing its premise. And perhaps you would care to elaborate and explain how this occurs in molecular terms and why most people do not suffer from such allergies. As I said, I'm no immunologist, but your comments have contributed nothing to my understanding. If, as I assume, you have down-voted my question, you could point out to everyone which particular parts of my answer are incorrect. $\endgroup$
    – David
    Commented Aug 18, 2016 at 11:28

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .