← Back to the biomarker database

Transglutaminase-3-Antikörper (anti-TG3): the pathway in the body

Transglutaminase-3-Antikörper (anti-TG3) is part of the pathway “Gluten antibodies”. This page shows the whole pathway; the station of Transglutaminase-3-Antikörper (anti-TG3) is highlighted.

Where this laboratory value sits: Anti-TG3 and anti-TG6 — antibodies against relatives. TG3 is found mainly in the epidermis, TG6 in nerve cells; both are related to transglutaminase 2. In some people with an immune response to gluten, antibodies against these enzymes form as well. Source 5, 6

In brief

Gluten is a mixture of storage proteins from wheat, rye and barley. Some of its fragments are modified by transglutaminase 2 and can start an immune response with antibodies against gliadin and against this enzyme. Once modified, they bind more firmly to HLA-DQ2/DQ8.

11 stations · 6 sources
ORYIn the gutImmune responseTransglutaminase 2CalciumDigestive enzymesHLA-DQ2, HLA-DQ8into the gut liningGluten in foodWheat, rye, barleyLong fragmentsleft undigested33-mer peptideespecially resistantPassage into tissuebeneath the gut surfaceFragment in the tissuemeets an enzymeModified peptidenegatively chargedBinding to HLA-DQ2/DQ8on immune cellsT cells recognise itResponse beginsB cells are activatedtwo targetsAntibodies in the bloodanti-TG2 and AGAAnti-TG3 and anti-TG6antibodies against relatives

Swipe the graphic sideways

The pathway step by step

Each station states what the compound does there. Three signs: ↑ supplies — builds up or makes available · ↓ depletes — inhibits, consumes or withholds · ↕ both, depending on amount. Behind it stands what the statement rests on: established physiology, observed in studies, or contested. The signs do not grade; they name the direction.

  1. Gluten in food → Long fragments Digestive enzymes Stomach and gut enzymes break most proteins down to single amino acids. At proline-rich sites they do not cut, so longer fragments remain. They are long enough to be displayed to T cells. Source 1↑ supplies Because stomach and gut enzymes do not cut at proline-rich sites, fragments stay long enough to be displayed on HLA molecules later and recognised by T cells. established physiology Source 1, 2
    ⚖ When the balance tips

    too much — If many long fragments remain, more material is present in the gut contents that can reach the gut wall.

    too little — If the chains are broken down extensively, hardly any pieces form that are long enough to bind to HLA molecules.

    established physiology · Source 1, 2

  2. Long fragments → 33-mer peptide A fragment of 33 amino acids survives digestion almost unchanged. It is the best studied of these peptides. It contains several sections that T cells recognise. Source 1↑ supplies The 33-mer peptide contains several overlapping sections that T cells recognise and is a preferred target of transglutaminase 2. In experiments it strongly stimulated T cells from the gut lining. observed in studies Source 1
    ⚖ When the balance tips

    too much — If much of it stays intact, many recognisable sections lie within one piece; in experiments it was precisely this long form that strongly stimulated T cells.

    too little — If it is broken into smaller pieces after all, the recognisable sections fall apart; in experiments the peptide then lost its effect on T cells.

    observed in studies · Source 1

  3. 33-mer peptide → Passage into tissue A small share of these fragments passes through or between the gut cells into the lining beneath. There they meet the enzyme and immune cells. Source 2↑ supplies Only beneath the gut surface do the fragments meet transglutaminase 2 and immune cells. The passage thus links the gut contents with the immune defence in the lining. observed in studies Source 2
    ⚖ When the balance tips

    too much — If more fragments reach the tissue, more material is available there for the enzyme and the immune cells.

    too little — If hardly any fragments pass through, the pathway stops in the gut contents, and the immune cells of the lining receive little of them.

    observed in studies · Source 2

  4. Fragment in the tissue → Modified peptide Transglutaminase 2 · Calcium The enzyme converts individual glutamine units into glutamic acid. This process is called deamidation; afterwards the fragment carries negative charges. This makes it bind more firmly to HLA-DQ2 and DQ8. Source 3↑ supplies The negative charges make the fragment bind more firmly to HLA-DQ2 and HLA-DQ8. Only then does it become recognisable to many of the T cells that respond to gluten. established physiology Source 2, 3
    ⚖ When the balance tips

    too much — If many glutamine units are converted, more firmly binding fragments are present, and more HLA molecules carry them on the cell surface.

    too little — If little is converted, the fragments bind more weakly to DQ2 and DQ8 and are displayed to T cells less often.

    established physiology · Source 2, 3

  5. Modified peptide → Binding to HLA-DQ2/DQ8 · HLA-DQ2, HLA-DQ8 Immune cells display fragments on HLA molecules. Negatively charged fragments fit especially well into the binding groove of DQ2 and DQ8. Without DQ2 or DQ8 they barely find a hold. Source 2↑ supplies The HLA molecules hold the fragment on the cell surface and display it to T cells. Without DQ2 or DQ8 it barely binds, and the pathway breaks off at this point. established physiology Source 2
    ⚖ When the balance tips

    too much — If cells carry many DQ2 or DQ8 molecules with a bound fragment, the matching T cells meet their target more often.

    too little — If DQ2 and DQ8 are absent, the fragments hardly find a fitting binding groove, and little of them is displayed to T cells.

    established physiology · Source 2

  6. Binding to HLA-DQ2/DQ8 → T cells recognise it Certain T cells recognise exactly this combination of HLA molecule and modified fragment. As a result they become active and divide. They then help B cells to produce antibodies. Source 2↑ supplies The active T cells release messenger substances into the lining and help B cells that recognise the same fragment to produce antibodies. established physiology Source 2
    ⚖ When the balance tips

    too much — If many of these T cells become active, they release more messenger substances into the lining, and more B cells receive their help.

    too little — If hardly any T cells become active, the help signal for the B cells does not come, and few antibodies against gliadin and transglutaminase 2 form.

    established physiology · Source 2, 3

  7. T cells recognise it → B cells are activated The T cells pass signals on to B cells. B cells that recognise the fragment itself become active, as do those that recognise the enzyme next to it. B cells against the enzyme internalise it along with the fragment. Source 3, 4↑ supplies B cells that recognise the enzyme internalise it together with the attached fragment. In this way they receive help from T cells that recognise gluten, even though the enzyme is the body's own. observed in studies Source 2, 3
    ⚖ When the balance tips

    too much — If many complexes of enzyme and fragment form, more B cells against the enzyme can receive this help.

    too little — If no more gluten fragments arrive, the B cells against the enzyme no longer receive help, and their response fades over time.

    observed in studies · Source 2, 3, 4

  8. B cells are activated → Antibodies in the blood The B cells produce antibodies: against gliadin (AGA) and against the body's own transglutaminase 2 (anti-TG2). Both can be measured in the blood. Anti-TG2 also binds to the enzyme in the gut lining. Source 3, 4↓ depletes Antibodies against transglutaminase 2 also attach to the enzyme in the gut lining and occupy it. In cell culture they altered its activity; whether they have an effect in the body itself is open. contested Source 3
    ⚖ When the balance tips

    too much — If many antibodies form, more transglutaminase 2 in the tissue is occupied by them; studies describe inconsistently what this means for the enzyme's work.

    too little — If few antibodies form, the enzyme in the lining stays largely free, and hardly any antibodies against gliadin and transglutaminase 2 are found in the blood.

    contested · Source 3

  9. Antibodies in the blood → Anti-TG3 and anti-TG6 TG3 is found mainly in the epidermis, TG6 in nerve cells; both are related to transglutaminase 2. In some people with an immune response to gluten, antibodies against these enzymes form as well. Source 5, 6↓ depletes The antibodies bind to TG3 in the skin or to TG6 in nerve tissue. In skin samples, TG3 and IgA were found deposited together; whether binding inhibits the enzymes in their work is open. observed in studies Source 5, 6
    ⚖ When the balance tips

    too much — If many of these antibodies form, studies find more complexes of antibody and enzyme deposited in tissue, for example in the upper layers of the skin.

    too little — If none of these antibodies form, TG3 and TG6 in skin and nerve tissue stay unoccupied; the response against transglutaminase 2 can still be present.

    observed in studies · Source 5, 6

Further stations

Cofactors in this pathway

Sources

  1. Shan L, Molberg O, Parrot I et al. Structural basis for gluten intolerance in celiac sprue. Science 2002 · PubMed 12351792
  2. Sollid LM, Jabri B. Triggers and drivers of autoimmunity: lessons from coeliac disease. Nat Rev Immunol 2013 · PubMed 23493116
  3. Rauhavirta T, Hietikko M, Salmi T et al. Transglutaminase 2 and Transglutaminase 2 Autoantibodies in Celiac Disease: a Review. Clin Rev Allergy Immunol 2019 · PubMed 27263022
  4. Lindfors K, Ciacci C, Kurppa K et al. Coeliac disease. Nat Rev Dis Primers 2019 · PubMed 30631077
  5. Sárdy M, Kárpáti S, Merkl B et al. Epidermal transglutaminase (TGase 3) is the autoantigen of dermatitis herpetiformis. J Exp Med 2002 · PubMed 11901200
  6. Hadjivassiliou M, Aeschlimann P, Strigun A et al. Autoantibodies in gluten ataxia recognize a novel neuronal transglutaminase. Ann Neurol 2008 · PubMed 18825674

Whole pathway: Gluten antibodies

Related pathways

As of 2026-09-16. Draft written by Claude to schema v2; sources checked in PubMed; expert review pending
Legal notice Privacy policy All biomarkers