Article Text


HLA-DQ8 as an Ir gene in coeliac disease
  1. K E A Lundin
  1. Department of Medicine and Institute of Immunology, Rikshospitalet, N-0027 Oslo, Norway; knut.lundin{at}


DQ8 restricted gliadin peptide is immunogenic in the intestinal mucosa of HLA-DQ8 positive patients, representing the first demonstration that a given peptide may be of pathogenic significance only for a subset of coeliacs, and strongly suggests that DQ2 and DQ8 act as immune response (Ir) genes in this disease

  • coeliac disease
  • gliadin peptides
  • HLA restriction
  • organ culture
  • mucosal immunity
  • Ir genes coeliac

Statistics from

Coeliac disease (CD) is an excellent model for dissecting human immunological disease. It is one of the few human non-infectious diseases where the triggering agent can be removed and thereby a causal treatment exists—the gluten free diet. Removal of the harmful agent is followed by complete normalisation of the small intestinal mucosa. The subsequent introduction in vivo and in vitro of the disease process offers a unique possibility to dissect this process. Clearly, the in vivo short or long term challenge most directly addresses the range of possible pathological processes that eventually lead to the overt enteropathy. However, the in vitro challenge of the coeliac mucosa stands as an excellent surrogate system.

In this issue of Gut, Mazzarella and colleagues,1 representing some of the leading groups in coeliac disease, present work on the coeliac mucosa in an in vitro challenge system [see page 57]. Such a challenge can be done on the damaged mucosa from untreated patients (readout: lack of normalisation on withdrawal of gluten) or on a normalised mucosa from treated patients (readout: induction of pathology on introduction of gluten). The group chose the latter and investigated several aspects of immunological activation in biopsies following a 24 hour challenge. One of the few critical comments of this excellent study is that a bucket full of synthetic peptide was used for the challenge. As a putative immunodominant epitope, a concentration of 1 mg/ml is certainly very high and determination of a dose-response could clearly have been included.

The first important observation provided in this paper is that a peptide identified as an epitope for small intestinal T cells can also induce a mucosal response, strongly suggesting that the epitope is of immunobiological and immunopathological importance. The outcome of the challenge is identical regardless of the peptide being used in its native form or in its deamidated form. The process of deamidation results in conversion of glutamines in the primary gliadin sequence to glutamic acid, leading to better binding of the peptides to HLA and more efficient recognition by gliadin specific T cells. 2 The present findings corroborate other data suggesting that the enzyme tissue transglutaminase deamidates gliadin in situ.3 The study of Mazzarella et al is an important step forward, although regrettably it did not address in a dose-response assay if the already deamidated peptide was more potent than the native peptide. Similar experiments involving any of the other gliadin T cell epitopes have not been reported. As a natural extension of these studies, one could envisage that all newly described T cell epitopes should be studied in similar experiments. This would be a daunting task but certainly much more amendable than complete in vivo toxicity testing.4 Such a combination of epitope identification using T cell clones and epitope confirmation using in vitro biopsy challenge will generate a valuable map of the gliadin fragments involved in disease pathogenesis.

The second important observation the authors make sheds light on a longstanding battle in progress within the inner circle of scientists and clinicians working with CD. Two mutually exclusive pathways have been suggested to contribute to the disease, as is clearly discussed in the paper. One is thought to be a rapid effect on the epithelium and intraepithelial T cells that involves several aspects of the innate immune system. The other is a delayed-type hypersensitivity-like reaction involving CD4+ T cells that recognise processed gliadin peptides presented by HLA-DQ2 and HLA-DQ8. By and large, this pathway represents the adaptive immune system. At the recent 10th International Symposium on Coeliac Disease in Paris, the present status of this battle was summarised by one speaker as: “It is the epithelium, after all!” However, good evidence that these two pathways are in reality distinct from one another is lacking. The present paper suggests that even an innate-type immune response such as induction of Fas on enterocytes is controlled by a distinct interaction between gliadin peptides, lamina propria, HLA-DQ8, and HLA-DQ8 restricted T cells. In fact, no signs of innate immune system activation were observed in DQ8 negative patients. Maybe “it is the adaptive immune system, after all?”

The third important aspect to this manuscript should arouse the immunological society. We know that whereas the vast majority of coeliacs carry HLA-DQ2, the rest carry DQ8. Individuals without DQ2 or DQ8 are extremely unlikely to develop CD. One important aspect that must be remembered is that most individuals with DQ2 and DQ8 will not go on to develop CD. However, the HLA association per se strongly suggests that DQ2 and DQ8 act as immune response (Ir) genes for the mucosal immune response. The term Ir genes was defined when immunology as a science was in its infancy.5 Later it was established that the Ir gene phenomenon was due to specific binding and presentation by antigenic peptides by the major histocompatibility complex encoded molecules.6 Moreover, demonstration that the Ir gene phenomenon is operable in humans has been very difficult. Mazzarella et al demonstrated that a DQ8 restricted peptide acts pathogenically only in DQ8 positive and not in DQ8 negative coeliacs. This represents the first demonstration that a given peptide may be of pathogenic significance only for a subset of coeliacs, and strongly suggests that DQ2 and DQ8 act as Ir genes in the disease.

At the end of the day, our efforts will not be measured by how much we understand of CD but to what extent we can translate our knowledge for the benefit of patients. Through basic medical science we have learned a lot about the disease over the past 10 years. The present paper represents both good news and not so good news. On the positive side, our understanding of the disease has made a significant step forward. On the negative side, this paper clearly illustrates that any attempt to design methods of specifically intervening with the disease process must take the HLA type of the patient into account. This confirms the fears that this difficult task has just become more complex.


View Abstract

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Linked Articles