• coeliac disease
• gluten
• gliadin peptides
• HLA restriction
• mucosal immunity
• T cells
• immunomodulation

Despite its heterogeneous clinical appearance, coeliac disease has a remarkable uniform human leucocyte antigen (HLA) association. The majority of coeliac disease patients carry a certain variant of HLA-DQ2, the rest carry HLA-DQ8.^1,2 As the physiological role of the HLA system is to present peptide fragments of antigens to T cells, it would seem logical that the HLA-DQ2 and HLA-DQ8 molecules predispose to coeliac disease by presenting peptides to T cells in the intestinal mucosa. The peptides these T cells recognise could derive form gluten peptides as coeliac disease is precipitated by intestinal exposure to wheat gluten and related proteins of other cereals. In fact, this simple concept has gained substantial experimental evidence in recent years.^1,2 A paper in this issue of Gut³ adds further credence to this model [see page 212].

The small intestinal lesion is characterised by inflammation and infiltration of T cells both in the lamina propria and epithelium. Intraepithelial T cells are typical of coeliac disease but their significance is uncertain and they are unlikely to recognise gluten presented by HLA-DQ2. They may represent an innate (“primitive”) immune response to gluten. Their lamina propria counterparts however are much more likely to be key players. In the lamina propria, we find CD4⁺ T cells and dendritic cell type of antigen presenting cells expressing HLA-DQ2. Therefore, when we first isolated gluten reactive T cells some 10 years ago from the small intestinal mucosa of coeliac disease patients, some of us hoped that an immediate solution to several conceptual problems was at hand.⁴ It turned out to be a bit more complex.

Following years of work from a small handful of devoted laboratories, we understand several details with respect to these T cells and what they recognise. We know of several of the peptide epitopes involved in the disease. T cell recognition of many, but not all, is dependent on the action of the small intestinal enzyme tissue transglutaminase which converts glutamine residues in gluten to glutamic acid.^1,2

It is in this field that the research group of Professor Ciclitira at St Thomas' Hospital has spent some years of work. It is therefore reassuring that they now report results that are in keeping with previous observations.³ Four adult coeliac disease patients were examined and small intestinal biopsies were challenged ex vivo with gluten. Thereafter, a cell suspension was made and T cells expanded. After some weeks of cell culture work they obtained enough T cells to perform functional testing. It can be read from the paper that this job was not easy and that several obstacles were encountered. They established both polyclonal and monoclonal T cell lines and found that in most cases T cells were reactive to an already characterised epitope of an α-gliadin with which the intestinal T cells of most adult coeliac disease patients are reactive.⁵ The present data thus fit nicely with these previous observations.

No new epitopes are characterised in this paper. However, Ellis et al extend the focus of the investigations to a field that may be unknown to some readers of Gut—that is, the field of “the altered peptide ligands”. As stated previously, T cells recognise a peptide in the cleft of the HLA molecules. If certain amino acids in the peptide are changed, the result might be that the peptide cannot bind to the HLA molecule and therefore gives no stimulation. Another situation exists if an amino acid that points to the T cell is changed. In this case the T cell might still respond, or it may not respond at all, or it might respond qualitatively different. Such different responses can involve skewing of the cytokine secretion of T cells, induction of longstanding anergy (they will not respond to subsequent antigen challenge), or suppression of bystanders (T cells recognising other parts of the same antigen). The phenomenon is well described in rodents.⁶ Unfortunately, the situation is more complex when applied to human autoimmune disease and to this end the results have been unimpressive.⁷ The results presented by Ellis et al demonstrate that substituting amino acids within the core region of a T cell epitope often impairs T cell recognition, as expected. Although no convincing data are presented as to whether altered gluten peptide ligands can induce qualitatively different T cell responses (for example, T cell anergy or bystander suppression), this remains a possibility that ultimately may be extended to therapy. The fact that there exists multiple gluten T cell epitopes in coeliac disease¹ and that the T cell receptor usage by gluten specific T cells is diverse⁴ however makes this approach less attractive.

Other more promising alternatives for new treatment modalities include inhibition of T cell activation by compounds that block peptide binding to HLA-DQ2, inhibitors of tissue transglutaminase that prevent gluten deamidation, and peroral peptidase supplementation that aids complete digestion of immunostimulatory peptides.^1,8 At any rate, much work remains before we know whether any of these treatment modalities represent real treatment alternatives to the currently safe, but cumbersome, gluten free diet.