Activation of Celiac Disease Immune System by Specific ax-Gliadin Peptides

Cereal Chem. 72(5):475-479 Two different gliadin molecules (designated oc-gliadin and ot/P-gliadin) were synthesized as 52 and 58 ten amino acid (aa) long overlapping peptides for the determination of their B-cell epitopes. Monoclonal antibodies and human serum pools revealed two epitopes common for both gliadins (peptide 14 aa:s 66-75 and peptides 340x aa:s 166-175, 36Wx/p aa:s 176-185) and two unique epitopes (x-gliadin peptides 48 aa:s 236-245 and aIp-gliadin peptide 52 aa:s 256-275). In addition, peptide 9 (QPYPQPQPFP) aa:s 41-50 and peptide 42 (LGQGSFRPSQ) aa:s 206-215 were detectable by monoclonal antibodies and serum pools from patients with untreated celiac disease but not by serum pools from disease control patients who had antigliadin antibodies. Patients Celiac disease, or gluten-sensitive enteropathy, is provoked in genetically susceptible individuals by dietary exposure to wheat gluten and similar proteins in other closely related cereals (Trier 1991, Marsh 1992). The typical lesion in small intestinal epithelium is villous atrophy with crypt hyperplasia leading to malabsorption of most nutrients. Common symptoms that result from the malabsorption include diarrhea and wasting away in adults or a failure to thrive in children. Although symptoms are wide ranging, monosymptomatic or symptomless manifestations are also often present (Visakorpi and Maki 1994). Removal of wheat, rye, barley, and oats from the diet results in histological and clinical recovery. Most work has focused on the gliadin proteins and the peptides derived from them by digestion with pepsin and trypsin as being primarily responsible for initiating changes in susceptible individuals that ultimately lead to epithelial damage (Shewry et al 1992). All the different types of gliadins appear to be active (Howdle et al 1984). The mechanisms by which the gliadintriggered immunological changes in the epithelium are initiated and lead to damage are not understood in any significant detail. At present, the most favored hypothesis involves interactions between human leucocyte antigen (HLA) class II peptide complex on the antigen-presenting cells and cell surface receptors of T-cell lymphocytes and other cells of the immune system (Marsh 1992). The most susceptible HLA class II DQ oc/x heterodimer (DQ2) on antigen-presenting cells in celiac disease is encoded by the DQA1 *0501 and DQB 1*0201 alleles. The presence of high levels of circulating antibodies to gliadins is a common (but not specific) feature of active celiac disease (Savilahti et al 1983). Whether or not there is a connection between the gliadin peptide sequences that stimulate production of circulating antibodies and the peptide sequences that initiate 'Medical School, University of Tampere, Finland. Corrseponding auhtor: Medical School, University of Tampere, POB 607, SF33101, Finland. Department of Dermatology, University of Turku, Finland. National Public Health Institute, Helsinki. Finnish Red Cross Blood Transfusion Service, Helsinki. 1nstitute of Medical Technology, University of Tampere, Finland. Department of Pediatrics, Tampere University Hospital, Finland. © 1995 American Association of Cereal Chemists, Inc. with celiac disease were also studied for their human leukocyte antigen (HLA) class II status (the presence of genetically determined proteins on antigen-presenting cells that are important for immunological recognition). Antigliadin antibody response to peptide QPYPQPQPFP was restricted by celiac disease (and HLA class II) because relative amounts of the antipeptide antibodies were significantly (P < 0.05) increased in celiac disease patients. The HLA alleles DQAl*0501 and DQB1*0201 are strongly associated with celiac disease. The difference between patients with celiac disease and healthy control subjects with regard to peptide QPYPQPQPFP suggest that this region in the gliadin molecule is of pathogenetic importance in celiac disease. the process leading to intestinal damage is not known. The most immunogenic stretches of amino acid sequences in gliadins are also not well characterized. A few specific ox-gliadin peptides have been studied both in vivo and in vitro (Karagiannis et al 1987, Mantzaris et al 1990, Sturgess et al 1994), but there have not been any systematic investigations published in which a series of peptides scanning through complete gliadin protein sequences have been investigated for the ability either to react with circulating antibodies or to produce cell-mediated responses that might lead to intestinal damage. In this investigation, we have evaluated the reactivity of linear B-cell epitopes (immunogenic peptide sequences) of an oXgliadin and an od13-gliadin (sequence based on a cDNA clone that might correspond to either an oxor 3-gliadin because they are structurally similar) by analyzing the reactivity with several mouse monoclonal antibodies (MAbs) and with serum pools from celiac patients of a series of 10-residue peptides that span the complete sequences in five-residue increments. We found several highly active peptides that react with circulating antibodies in human sera. MATERIALS AND METHODS Study Population During 1990-1992, 98 children attending the Department of Pediatrics at the Tampere University Hospital for suspected celiac disease underwent small bowel biopsy. Venous blood samples were taken on admission at the time of their initial visit. Seventeen children showed severe jejunal mucosal damage (subtotal villous atrophy) with crypt hyperplasia consistent with celiac disease. They were treated with gluten-free diet and they fulfilled the ESPGAN criteria for celiac disease (Maki et al 1989, Walker-Smith et al 1990). The other children who were suspected for celiac disease but had normal jejunal mucosal morphology were thus excluded for celiac disease (disease controls). Serum Pool Study Groups For the epitope scanning, patients were grouped by serum pools consisting of four individuals each. The first study group comprised three pools derived from 12 patients with untreated celiac disease. One of these pools consisted of patients with known HLA class II antigen (DQA1*0501 and DQB1*0201). The Vol. 72, No. 5,1995 475 celiac disease patients in the other two pools of this study group were not tested for the DQA and DQB alleles. All these patients were positive for antigliadin IgG antibodies. The second study group comprised three pools derived from 12 patients who had antigliadin IgG antibodies but were excluded for celiac disease. Two of these pools consisted individuals negative for HLA DQA1*0501 and DQB1*0201. Individuals in one pool were not tested. The third study group was made up of one pool with four disease control patients who were negative for HLA DQA1 *0501 and HLA DQB 1*0201 and who had no antigliadin antibodies. An additional pool of three MAbs was also used in epitope scanning. MAb clone numbers 15E10-1, 16C4-6, and 2C11-12 were raised in mice against commercial gliadin (a mixture of oc-, 3-, y-, and co-gliadins) polypeptides (30-40 kDa) obtained by preparative gel electrophoresis (Vainio and Varjonen 1995). They were of IgGI isotype except for 2C1 1-12 (not determined for isotype). Individual Serum Samples The antipeptide antibody levels were further determined for 17 patients with celiac disease (mean age 8.2 years, range 1.3-16.3 years) and 21 patients suspected for celiac disease but with normal jejunal morphology (disease controls) (mean age 7.6 years, range 0.4-13.8 years). The sera were from the diagnostic phase, and all the celiac disease patients and 19 disease control patients were positive for antigliadin antibodies (mean IgG level 82 and 84 EIU, respectively). The celiac disease patients were positive for HLA DQA1*0501 and DQB 1*0201, and the disease