V delta 1 bearing T cells comprise the major population of gamma/delta T cells in the human intestinal tract. To gain insight into mechanisms involved in the generation of these cells and the diversity of their repertoire, we have characterized the junctional sequences of V delta 1 T cell receptor transcripts in the human small intestine and colon. Mucosal biopsies obtained from defined regions along the length of the small intestine or colon contained a high frequency of either one or a few identical in frame V delta 1 sequences. Less abundant sequences were also detected repeatedly throughout the length of small intestine or colon. Moreover, the intestinal V delta 1 repertoire in the small intestine and colon appeared compartmentalized and showed no overlap with the V delta 1 repertoire in peripheral blood. Dominant V delta 1 transcripts in each subject differed between the small intestine and colon, and the dominant transcripts within these sites differed among individuals. Analysis of small intestinal transcripts obtained at a 1-yr interval revealed that the V delta 1 repertoire was stable over time. The fact that the majority of V delta 1 transcripts, both dominant and rare, are distributed throughout a several meter length of the adult intestinal tract and are stable over time suggests they are not generated by an ongoing process of in situ VDJ gene rearrangement. Our results favor a model in which the repertoire of V delta 1 T cells in the intestinal tract is shaped by positive selection in response to a limited array of ligands before the migration of V delta 1 cells throughout the small intestine or colon.

Murine L3T4 T cells freshly isolated from Peyer's Patch were fractionated based on differential adherence to Vicia villosa agglutinin (VVA). VVA adherent cells secreted IL-5, but not IL-2, after stimulation with Con A and IL-1. In striking contrast, VVA nonadherent PP L3T4 T cells secreted IL-2, but not IL-5, under the same conditions. In addition, supernatants from VVA adherent, but not from VVA nonadherent cells cultures, enhanced IgA secretion by LPS-stimulated splenic B cells to the same extent as purified IL-5. Thus, IL-5-producing T cells are present in PP in situ and may play an important role in the development of mucosal immunity. Further, differential adherence to VVA can be used to separate T cell populations that preferentially secrete IL-5 or IL-2.

This study is the first to describe a molecular marker that distinguishes the celiac disease HLA-D region haplotype from a serologically identical haplotype in unaffected controls. Using a DQ beta chain cDNA probe and the restriction endonuclease Rsa I, we have detected a polymorphic 4.0 kb fragment which, in DQw2 individuals, is associated with a 40-fold increased relative risk of developing celiac disease. This finding should permit the identification of the celiac disease susceptibility gene(s) in the HLA-D region and facilitate a more precise dissection of the molecular and immunogenetic mechanisms involved in the pathogenesis of that disease.

We have examined cytotoxic responses of lymphocytes derived from the gut epithelium of mice primed systemically and enterically with alloantigens. Both gut intraepithelial (IEL) and splenic lymphocytes from alloantigen-primed mice were found to contain antigen-specific cytotoxic T cell activity. However, after priming, gut IEL also developed high levels of natural killer and spontaneous cytotoxic cell activities. We suggest that this nonspecific activation of additional cytotoxic effector populations during an antigen-specific response is an important host immune defense within the intestinal mucosa.

Celiac disease in humans is activated by the dietary ingestion of wheat, rye, triticale, barley, and possibly oats. Gliadins in wheat and similar proteins in the other grains are known to activate disease in susceptible individuals. There is a striking association between celiac disease and HLA-B8, -DR3 and/or -DR7, and -DC3. Nonetheless, less than 0.2% of individuals with those serologic HLA specificities develop celiac disease and disease is not always concordant among monozygotic twins. We propose that additional environmental factors may be important in the pathogenesis of celiac disease. To investigate that possibility, we examined a data bank of protein sequences for other proteins that might share amino acid sequence homologies with A-gliadin, an alpha-gliadin component known to activate celiac disease and whose complete primary amino acid sequence is known. These studies demonstrate that A-gliadin shares a region of amino acid sequence homology with the 54-kD E1b protein of human adenovirus type 12 (Ad12), an adenovirus usually isolated from the intestinal tract. The region spans 12 amino acid residues, includes 8 residue identities and an identical pentapeptide, and is hydrophilic in both proteins. Antibody reactive with the 54-kD Ad12 E1b protein cross-reacts with A-gliadin, a 119 amino acid cyanogen bromide peptide of A-gliadin that spans the region of homology and a synthetic heptapeptide of A-gliadin from within the region of homology. We suggest that an encounter of the immune system with antigenic determinants produced during intestinal viral infection may be important in the pathogenesis of celiac disease.

Peyer's patch T cells may serve an important role in the interaction of the host with intraluminal gut antigens. Studies presented in this paper demonstrate that T cells in murine Peyer's patches can be carrier primed for helper function in the induction of an antihapten response by feeding antigen. Carrier priming was assessed by measuring the ability of Peyer's patch cells from mice fed heterologous erythrocytes to enhance an antitrinitrophenyl (TNP) response in vitro when added to normal Peyer's patch cells cultured with TNP coupled to the erythrocyte used for feeding. Priming of T helper cells in Peyer's patches was antigen specific and occurred when erythrocytes were administered orally but not when erythrocytes were injected intravenously or intraperitoneally. Murine Peyer's patches are naturally deficient in a cooperating accessory adherent cell type(s) required for B-cell induction to humoral antibody synthesis in vitro and antigen feeding does not result in significant induction of Peyer's patch B cells to humoral antibody synthesis in vivo. Since Peyer's patch T cells can be carrier-antigen primed for helper function in the absence of B-cell induction to humoral antibody synthesis, these studies may indicate that T-cell priming is less dependent than B-cell induction on cooperating accessory adherent cells.