The cross-talk between enterocytes and intraepithelial lymphocytes

Adaptive immune response to gluten in the small intestinal mucosa of celiac patients

Celiac disease (CD) is a small intestinal disorder provoked by a dysregulated immune response to wheat gluten and related proteins of barley and rye. This multifactorial disease, caused by both genetic and environmental factors, is characterized by a mild to severe enteropathy and a humoral response to the tissue-transglutaminase autoantigen. HLA class II DQ2 and DQ8 genes are the main risk factors for CD [8] and present gluten immunogenic peptides to lamina propria T lymphocytes, as documented by several studies [9]. Following gluten stimulation, a marked proliferation of Th1 and Th0 lymphocytes, and an increased production of inflammatory cytokines, mainly interferon-γ (IFN-γ), characterize the celiac intestinal mucosa. Up to now, several peptides, derived from gliadins (α-, ω-, and γ-gliadins) and from glutenins, have been identified to stimulate CD4+ T cells in CD subjects [9, 10]. Remarkably, gluten-specific CD4+ T cells can be isolated from the small intestinal biopsies of CD patients and not from non-celiac controls, thus underlining their pathogenic relevance [11].

Despite the pivotal role played by CD4+ T lymphocytes in the pathogenesis of CD, recent evidences suggested the involvement of CD8+ T cells in the inflammatory cascade elicited by gluten. The extensive infiltration of CD8+ T lymphocytes in the mucosa of CD patients, both in the epithelium and lamina propria, represents a histopathological hallmark [12]. Notably, gliadin contains peptides able to bind HLA class I molecules on the surface of antigen-presenting cells [11]. We have found that a short peptide of 10 amino acid length (QLIPCMDVVL), mapping the 123-132 region of α-gliadin stimulates CD8+ T lymphocytes in both peripheral blood and intestinal mucosa of HLA A2-positive celiac patients [13]. A further study with organ culture of celiac mucosal explants has demonstrated that this peptide specifically stimulates cytotoxic CD8+ T cells resident in the lamina propria and induces the apoptosis of enterocytes [14]. Interestingly, initial findings associated the HLA class I A*01 and B*08 genes with susceptibility to celiac disease (CD), though further analysis showed a primary association with alleles encoding for HLA DQ2/8 molecules [15]. By using bioinformatic algorithms for the prediction of peptides with binding motives to surface molecules encoded by A*01 and B*08 genes, we found several sequences in all three main gliadin families that bound to either HLA A1 or B8 molecules with high affinity. Selected peptides also recalled IFN-γ responses in A*01+ and B*08+ CD patients (Picascia et al. manuscript in preparation). A recent genome-wide association study (GWAS) has highlighted the role of B*08 gene, in strong linkage disequilibrium with DR3-DQ2.5, as additional susceptibility factor for CD [16]. Altogether, these findings sustain our previous data on HLA class I-restricted T cell responses in CD and strongly encourage further studies to dissect this, still unexplored, pathway in CD pathogenesis.

Innate immune response to gluten in the small intestinal mucosa of celiac patients and the role of IL-15

A consistent number of studies have demonstrated that gliadin contains peptides, particular mapping the NH2-terminal part of α-gliadin, able to stimulate innate immuno-competent cells, or even to exert a direct cellular toxicity [17]. After a brief incubation with gliadin peptide p31-43, Maiuri and co-workers reported that the uninflamed CD mucosa up-regulated the expression of ICAM-1, CD69, and HLA-DR markers and the density of interleukin-15 (IL-15)-positive cells. Upon a longer stimulation, an increased expression of CD83 on dendritic cells and of CD25 on T cells and macrophages was observed. This peptide also induced intraepithelial migration of CD8+ T cells in treated CD mucosa and an epithelial damage mainly due to enterocyte apoptosis. Moreover, the same authors found that an early activation of the innate cells might favor the capacity of gliadin peptides, particularly p31-43, to stimulate the adaptive T cell response to immunodominant gliadin epitopes [18]. IL-15 seems to have a key role as mediator of the gliadin-elicited innate activation, as the use of specific neutralizing antibodies inhibited a great part of the above pathways [18].

IL-15 is a pleiotropic pro-inflammatory cytokine, member of the IL-2 family, involved in several mechanisms of both immune and non-immune systems. It is expressed on the surface of mainly monocytes, macrophages, dendritic and epithelial cells in response to inflammatory stimuli. IL-15 has numerous activities, including anti-apoptotic function, stimulation of T cell proliferation, generation of cytotoxic T lymphocytes, and regulation of NK cell survival and function, as well as the induction of immunoglobulin synthesis by B cells. IL-15 is also an important growth and activator factor for both IELs and IECs. In small intestinal biopsies of healthy subjects, IL-15 protein is almost absent in the lamina propria but expressed, though at low intensity, on villous enterocytes. By contrast, in inflammatory condition, as in active CD mucosa, IL-15 is highly expressed by lamina propria mononuclear cells and by intestinal epithelial cells. The overexpression of IL-15 in the epithelium layer has a key role in TCR-independent expansion and activation of IELs in CD mucosa with villous atrophy [19]. Mention et al. reported that IL-15 was massively expressed at the surface of IECs and it was associated with an enhanced transcription of IFN-γ in untreated CD and in refractory celiac sprue patients (RCS), a pathological condition, characterized by massive intraepithelial infiltration with abnormal phenotype and unresponsive to the gluten-free diet [20]. The increased IL-15 levels on enterocytes selectively promoted the growth and activation of clonal CD103+ CD3− IELs in RCS.

Furthermore, it was demonstrated that in CD, IL-15 induced on the epithelial cells similar effects to those elicited by gliadin, as the TFR and FAS expression in both crypt and villous enterocytes, and the up-regulation of Ki67 expression, a marker of cell proliferation. Overall, these results suggested that gluten-induced IL-15 has a pivotal role in the epithelial layer and lamina propria modifications in CD mucosa [21].

There is a general consensus that the mechanisms involved in the villous atrophy in patients with full-blown CD include both the intensive epithelium infiltration by T cells and the increased expression of IL-15, and of non-classical MHC class I molecules, as MICA and MICB, on IECs. MICA and MICB proteins are ligands for the activating NKG2D receptors expressed on the surface of both TCRαβ and TCRγδ CD8+ T cells. MICs act as cellular stress signals and their recognition by NKG2D receptors induces several immune mechanisms, such as cellular cytotoxicity, proliferation, and cytokine secretion. IECs in the inflamed intestinal mucosa of CD patients show an increased expression of non-classical MHC class I molecules. At the same time, the IELs from CD subjects express high levels of the activating NKG2D and CD94/NKG2C receptors and mediate the killing of enterocytes via NKG2D. Additionally, IL-15, by up-regulating the NKG2D receptors, switches on the cytotoxic activity of IEL directed on the neighbor IECs [22]. As previously mentioned, in normal conditions, the cross-talk between IECs and IELs mediated by NKG2D/MIC might maintain the epithelium in a healthy condition eliminating infected cells. By contrast, in the case of uncontrolled IL-15 production by IECs induced by gluten, as in celiac intestinal mucosa, intraepithelial CD8+ T cells are converted into activated killer cells and contribute to epithelial cell destruction and tissue atrophy through a TCR-independent NKG2D signaling pathway [23]. A recent study by Setty and co-workers showed that IECs from healthy subjects with a family history of CD expressed higher levels of IL-15 and heat shock proteins (HSPs) than controls and have ultrastructural epithelial alterations [24]. In addition, they found that the cytotoxic IEL from relatives of CD patients expressed higher levels of activating NK receptors than cells from controls, although at lower levels than patients with active CD. By contrast, IEL from subjects with potential CD failed to up-regulate activating NK receptors, thus suggesting that the fine balance between inhibitory and activating molecules expressed by NK cells is variable in the different forms of the CD. These results re-mark that the innate epithelial stress is an important event to convert IELs into active killer cells and to induce villous atrophy.