A central role of the gut epithelium in oral tolerance is suggested by the observation that its experimental induction depends on the preserved integrity of the mucosal barrier (Nicklin and Miller, 1983; Strobel et al., 1983). Suppressive effects resulting from interactions between the dominating TCRa/p CD8+ IEL subset and a normal epithelium represent one intriguing possibility, and there is some supporting evidence to this effect (Sachdev et al., 1993). It is possible that luminal antigenic peptides are presented by resting enterocytes, with inadequate co-stimulation to IELs or subepithelial CD4+ T-cells (Hoyne et al., 1993). Experiments in CD8 knockout mice have suggested that CD8+ T-cells are crucial for the down-regulation of enterically-elicited mucosal immunity but not for mucosally-induced suppression of systemic antibody responses (Grdic et al., 1998). Moreover, the chief effect obtained when enterocytes have been used as unconventional APCs in various test systems has been stimulation of CD8+ T-cells with a suppressor function (Hershberg and Mayer, 2000; Mayer et al., 2001). Human enterocytes express a ligand (gp180) that, by interaction with the a chain of CD8, may rapidly activate the tyrosine kinase p561ck and thereby trigger preferentially CD8+ T-cells (Li et al., 1995). Antigen presentation by MHC or CD1d molecules on enterocytes in this context could theoretically leave cognate IELs and even CD4+ lamina propria Th cells in an unresponsive state or induce an active down-regulatory potential by a deviated cytokine profile (Fig. 14.5). Moreover, basolateral exosomes with MHC class II-dependent antigen-presenting capacity may be released from the gut epithelium and act as 'tolerosomes' (Karlsson et al., 2001), either locally or at distant sites, such as mesenteric lymph nodes or the liver (Fig. 14.5).
The additional involvement of TCR7/8+ IELs in oral tolerance is also an intriguing possibility (Fig. 14.5), in view of the suggestion that this subset in the mouse may act as 'contrasuppressor cells', thereby being able to release intestinal IgA responses from T-cell-mediated suppression (Fujihashi et al., 1992). Subsequent studies have shown that this effect can probably be ascribed to IL-10 secreted by CD4+ T-cells, which are controlled by 7/8 T-cells operating through this down-regulatory cytokine in low-dose tolerance (Fujihashi et al., 1999). If this mechanism also operates in humans, the preferential expansion of intraepithelial 7/8 T-cells in the coeliac lesion might contribute to the striking
Peyer's patch Intestinal mucosa
Peyer's patch Intestinal mucosa
increase of Ig-producing immunocytes and activated lamina propria CD4+ T-cells seen in untreated patients (Scott et al., 1997). However, the increase of TCR7/8+ IELs in coeliac disease could instead reflect that they are cytotoxic cells involved in the clearance of microorganisms or damaged epithelium to preserve the surface barrier (Brandtzaeg, 1996a; Groh et al., 1998; Hershberg and Mayer, 2000).
Productive T-cell activation with appropriate proliferation and cytokine secretion requires two signalling events, one through the TCR and another through a receptor for some co-stimulatory molecule (Fig. 14.6). Without the latter signal, the T-cells mount only a partial response and, more importantly, may be sub jected to active tolerance induction (Nagler-Anderson, 2000) or anergy, with no capacity for production of their own growth-factor IL-2 upon restimulation (Janeway and Bottomly, 1994). The required co-stimulation for productive immunity is provided by soluble mediators, such as IL-1, and through cellular interactions, especially ligation of B7 (CD80/CD86) on professional APCs with CD28 on the T-cells (Robey and Allison, 1995). There is particularly great interest in the role of DCs in shaping the phenotypes of naive T-cells during such initial priming. Also, because DCs have migratory properties, they largely determine the tissue site in which primary immune responses will take place (Holt and Stumbles, 2000; Lanzavecchia and Sallusto, 2001).
Immature DC subsets are found both in the circulation and in most peripheral tissues, from which, after endocytosis of antigen, they generally migrate via draining lymphatics into regional lymph nodes to perform antigen presentation (Sallusto and Lanzavecchia, 1999). The actual expression level of various co-stimulatory molecules on the matured and activated DCs during the priming process influences the differentiation of naive T-cells in terms of cytokine
production - that is, a Th1 (IFN-7, IL-2 and tumour necrosis factor (TNF)-a) versus a Th2 (IL-4, IL-5, IL-10 and IL-13) profile. Interaction of the T-cell CD28 receptor with B7.1 (CD80) appears to favour the former and with B7.2 (CD86) the latter cytokine profile (Kuchroo et al., 1995). This Th1/Th2 paradigm is important in relation to atopic allergy, because IgE production as a basis for type I hypersensitivity is highly dependent on IL-4 and IL-13 (Corry and Kheradmand, 1999). Also, homoeostatic cross-regulation should ideally take place between the Th1 and Th2 responses (Romagnani, 2000).
Considerable information exists about putative aberrant immunoregulatory functions of non-professional APCs, such as keratinocytes, because they lack the appropriate co-stimulatory molecules necessary for productive immunity (Nickoloff and Turka, 1994). As alluded to above, this also applies to entero-cytes (Fig. 14.5). Thus, both B7 and intercellular adhesion molecule 1 (CD54) are virtually absent on normal human enterocytes (Bloom et al., 1995). Low levels of B7 might actually engage the high-affinity co-stimulatory molecule cytotoxic T lymphocyte antigen (CTLA)-4 on Th cells (Chambers and Allison, 1999), which could result in a down-regulatory response contributing to oral tolerance (Read et al., 2000).
In the normal state, even the subepithelial professional APCs in human gut mucosa, which have both macrophage and DC properties, show an extremely low level of B7 expression (Rugtveit et al., 1997; Brandtzaeg, 2001) and might therefore ligate CTLA-4 rather than CD28 on T-cells. Also, only B7.2 (CD86) is normally detectable, and this molecule has been shown in animal experiments to be important for low-dose oral tolerance (Liu et al., 1999). Functional characteristics of normal human lamina propria CD4+ T-cells do suggest that they are tightly controlled by suppression. First, they are remarkably unresponsive to signalling via the classical TCR/CD3 pathway alone, whereas anti-CD2 (particularly together with engagement of CD28) induces proliferation and cytokine secretion (Boirivant et al., 1996; Fuss et al., 1996). Second, they appear to be particularly susceptible to Fas (CD95)-mediated apoptosis, which might contribute to the limitation of clonal proliferation in the normal gut (De Maria et al., 1996). Third, they may be kept in check by prostaglandin E2 released by the gut epithelium or lamina propria macrophages (Newberry et al., 1999).
The fact that resident APCs from normal human gut mucosa are quite inert in terms of immune-productive stimulatory properties (Qiao et al., 1996) supports the notion that they play a central role in the induction of oral tolerance. One possibility is that, in the normal state (i.e. when subjected to only low-grade activation), they carry penetrating dietary and innocuous microbial antigens away from the mucosa, thereby avoiding local hyperactivation of immune cells (Fig. 14.5). Indeed, normal human intestinal mucosa shows only very low expression levels of mRNA for IFN-7, the key cytokine of activated Th1 cells (Nilsen et al., 1998). The same is true for Th2 cytokines, such as IL-4 and IL-5. Moreover, animal experiments have demonstrated that intestinal APCs can be triggered by pro-inflammatory factors to become mobilized (MacPherson et al., 1995) and even constitutively migrate rapidly with acquired epithelial elements and antigens away from the intestinal mucosa (Gutgemann et al., 1998; Huang et al., 2000). Such successful 'silent' antigen clearance probably depends on relatively low doses of absorbed antigen and may result in systemic T-cell-dependent tolerance induction (Fig. 14.5). Interestingly, in vivo expansion of the intestinal APC population enhanced the induction of oral tolerance in mice (Viney et al., 1998), whereas concurrent APC activation by immunization with cholera toxin or treatment of the animals with IL-1 resulted in productive immunity against the fed antigen (Williamson et al., 1999).
Animal studies have suggested differential effects of antigen dose and feeding frequency on the mechanisms of tolerance induction (Brandtzaeg, 1998). At very high doses, both Th1 and Th2 cells were shown to be deleted following initial activation, an event apparently depending on apoptosis in Peyer's patches (Chen et al., 1995). Anergy and clonal deletion would be antigen-specific events, in contrast to active suppression resulting from deviation of cytokine profiles induced by T-cell stimulation locally or in regional lymph nodes or the liver (Knolle et al., 1999; Limmer et al., 2000) after distant transport of antigen in APCs or epithelial exosomes (Fig. 14.5). Experiments performed to induce therapeutic tolerance via the gut in various autoimmune disease models have relied on a bystander effect of stimulated T-cells, which, through immune deviation, have preferentially secreted down-regulatory cytokines, particularly TGF-p (Weiner et al., 1994). It has been suggested that the gut harbours T-cells with a propensity for secretion of TGF-p (so-called Th3 cells), which appear to be particularly resistant to apoptosis (Chen et al., 1995), but this subset has not been clearly identified in humans. Another regulatory T-cell subset (Tr1), with a remarkable propensity for IL-10 production, has been identified in both the murine and the human gut (Groux et al., 1997; Khoo et al., 1997). This subset probably belongs to the activated (CD25+) and CTLA-4-expressing suppressive CD4+ T-cells induced after antigen feeding in mice (Zhang et al., 2001).
Altogether, a complex scenario may be proposed for oral tolerance, depending on apoptosis, when intestinal antigen exposure is excessive, and on anergy, due to lack of co-stimulatory APC molecules, antigen clearance from the mucosa and induction of immune deviation (skewing of T-cell cytokine profile) at lower antigen doses (Fig. 14.5). This scenario is further complicated by the fact that several cytokines contributing to the local profile are produced not only by T-cells, but also by APCs and epithelial cells - for instance, the down-regulatory cytokines TGF-p and IL-10. Furthermore, it remains unclear whether the most important immunoregulatory events for oral tolerance against dietary antigens take place in Peyer's patches, in the lamina propria, in systemic lymphoid organs or in the liver (Chen et al., 2000; Alpan et al., 2001; Fujihashi et al., 2001).
Was this article helpful?