Mucosal responses to novel food antigens in the normal and transplanted intestine

• Main Author: Weale, Andrew Robert
• Published: University of Bristol 2008
• Subjects: 616.079
• Online Access: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.486099

Despite recent advances, the intestine remains one ofthe most challenging solid organs to transplant, with graft and patient survival rates inferior to all other solid organs. The reasons for this are complex and reviewed in this thesis. We hypothesised that the transplantation ofa non-self mucosal immune system may have effects on the ability ofthe intestinal transplant recipient to control immune responses to food and commensal bacterial antigens in an appropriate way. The clinical implication of this hypothesis for intestinal transplantation is that, unrecognised, inappropriate inflammatory responses to antigens offood or commensal flora may contribute to graft . failure. Such episodes may be relatively preventable or controllable through dietary manipulation. It was appropriate within the thesis to firstly review the current understanding ofhow the non-transplanted gut mucosal immune system normally handles such antigens. The phenomenon oforal tolerance, demonstrable experimentally by a reduction in systemic responses to an antigen by prior oral administration of that antigen, is well recognised. However, less well recognised is that prior to the onset of systemic hypo-responsiveness, there is a period following feeding where the individual is primed. We hypothesised that this systemic response may be accompanied by early transient mucosal changes associated with T cell activation. We tested this hypothesis in a rat model, using quantitative immunohistology. However we found that feeding a novel, soluble protein antigen to rats at doses demonstrated to produce oral tolerance, appeared not to produce immunological (MHC class II expression; leucocyte or antigen presenting cell infiltration), or morphological changes in the epithelium or lamina propria, consistent with transient hypersensitivity preceding tolerance. Having investigated responses to fed antigen in the mucosal immune system of nontransplanted rats, we then wished to investigate the responses in a model of small intestinal transplantation. Donor lymphocytes within the graft ofthe transplanted intestine are progressively replaced by cells of recipient origin, independent of rejection, whilst stromal cells generally remain of donor origin. It has been shown in vitro that those stromal cells which express MHC class II (in humans, enterocytes), can present antigens to T lymphocytes. We hypothesised that this MHC-mismatch between T-lymphocytes and stroma, may affect the ability of the graft to mount appropriate immune responses to mucosally derived antigens. Thus in a previously stable intestinal graft, in which donor graft T-lymphocytes, restricted to donor MHC, do not recognise antigens presented in the context of stromal graft MHC, we predicted that feeding a novel protein would produce signs of active intestinal inflammation. To test the hypothesis a stable animal model of intestinal transplantation was required. We gained Home Office Approval to perform one-stage orthotopic intestinal transplants with sutured vascular anastomoses in rats, and established a small animal operating environment. Following modifications to our technique, we obtained a technical success rate (>72 hours survival) of33.3%. Unfortunately only 2 animals survived to 28 days - the point of planned oral challenge. Although we can make no firm or statistically based conclusions regarding the response to novel antigens in the intestinal graft, we made a number of observations which are at least consistent with the hypothesis that an active response was occurring. These included a greater degree of graft infiltration by host lymphocytes, a greater level of epithelial and lamina propria class II staining and higher levels of class WCD 45 dual positive staining in the graft ofthe animal fed ovalbumin compared to the control. The positive findings ofthe study are encouraging and have stimulated us to consider how to most effectively develop the model so that the hypothesis can be formally tested.