Dendritic cells (DCs) are regulators of the immune response and are thought to be critical in maintaining tolerance towards the intestinal microbiota. Recent data have identified distinct subsets of DCs with specific functional properties. The objective of this thesis was to further define CD103⁺ and CX3CR1⁺ DCs in the intestine and to determine how DCs and regulatory T (Treg) cell responses are influenced by the microbiota. Using multicolour flow cytometry, we identified two CD103⁺ DC subsets with differential aldehyde dehydgrogenase (ALDH) activity and two populations of CX3CR1⁺ cells. In the mesenteric lymph node CD103⁺ALDH⁺ DCs were highly mature (CD86^hi, MHCII^hi), likely migratory (CCR7⁺) and enhanced Treg cell induction compared with ALDH⁻ DCs. CX3CR1^int cells accumulated during bacterially-induced colitis suggesting a pro-inflammatory role whereas CX3CR1^hi cells were associated with the production of the anti-inflammatory cytokine IL-10 during homeostasis. We also assessed the generation of CD103⁺ DCs from bone marrow progenitors. Although only small proportions of CD103⁺ DCs were detected in culture with FLT3L or GM-CSF alone, the combination of FLT3L and GM-CSF induced CD103⁺ DCs with a phenotype similar to those found in the small intestine. Using this system we showed that TLR ligands and retinoic acid induce ALDH enzyme activity in vitro. In order to assess how DCs and Treg cells respond to changes in the microbiota we employed broad-spectrum antibiotic treatment to deplete endogenous bacteria and also analyzed the impact of colonization with the model organism Helicobacter hepaticus. Interestingly, we did not detect alterations in the proportions of different DC subsets following antibiotic treatment or H. hepaticus infection. However, using a novel FoxP3^huCD2-IL-10^GFP reporter mouse, we found that IL-10 production by Treg cells was ablated following antibiotic treatment and significantly elevated following H. hepaticus infection. Preliminary investigation of the mechanism underlying this effect suggests a role for IL-27. In summary, this thesis provides further detail on the phenotype of intestinal DCs and shows that Treg cell IL-10 production is sensitive to the composition of the microbiota.

The mammalian immune system has developed diverse strategies to protect the gastrointestinal tract, as this tissue locale represents a huge absorptive surface and is susceptible to microbial breach. Paradoxically, one key aspect of this protective strategy is the maintenance of selected commensal microorganisms. These commensals serve essential roles in digestion, interfere with pathogenic microbial invasion and stimulate development of the host immune system. Therefore, immune responses which deplete these commensal populations are detrimental to the host. One effective intestinal immune response which selectively promotes the survival of commensals is production of antibodies of the IgA isotype which bind to bacteria without triggering inflammatory cytokines. Proteins of the tumor necrosis factor (TNF) family such as Lymphotoxin and BAFF contribute to the induction of IgA responses. Lymphotoxin is required for generation and organization of most organized lymphoid tissues, where B cell differentiation occurs, while BAFF is necessary for B cell survival and induces B cells to produce IgA. In this thesis, I describe work I have done in examining the roles of the TNF family members Lymphotoxin, BAFF and two related TNF family member cytokines, LIGHT and APRIL, in the regulation of IgA production in mice and in humans. Specifically, LIGHT over-expression drives immense production of IgA, leading to renal deposition of immune complexes in mice. Similar to LIGHT, BAFF over-expression drives increases in IgA production in the intestine, however I have shown that the effects of the BAFF pathway on IgA hyper-production are independent of LIGHT activity. Secondly, examining the phenotype of BAFF-over-expressing mice, I have shown that this phenotype resembles human IgA nephropathy (IgAN) and is dependent on intestinal commensals. Finally, I have described a lymphotoxin-dependent chemokine system in the intestinal lamina propria that could be responsible for organizing cells for the development of IgA responses in this mucosal site.

The mammalian immune system has developed diverse strategies to protect the gastrointestinal tract, as this tissue locale represents a huge absorptive surface and is susceptible to microbial breach. Paradoxically, one key aspect of this protective strategy is the maintenance of selected commensal microorganisms. These commensals serve essential roles in digestion, interfere with pathogenic microbial invasion and stimulate development of the host immune system. Therefore, immune responses which deplete these commensal populations are detrimental to the host. One effective intestinal immune response which selectively promotes the survival of commensals is production of antibodies of the IgA isotype which bind to bacteria without triggering inflammatory cytokines. Proteins of the tumor necrosis factor (TNF) family such as Lymphotoxin and BAFF contribute to the induction of IgA responses. Lymphotoxin is required for generation and organization of most organized lymphoid tissues, where B cell differentiation occurs, while BAFF is necessary for B cell survival and induces B cells to produce IgA. In this thesis, I describe work I have done in examining the roles of the TNF family members Lymphotoxin, BAFF and two related TNF family member cytokines, LIGHT and APRIL, in the regulation of IgA production in mice and in humans. Specifically, LIGHT over-expression drives immense production of IgA, leading to renal deposition of immune complexes in mice. Similar to LIGHT, BAFF over-expression drives increases in IgA production in the intestine, however I have shown that the effects of the BAFF pathway on IgA hyper-production are independent of LIGHT activity. Secondly, examining the phenotype of BAFF-over-expressing mice, I have shown that this phenotype resembles human IgA nephropathy (IgAN) and is dependent on intestinal commensals. Finally, I have described a lymphotoxin-dependent chemokine system in the intestinal lamina propria that could be responsible for organizing cells for the development of IgA responses in this mucosal site.

Changes to the composition and function of the gut mononuclear phagocyte (MNP) compartment are associated with the development of intestinal inflammation. Much work has focused on the role of MNPs in gut-associated lymphoid tissue in maintaining homeostasis, however little is known regarding the roles of MNPs during colitis. We have investigated MNPs in the large intestinal lamina propria during the steady state and inflammation. One of our primary aims was to determine the contribution of MNP subsets to intestinal pathology. For our studies of inflammation, we focused mainly on the Helicobacter hepaticus infection + anti-IL-10R model, which induces inflammation of the colon and caecum (typhlocolitis). We defined the composition of the MNP compartment alongside intestinal pathology scores throughout Hh + anti-IL-10R typhlocolitis. Peak pathology, 2-3 weeks after induction of colitis, coincided with peak frequencies of CX₃CR1^int Ly6C^+ MNPs. Having observed the accumulation of CX₃CR1^int CD64^+ monocyte/macrophage MNPs in the inflamed lamina propria, we conducted comparative whole genome microarray analysis of these cells isolated from the large intestine three weeks after Hh + anti-IL-10R treatment. CX₃CR1^int CD64^+ MNPs selectively expressed a variety of pro- and anti-inflammatory genes, including a number of genes which individually can both promote and negatively regulate inflammation. IL-23 is essential for Hh + anti-IL-10R-induced intestinal pathology. We investigated the role of MNPs as a source of IL-23 which drives Hh + anti-IL-10R colitis. Unexpectedly, our results indicate that normally hyporesponsive CX₃CR1^hi macrophages may act as the initial source of IL-23, which induces development of colitis. Recruitment of Ly6C^+ MHCII^+ MNPs to the lamina propria was IL-23-dependent, and these cells also expressed IL-23, which may establish a positive feedback loop of immune cell recruitment, activation and IL-23 production. Finally, we also examined how MNPs might be recruited to the colonic lamina propria during inflammation. Our studies support the conclusion that CCR6 is not required for accumulation of monocyte-derived populations in the inflamed intestine. We cannot rule out a role for CCR2, however preliminary data from the Hh + anti-IL-10R colitis model suggest a potential role for CCR1 or its close relation CCRL2. Such pathways could represent new therapeutic targets in inflammatory bowel disease.