Regulation of viral and autoimmune responses.

We have studied the induction and effector function of Th2-like regulatory cells in mouse models for type 1 diabetes (NOD and RIP-LCMV). CD4+ lymphocytes with specificity for insulin can be induced by immunization with the insulin B chain via the oral route or by DNA vaccination. Such cells are protective upon adoptive transfer and prevent diabetes development in syngeneic pre-diabetic recipients. In comparison to non-regulatory insulin B-specific cell lines, they produce high amounts of interleukin (IL)4 and IL10, whereas interferon (IFN)gamma and tumour necrosis factor (TNF)alpha levels are comparable. Indeed, IL4 is essential for the protective capability, as evidenced by use of IL4-deficient mice and sorting of IL4+ versus IL4- lymphocytes prior to transfer. Mechanistically, these cells act as bystander suppressors in the pancreatic draining node, the location where their cognate antigen, insulin B, is presented during the pre-diabetic inflammatory process. As a consequence, the autoaggressive response is locally dampened. We propose that this is achieved by modulation of antigen presenting cells that lose the ability to propagate aggressive responses after exposure to IL4 or IL10 in vitro. The clinically attractive side of our strategy is that it only acts as the site of inflammation, thus circumventing systemic side effects. In order to avoid induction of insulin B-specific autoaggressive T cells we have demonstrated that administration of IL4 or IL10 at the time of immunization is beneficial and therefore should be part of a potential future clinical application. Interestingly, these Th2-like regulators share in our systems no features with the so-called CD25+ regulatory cells, whose antigen specificity is still unclear. However, we have recent evidence that virus specific CD25+ cells can be generated and are able to affect antiviral responses in vivo.

Colitogenic Th1 cells are present in the antigen-experienced T cell pool in normal mice: control by CD4+ regulatory T cells and IL-10.

CD4(+) regulatory T cells have been shown to prevent intestinal inflammation; however, it is not known whether they act to prevent the priming of colitogenic T cells or actively control these cells as part of the memory T cell pool. In this study, we describe the presence of colitogenic Th1 cells within the CD4(+)CD45RB(low) population. These pathogenic cells enrich within the CD25(-) subset and are not recent thymic emigrants. CD4(+)CD45RB(low) cells from germfree mice were significantly reduced in their ability to transfer colitis to immune deficient recipients, suggesting the presence of commensal bacteria in the donor mice drives colitogenic T cells into the Ag-experienced/memory T cell pool. This potentially pathogenic population of Ag-experienced T cells is subject to T cell-mediated regulation in vivo by both CD4(+)CD25(+) and CD4(+)CD25(-) cells in an IL-10-dependent manner. Furthermore, administration of an anti-IL-10R mAb to unmanipulated adult mice was sufficient to induce the development of colitis. Taken together, these data indicate that colitogenic Th1 cells enter into the Ag-experienced pool in normal mice, but that their function is controlled by regulatory T cells and IL-10. Interestingly, IL-10 was not absolutely required for CD4(+)CD25(+) T cell-mediated inhibition of colitis induced by transfer of naive CD4(+)CD45RB(high) cells, suggesting a differential requirement for IL-10 in the regulation of naive and Ag-experienced T cells.

Reduction of antiviral CD8 lymphocytes in vivo with dendritic cells expressing Fas ligand-increased survival of viral (lymphocytic choriomeningitis virus) central nervous system infection.

In vivo administration of APC expressing Fas ligand (Fas-L(+) dendritic cells (DCs)) has shown promise in dampening allergic reactions and transplant rejection. Since the effect in these studies was mainly on CD4 lymphocytes, our goal was to evaluate the ability of such killer DCs to eliminate antiviral CD8 lymphocytes and in this way ameliorate viral immunopathology or, conversely, impede viral clearance. Intravenous administration of Fas-L(+) DCs resulted in a 50% reduction of lytic CD8 precursors following intracerebral infection with lymphocytic choriomeningitis virus (LCMV), and accordingly, immunopathology and survival of LCMV meningitis were improved, whereas viral clearance remained unaffected. In transfer studies the effect of the Fas-L(+) DCs was only quantifiable on experienced, not naive, CD8 lymphocytes. Importantly, loading of Fas-L(+) DCs with viral Ag before therapy was not necessary to achieve this effect, indicating that non-LCMV-infected Fas-L(+) DCs acquired viral Ag during acute LCMV infection in vivo. Our studies delineate important aspects for the clinical use of Fas-L(+) DCs in vivo. One should expect that they acquire viral Ags and suppress antiviral CD8 responses to some degree when given while an acute infection is ongoing. In terms of safety it is encouraging that resolution of the infection, at least in the case of LCMV, is not inhibited.