Dissecting the interferon response required for triggering autoimmune diabetes

© 2016 Dr. Hong Sheng QuahType 1 diabetes (T1D) is characterized by the autoimmune destruction of β cells in the islets of Langerhans by immune cells. The interferon (IFN) family of cytokines has been implicated in diabetes pathogenesis, possibly via the activation of innate immune pathways. However, how IFNs contribute to the pathogenesis of autoimmune diabetes remains unclear. The overall aim of this thesis is to dissect the role of IFNs in the triggering of autoimmune diabetes using the non-obese diabetic (NOD) mouse model. Chapter 2 describes the contribution of type I IFN to the development of autoimmune diabetes in NOD mice. The islets of NOD mice displayed a unique mRNA expression pattern of IFN-induced genes that peaked at 3-4 weeks of age. Genetic ablation of type I IFN receptor in NOD mice (NOD.Ifnar1-/-) significantly reduced the expression of these genes. However, lack of IFNAR1 did not affect the development of diabetes. The overlapping function of type III IFN could be a reason for the lack of effect of IFNAR1-deficiency. Another reason could be that the natural level of type I IFN produced in wild-type NOD mice is insufficient to have an effect on diabetes development. In Chapter 3, the possible role of type III IFN in autoimmune diabetes was examined. β cells isolated from humans and mice expressed the type III IFN receptor and β cells were able to respond to type III IFN stimulation. Type III IFN was detected in the islets in NOD mice. These results indicate that type III IFN may contribute to diabetes development in NOD mice. Future experiments using NOD mice deficient in the type III IFN receptor will determine whether type III IFN contributes to the development of autoimmune diabetes. Engagement of pattern recognition receptors (PRRs) with exogenous and endogenous danger signals can trigger type I IFN production. Endogenous danger signals, such as aberrant cytosolic DNA, are normally eliminated to prevent unnecessary induction of PRRs. Nucleases, such as the 3’ exonuclease TREX1, are important in degrading aberrant DNA, and have been implicated in protection from autoimmunity. TREX1 is normally found in the SET complex, and one protease that can activate this complex is granzyme A. In Chapter 4, I show that granzyme A deficiency resulted in increased diabetes in NOD mice. Single-stranded DNA accumulated in the cytoplasm of dendritic cells and NK cells in the islets and spleens, and this was observed more frequently in NOD.Gzma-/- mice. Consistent with poor clearance of DNA and increased PRR activation, expression of IFN-induced genes was higher in the islets of NOD.Gzma-/- mice than NOD mice at 4 weeks of age. When NOD.Gzma-/- mice were crossed with NOD.Ifnar1-/- mice, diabetes returned to the rate observed in NOD mice. Overall, the data indicate that the natural level of type I IFN produced in wild-type NOD mice is not sufficient to cause diabetes, but excessive type I IFN can increase diabetes development in NOD.Gzma-/- mice. The results provide mechanistic insight into the triggering of autoimmune diabetes, suggesting that aberrant accumulation of cytoplasmic DNA and type I IFN production are important. In this study, these were caused by loss of granzyme A, but it is also possible that virus infection could similarly activate type I IFN and possibly also type III IFN to trigger T1D