The analysis of thymic iNKT cells showed higher frequency and absolute number of iNKT17 cells in NOD mice compared with C57BL/6 mice. Furthermore the analysis of the thymic stage 2 CD4− iNKT cell subset (containing iNKT17 cells) showed an enhanced expression of RORγt and IL-23R mRNA, two key molecules controlling IL-17 lineage 21. Thus, selleck compound our data suggest that the high frequency of iNKT17 cells in the peripheral tissues is subsequent

to an elevated frequency of iNKT17 cells in the thymus of NOD mice, which could be due to an elevated expression of RORγt in thymic iNKT cells upon their IL-17 lineage commitment. Not only are iNKT17 cells present at high frequency in NOD mice but more importantly, they infiltrate pancreatic islets of NOD mice. NOD pancreatic islets express the adhesion molecule E-cadherin, which interacts with the integrin CD103 36. Interestingly, 60% of pancreatic iNKT17 cells expressed CD103 integrin and retention of iNKT17 cells in the pancreas could be due to CD103/E-cadherin interactions as previously described for diabetogenic CD8 T cells in the context of islet allografts 37. Moreover, CD103 can act

as a co-activation molecule in human T lymphocytes 38 and could play a similar role in the activation of iNKT17 cells in the pancreas. While CCR6 is involved in the recruitment of Th17 cells in the target tissue in autoimmune CIA 39, the recruitment of iNKT17 cells in the pancreas is probably independent

of CCR6 since most of them do not express this molecule. Alternatively, Everolimus in vivo lack of expression of CCR6 might be due to downregulation upon entry into inflamed pancreas. Even though it has been suggested that iNKT17 cells are characterized by CCR6 and CD103 expression, the expression of these molecules by iNKT17 cells varies Pregnenolone depending on tissues. Since IL-17 protein is not detectable in absence of exogenous activation 19, 20, we analyzed IL-17 mRNA and other mRNAs associated with the IL-17 response. Importantly, IL-17 mRNA level was much higher in iNKT cells from the pancreatic islets than from PLNs and ILNs. No such difference in the mRNA level was observed for RORγt and IL-23R between these three tissues. Flow cytometry data showed that iNKT17 cells represent respectively 40% of iNKT cells in ILNs, 12% in PLNs and 6% in pancreas. The discrepancy between the frequency of iNKT17 cells in these three tissues and the spontaneous level of IL-17 mRNA suggests that pancreatic iNKT17 cells are locally activated in this tissue. Interestingly, IL-17, but not IFN-γ, mRNA expression by pancreatic iNKT cells was strongly decreased in mice lacking peripheral CD1d expression, demonstrating that local iNKT17 cell activation involves CD1d recognition. The residual expression of IL-17 mRNA in the absence of peripheral CD1d expression suggests that other local factors, such as IL-23 or IL-1β, could participate in the activation of iNKT17 cells 40.