Aims/hypothesis Immunomodulators blocking cytotoxic T-lymphocyte-associated protein 4 (CTLA-4) and programmed cell death protein 1 (PD-1) or programmed death-ligand 1 (PD-L1) have improved the treatment of a broad spectrum of cancers. These immune checkpoint inhibitors (ICIs) reactivate the immune system against tumour cells but can also trigger autoimmune side effects, including type 1 diabetes. Mesenchymal stem cell (MSC) therapy is the most prevalent cell therapy, with tissue-regenerating, anti-fibrosis and immunomodulatory functions provided by the secretome of the cells. Here, we examined whether systemic MSC treatment could prevent the development of type 1 diabetes in a NOD mouse model. Methods The purified PD-L1 monoclonal antibody was administered to induce diabetes in male NOD mice which normally do not develop diabetes. Human adipose-derived MSCs were administered by tail vein injections. T cells, macrophages and monocyte-derived macrophages expressing C-X-C motif chemokine ligand 9 (CXCL9) in pancreatic sections of NOD mice and a cancer patient who developed diabetes following the ICI treatments were analysed by immunofluorescence. Tissue localisation of the injected MSCs, plasma exosome levels and plasma cytokine profiles were also investigated. Results PD-1/PD-L1 blockade induced diabetes in 16 of 25 (64%) NOD mice which received anti-PD-L1 mAb without hMSCs [MSC(−)], whereas MSC administration decreased the incidence to four of 21 (19%) NOD mice which received anti-PD-L1 mAb and hMSCs [MSC(+)]. The PD-1/PD-L1 blockade significantly increased the area of CD3-positive T cells (6.2-fold) and macrophage-2 (Mac-2) antigen (2.5-fold)- and CXCL9 (40.3-fold)-positive macrophages in the islets. MSCs significantly reduced T cell (45%) and CXCL9-positive macrophage (67%) accumulation in the islets and the occurrence of diabetes. The insulin content (1.9-fold) and islet beta cell area (2.7-fold) were also improved by MSCs. T cells and CXCL9-positive macrophages infiltrated into the intricate gaps between the beta cells in the islets by PD-1/PD-L1 blockade. Such immune cell infiltration was largely prevented by MSCs. The most striking difference was observed in the CXCL9-positive macrophages, which normally did not reside in the beta cell region in the islets but abundantly accumulated in this area after PD-1/PD-L1 blockade and were prevented by MSCs. The CXCL9-positive macrophages were also observed in the islets of a cancer patient who developed diabetes following the administration of ICIs but few CXCL9-positive macrophages were observed in a control patient. Mechanistically, the injected MSCs accumulated in the lung but not in the pancreas and strongly increased plasma exosome levels and changed plasma cytokine profiles. Conclusions/interpretation Our results suggest that MSCs can prevent the incidence of diabetes associated with immune checkpoint cancer therapy and may be worth further consideration for new adjuvant cell therapy. Graphical abstract

No abstract

No abstract

The fat-derived factor, adiponectin, is considered a salutary circulating factor. We recently demonstrated that native adiponectin binds T-cadherin and promotes intracellular biogenesis and secretion of the exosome. Exosomes play important roles in various aspects of homeostasis, including glucose and energy metabolism. However, it remains unclear whether and how the promotion of exosome production by adiponectin in vivo is beneficial for glucose and lipid metabolism. In the present study, overexpression of human adiponectin in mice resulted in an increased number of circulating exosomes, but it did not significantly improve glucose metabolism, change body weights, or change triglyceride clearance under a high-fat diet. Multiple small doses of streptozotocin increased blood glucose and decreased triglyceride clearance similarly in both wildtype and transgenic mice. Thus, these results indicated that human adiponectin overexpression in mice increases plasma exosomes but does not significantly influence glucose and lipid metabolism.