β-cell dedifferentiation prior to insulitis prevents T1DM

The mechanisms by which β-cells contribute to their own destruction in the pathogenesis of type 1 diabetes mellitus (T1DM) are unclear. A study in Cell Metabolism now shows that deletion of the unfolded protein response (UPR) sensor IRE1α in β-cells triggers a transient β-cell dedifferentiation and prevents T1DM in a mouse model.

The UPR enables a cell to adapt to substantial endoplasmic reticulum (ER) stress. However, long-term ER stress can cause the UPR to become pro-apoptotic. Furthermore, the UPR is impaired in preclinical models of T1DM and in humans with T1DM. Using the non-obese diabetic (NOD) mouse model of T1DM, Feyza Engin and colleagues generated a mutant in which IRE1α was deleted from β-cells in mouse pups, before the development of insulitis (IRE1α^β–/– NOD mice). Islet morphology and architecture was assessed, islet cells were immunophenotyped and RNA sequencing was carried out on intact islets as well as on single islet cells.

Credit: The image shows altered islet composition in IRE1α^β–/– non-obese diabetic mice: glucagon-positive cells (red) and insulin-positive cells (green). Image courtesy of Feyza Engin/University of Wisconsin-Madison.

“IRE1α deletion in NOD β-cells before insulitis causes a transient dedifferentiation of β-cells, and these dedifferentiated β-cells show diminished expression of β-cell autoantigens and MHC class I molecules, while they increase their expression of immune inhibitory markers,” explains corresponding author Engin. Previous work found that deletion of IRE1α in prenatal or adult β-cells was associated with a diabetic phenotype. By contrast, the IRE1α^β–/– NOD mice showed only transient hyperglycaemia and were protected from T1DM up to 50 weeks of age.

Adoptive transfer experiments were carried out, transferring T cells from IRE1α^β–/– NOD mice or control mice to immunodeficient NOD Rag1^–/– mice. Recipient mice became diabetic 20 weeks after cell transfer if receiving T cells from control mice, whereas they did not develop diabetes mellitus if they received T cells from IRE1α^β–/– NOD mice. “IRE1α^β–/– NOD mice exhibit impaired T cell diabetogenic activity, which leads to protection from autoimmune destruction and diabetes mellitus,” says Engin. Of note, the pancreas of IRE1α^β–/– NOD mice had substantially reduced numbers of CD8⁺ T cells compared with control mice. By contrast, no significant alterations were observed in the levels of CD4⁺ T cells, B cells or macrophages. These findings suggest that modulating the UPR in β-cells before the immune infiltration can induce β-cell dedifferentiation that triggers immune tolerance and protects against disease.

“Whether stressed human β-cells adopt a similar mechanism for stress adaptation and escape from immune attack needs to be tested,” concludes Engin. “If this finding is replicated, then there might be a critical therapeutic window for intervention specifically for autoantibody-positive high-risk individuals to briefly alter their β-cell identity, induce immune tolerance and prevent T1DM.”