Footprint of pancreas infiltrating and circulating immune cells throughout type 1 diabetes development.

Introduction: Type 1 diabetes (T1D) is defined by immune cell infiltration of the pancreas, in particular the islets of Langerhans, referred to as insulitis, which is especially prominent during the early disease stages in association with decreased beta cell mass. An in-depth understanding of the dynamics and phenotype of the immune cells infiltrating the pancreas and the accompanying changes in their profiles in peripheral blood during T1D development is critical to generate novel preventive and therapeutic approaches, as well as to find biomarkers for the disease process. Methods: Using multi-parameter flow cytometry, we explored the dynamic changes of immune cells infiltrating the pancreas and the pancreatic draining lymph nodes (PLN), compared to those in peripheral blood in female and male non-obese diabetic (NOD) mice during T1D progression. Results: The early stages of T1D development were characterized by an influx of innate dendritic cells and neutrophils in the pancreas. While dendritic cells seemed to move in and out (to the PLN), neutrophils accumulated during the pre-symptomatic phase and reached a maximum at 8 weeks of age, after which their numbers declined. During disease progression, CD4+ and CD8+ T cells appeared to continuously migrate from the PLN to the pancreas, which coincided with an increase in beta cell autoimmunity and insulitis severity, and a decline in insulin content. At 12 weeks of age, CD4+ and especially CD8+ T cells in the pancreas showed a dramatic shift from naïve to effector memory phenotype, in contrast to the PLN, where most of these cells remained naïve. A large proportion of pancreas infiltrating CD4+ T cells were naïve, indicating that antigenic stimulation was not necessary to traffic and invade the pancreas. Interestingly, a pre-effector-like T cell dominated the peripheral blood. These cells were intermediates between naïve and effector memory cells as identified by single cell RNA sequencing and might be a potential novel therapeutic target. Conclusion: These time- and tissue-dependent changes in the dynamics and functional states of CD4+ and CD8+ T cells are essential steps in our understanding of the disease process in NOD mice and need to be considered for the interpretation and design of disease-modifying therapies.

Preventing type 1 diabetes in late-stage pre-diabetic NOD mice with insulin: A central role for alum as adjuvant.

Background: Restoration of immune tolerance to disease-relevant antigens is an appealing approach to prevent or arrest an organ-specific autoimmune disease like type 1 diabetes (T1D). Numerous studies have identified insulin as a key antigen of interest to use in such strategies, but to date, the success of these interventions in humans has been inconsistent. The efficacy of antigen-specific immunotherapy may be enhanced by optimising the dose, timing, and route of administration, and perhaps by the inclusion of adjuvants like alum. The aim of our study was to evaluate the effect of an insulin peptide vaccine formulated with alum to prevent T1D development in female non-obese diabetic (NOD) mice when administered during late-stage pre-diabetes. Methods: Starting at 10 weeks of age, female NOD mice received four weekly subcutaneous injections of an insulin B:8-24 (InsB:8-24) peptide with (Ins+alum) or without Imject® alum (Ins) as adjuvant. Diabetes incidence was assessed for up to 30 weeks of age. Insulin autoantibodies and C-peptide concentrations were measured in plasma and flow cytometric analysis was performed on pancreatic-draining lymph nodes (PLN) and pancreas using an InsB:12-20-reactive tetramer. Results: InsB:8-24 peptide formulated in alum reduced diabetes incidence (39%), compared to mice receiving the InsB:8-24 peptide without alum (71%, P < 0.05), mice receiving alum alone (76%, P < 0.01), or mice left untreated (70%, P < 0.01). This was accompanied by reduced insulitis severity, and preservation of C-peptide. Ins+alum was associated with reduced frequencies of pathogenic effector memory CD4+ and CD8+ T cells in the pancreas and increased frequencies of insulin-reactive FoxP3+ Tregs in the PLN. Of interest, insulin-reactive Tregs were enriched amongst populations of Tregs expressing markers indicative of stable FoxP3 expression and enhanced suppressive function. Conclusion: An InsB:8-24 peptide vaccine prevented the onset of T1D in late-stage pre-diabetic NOD mice, but only when formulated in alum. These findings support the use of alum as adjuvant to optimise the efficacy of antigen-specific immunotherapy in future trials.