Dual role of interferon-gamma signalling pathway in sensitivity of pancreatic beta cells to immune destruction.

AIMS/HYPOTHESIS: Disruption of the interferon-gamma (IFN-gamma) signalling pathway at the level of interferon regulatory factor-1 (IRF-1) protects islets against cytokine-induced nitric oxide production and cell death in vitro. The aim of this study was to investigate the effects of a global disruption of IFN-gamma signalling, or a selective disruption of IRF-1, on beta-cell sensitivity to in vivo immune destruction. METHODS: In a first set of experiments, IFN-gamma receptor knockout mice (IFN-gammaR-/-) and interferon regulatory factor-1 knockout mice (IRF-1-/-) were rendered diabetic by injections of 50 mg streptozotocin i. p. on 5 consecutive days (MLDSTZ). RESULTS: Whereas no difference in sensitivity to MLDSTZ-induced diabetes could be observed between IFN-gammaR-/- mice and their 129/Sv/Ev controls (50% vs 55%, NS), there was an increased incidence of diabetes in IRF-1-/- mice (100% vs 67% in C57B1/6 mice, p < 0.05). A similar increased sensitivity to immune destruction of IRF-1-/- islets was observed when these islets were used as allografts. Islet graft survival rate of IFN-gammaR-/- and 129/Sv/Ev islets, when transplanted in alloxan-diabetic BALB/c recipients, was comparable (12.0 +/- 1.9 days vs 12.9 +/- 2.3 days, NS). Allograft rejection, however, of IRF-1-/- islets by BALB/c recipients occurred more rapidly than following transplantation to their C57B1/6 controls (9.1 +/- 2.0 days vs 13.1 +/- 1.5 days, p < 0.003). CONCLUSIONS/INTERPRETATION: These data indicate that IFN-gamma signal transduction at the beta-cell level is not essential for immune beta-cell destruction in vivo. Moreover, disruption of the IRF-1 gene in pancreatic islets increases susceptibility to beta-cell killing, suggesting that IRF-1 might be necessary for the expression of putative beta-cell "defence and/or repair" genes.

Early graft failure of xenogeneic islets in NOD mice is accompanied by high levels of interleukin-1 and low levels of transforming growth factor-beta mRNA in the grafts.

Early graft failure, graft rejection, and autoimmune recurrence remain unresolved issues in islet xenotransplantation in type 1 diabetes. The first aim of this study was to examine the existence of early graft failure in spontaneously diabetic autoimmune NOD mice after rat islet transplantation under technically controlled circumstances. The second aim was to examine the mediators of this early xenograft dysfunction. First, we demonstrated a higher percentage of early xenograft failure (48%) in spontaneously diabetic NOD mice as compared with chemically diabetic old NOD (13%, P < 0.05) and C57Bl/6 (7%, P < 0.01) mice. In addition, in spontaneously diabetic NOD mice, xenogeneic islets displayed early graft failure more frequently than allogeneic (23%, P < or = 0.05) or isogeneic islets (7%, P < 0.01). No early graft failure was observed in allotransplantation or isotransplantation in chemically diabetic mice. Reverse transcriptase-polymerase chain reaction analysis of cytokine mRNA in islet xenografts 8 h after transplantation showed higher levels of interleukin (IL)-1 mRNA in autoimmune diabetic mice compared with chemically diabetic old NOD mice (1.40 +/- 0.32 vs. 0.90 +/- 0.14 IL-1 copies/beta-actin copies, P < 0.05). In contrast, mRNA levels of transforming growth factor (TGF)-beta were lower in spontaneously diabetic NOD mice than in chemically diabetic old NOD mice (0.67 +/- 0.16 vs. 1.36 +/- 0.50 TGF-beta copies/beta-actin copies, P < 0.05). No differences in tumor necrosis factor-alpha, IL-6, and inducible nitric oxide synthase were seen between autoimmune and nonautoimmune diabetic mice. T-cell cytokines (IL-2, IL-4, IL-10, and gamma-interferon) were absent in all mice until 48 h after transplantation. These data suggest that early islet xenograft failure is more common in spontaneously diabetic NOD mice and could be due to a nonspecific inflammatory reaction locally in the grafts.

The role of interferon regulatory factor-1 in cytokine-induced mRNA expression and cell death in murine pancreatic beta-cells.

Combinations of cytokines, including interleukin-1beta (IL-1beta) and interferon-gamma (IFN-gamma), induce nitric oxide (NO) production and cell death in pancreatic islet cells. We have previously shown that these events are preceded by increased expression of the transcription factor interferon regulatory factor-1 (IRF-1). We utilized an IRF-1 knockout mouse (IRF-1-/-) to investigate the role of IRF-1 in cytokine-induced islet- and beta-cell gene expression and cell death. For this purpose, pancreatic islets or FACS-purified beta-cells were isolated from wild type (wt) or IRF-1-/- mice. These cells were exposed for different time points to IL-1beta (50 U/mI), IFN-gamma (1,000 U/ml) and/or TNF-alpha (1,000 U/ml) before being harvested for determination of viability (by nuclear dyes) and mRNA expression (by RT-PCR with specific primers). Following a 24 hours exposure to IL-1beta or IL-1beta + IFN-gamma, pancreatic islets isolated from IRF-1-/- mice presented a 30-50% reduction in medium nitrite accumulation and inducible NO-synthase (iNOS) expression. Interestingly, both wt and IRF-1-/- purified beta-cells failed to produce NO in response to IL-1beta alone, but presented a similar increase in nitrite accumulation and iNOS expression following exposure to IL-1beta + IFN-gamma. The basal expression of MHC class I mRNA was lower in IRF-1-/- islet cells (30% reduction), but there was a similar 2-4 fold-increase in MHC expression in islet cells from both strains following cytokine exposure. IL-1beta induced serine protease inhibitor-3 (SPI-3; a putative cellular "defense" protein) mRNA expression in both wt and IRF-1-/- islets or beta-cells. IFN-gamma decreased the IL-1beta-induced SPI-3 expression in wt islets or beta-cells, but induced a 5-fold increase in the expression of this mRNA in IRF-1-/- islets cells, suggesting that IRF-1 mediates an inhibitory effect of IFN-gamma on SPI-3 expression. Treatment of whole islets for 3 days with IL-1beta + IFN-gamma induced significantly more islet cell death in wt than in IRF-1-/- mice (respectively 85 +/- 3% versus 31 +/- 4% dead cells). On the other hand, prolonged exposure (3-9 days) of FACS-purified beta-cells to the same cytokines, or a combination of 3 cytokines, led to a similar increase in cell death in both IRF-1-/- and wt islets. In conclusion, IRF-1 contributes to cytokine-induced islet iNOS expression and cell death. These effects are absent in purified beta-cells, suggesting that IRF-1 may mediate its effects on whole islets via activation of non-endocrine cells (e.g. macrophages and ductal cells) present in these preparations.

Sex difference in resistance to dexamethasone-induced apoptosis in NOD mice: treatment with 1,25(OH)2D3 restores defect.

The NOD mouse, a model for type 1 diabetes, is characterized by resistance to apoptosis in immunocytes. The aim of this study was to investigate a link between apoptosis in NOD thymocytes and autoimmunity. First, we demonstrated that the sexual dimorphism in diabetes incidence in NOD mice (females are more diabetes-prone than males) is reflected by differences in apoptosis. Apoptosis in NOD thymocytes, 24 h after dexamethasone, was decreased in both sexes compared with C57B1/6, but it was lower in female mice (26 +/- 2%) than in male mice (50 +/- 3%, P < 0.001). Further, we demonstrated that sex hormones themselves play a central role in this difference, since castration of NOD male mice, which increases diabetes incidence, decreased apoptosis levels (32 +/- 2%), while treatment of NOD female mice with dihydrotestosterone, which protects against diabetes, restored apoptosis to male levels (42 +/- 1.5%). Finally, we demonstrated that 1,25-dihydroxyvitamin D3, a steroid hormone that prevents diabetes in NOD mice, restored apoptosis levels to C57B1/6 reference levels. This improved apoptosis was seen in male (68 +/- 1 vs. 50 +/- 3% in untreated NOD mice, P < 0.001) but especially in female NOD mice (51 +/- 5 vs. 26 +/- 2% in untreated NOD mice, P < 0.001). Fluorescence-activated cell sorter analysis of thymocyte subsets revealed marked differences, especially in CD4+CD8+ and CD4+ cells. We conclude that the sexual dimorphism in diabetes incidence in NOD mice is paralleled by a dimorphism in resistance to apoptotic signals in NOD thymocytes. This resistance to apoptosis is driven by sex hormones and is corrected by 1,25-dihydroxyvitamin D3.