BackgroundDysbiosis of the gut microbiota has been implicated in the pathogenesis of many autoimmune conditions including type 1 diabetes (T1D). It is unknown whether changes in the gut microbiota observed in T1D are due to environmental drivers, genetic risk factors, or both. Here, we have performed an analysis of associations between the gut microbiota and T1D genetic risk using the non-obese diabetic (NOD) mouse model of T1D and the TwinsUK cohort.ResultsThrough the analysis of five separate colonies of T1D susceptible NOD mice, we identified similarities in NOD microbiome that were independent of animal facility. Introduction of disease protective alleles at the Idd3 and Idd5 loci (IL2, Ctla4, Slc11a1, and Acadl) resulted in significant alterations in the NOD microbiome. Disease-protected strains exhibited a restoration of immune regulatory pathways within the gut which could also be reestablished using IL-2 therapy. Increased T1D disease risk from IL-2 pathway loci in the TwinsUK cohort of human subjects resulted in some similar microbiota changes to those observed in the NOD mouse.ConclusionsThese findings demonstrate for the first time that type 1 diabetes-associated genetic variants that restore immune tolerance to islet antigens also result in functional changes in the gut immune system and resultant changes in the microbiota.Electronic supplementary materialThe online version of this article (10.1186/s40168-018-0417-4) contains supplementary material, which is available to authorized users.

The microbial community making up the gut microbiota can profoundly influence intestinal homeostasis and immune system development, and is believed to influence the development of complex diseases including type 1 diabetes (T1D). T1D susceptible nonobese diabetic (NOD) mice have been shown to harbor a distinct microbiota to disease-protected mice. We hypothesized that the T1D susceptible genetic background of NOD mice would be resistant to the introduction of a C57BL/6-derived microbiota. NOD and C57BL/6 mice were cohoused either continually from birth, from birth until weaning or from weaning onwards, allowing transfer of microbiota between the mice. Cohousing NOD with C57BL/6 mice from before birth, resulted in moderate changes to the gut microbiota, whereas initiating cohousing at weaning only led to minimal changes. Terminating cohousing at weaning reduced the changes in the microbiota composition. However, diabetes onset was not significantly delayed and there was no reduction in intestinal inflammation or the proportion of regulatory T cells in the cohoused NOD mice. However, insulin but not islet-specific glucose-6-phosphatase catalytic subunit-related protein-specific CD8 T cells were reduced by cohousing suggesting an epitope-specific modulation of the autoreactive response by the gut microbiota. These results suggest that the T1D susceptible genetic background of the NOD mouse was resistant to the introduction of a C57BL/6-derived microbiota.

No abstract