Children at onset of type 1 diabetes show altered N-glycosylation of plasma proteins and IgG

Individual variation in plasma N-glycosylation has mainly been studied in the context of diabetes complications, and its role in type 1 diabetes onset is largely unknown. Our aims were to undertake a detailed characterisation of the plasma and IgG N-glycomes in patients with recent onset type 1 diabetes, and to evaluate their discriminative potential in risk assessment.

In the first part of the study, plasma and IgG N-glycans were chromatographically analysed in a study population from the DanDiabKids registry, comprising 1917 children and adolescents (0.6–19.1 years) who were newly diagnosed with type 1 diabetes. A follow-up study compared the results for 188 of these participants with those for their 244 unaffected siblings. Correlation of N-glycan abundance with the levels and number of various autoantibodies (against IA-2, GAD, ZnT8R, ZnT8W), as well as with sex and age at diagnosis, were estimated by using general linear modelling. A disease predictive model was built using logistic mixed-model elastic net regression, and evaluated using a 10-fold cross-validation.

Our study showed that onset of type 1 diabetes was associated with an increase in the proportion of plasma and IgG high-mannose and bisecting GlcNAc structures, a decrease in monogalactosylation, and an increase in IgG disialylation. ZnT8R autoantibody levels were associated with higher IgG digalactosylated glycan with bisecting GlcNAc. Finally, an increase in the number of autoantibodies (which is a better predictor of progression to overt diabetes than the level of any individual antibody) was accompanied by a decrease in the proportions of some of the highly branched plasma N-glycans. Models including age, sex and N-glycans yielded notable discriminative power between children with type 1 diabetes and their healthy siblings, with AUCs of 0.915 and 0.869 for addition of plasma and IgG N-glycans, respectively.

We defined N-glycan changes accompanying onset of type 1 diabetes, and developed a predictive model based on N-glycan profiles that could have valuable potential in risk assessment. Increasing the power of tests to identify individuals at risk of disease development would be a considerable asset for type 1 diabetes prevention trials.