Deciphering disease-specific glycosylation: unraveling diabetes subtypes through serum glycopattern.

Latent autoimmune diabetes in adults (LADA) is a slowly progressing form of diabetes that develops in adulthood, characterized by autoimmune destruction of pancreatic β-cells and subsequent insulin deficiency, akin to type 1 diabetes (T1D). Due to its shared genetic, immunological, and metabolic features with both T1D and type 2 diabetes (T2D), LADA is frequently misdiagnosed and inappropriately treated as T2D. To address this, we developed the A.NG algorithm, which identifies serum glycopatterns by calculating the ratio of upregulated to downregulated N-glycans, thereby facilitating the detection of subtle glycan alterations specific to each diabetes subtype. Our method, which utilizes matrix-assisted laser desorption ionization (MALDI) for N-glycan profiling, revealed distinct glycan patterns across T1D, T2D, and LADA, with observed correlations achieving an AUC of 0.918 in this cohort. While these findings demonstrate the technical feasibility of detecting subtype-associated glycosylation changes, their clinical utility for subtype differentiation requires validation in larger studies with refined quantification approaches. Furthermore, complementary ELISA and intact glycopeptide analyses showed that enzymes like FUT8 and FUCA1 contribute to altered glycan expression patterns on specific glycoproteins, which could serve as potential biomarkers for LADA. In conclusion, the A.NG algorithm represents a promising novel approach for distinguishing between LADA and T1D or T2D, with the potential to significantly improve the diagnosis and management of these diabetes subtypes.