Predicting disease-overarching therapeutic approaches for congenital disorders of glycosylation using multi-OMICS

Congenital Disorders of Glycosylation (CDG) are a rapidly expanding group of inherited metabolic diseases caused by defects in glycosylation. Although over 190 genetic defects have been identified, effective treatments remain available for only a few. We hypothesized that integrative analysis of multi-omics datasets from individuals with various CDG could uncover common molecular signatures and highlight shared therapeutic targets.

We compiled all publicly available RNA sequencing, proteomics and glycoproteomics datasets from patients with PMM2-CDG, ALG1-CDG, SRD5A3-CDG, NGLY1-CDDG, ALG13-CDG and PGM1-CDG, spanning different tissues, including induced cardiomyocytes, human cortical organoids, fibroblasts, and lymphoblasts. Differential expression and glycosylation analyses were performed, followed by Gene Set Enrichment Analysis (GSEA) to identify commonly dysregulated pathways. We then applied the EMUDRA drug prediction algorithm to prioritize candidate compounds capable of reversing these shared molecular signatures.

We identified four glycoproteins with consistent differential glycosylation across all eight glycoproteomics datasets. Six glycosylation sites and glycan structures were recurrently altered across CDG and showed partial correction with treatment. Pathway analysis revealed shared disruptions in autophagy, vesicle trafficking, and mitochondrial function. EMUDRA predicted several repurposable drug classes, including muscle relaxants, antioxidants, beta-adrenergic agonists, antibiotics, and NSAIDs, that could reverse key pathway abnormalities, particularly those involving autophagy and N-glycosylation.

Most dysregulated pathways were shared across CDG, suggesting the potential for common therapeutic strategies. Several candidate drugs targeting these shared abnormalities emerged from integrative analysis and warrant validation in future in vitro studies.