Genetic analysis of IgG N-glycosylation in health and disease
Item statusRestricted Access
Embargo end date31/12/2100
Glycosylation is among the most common post-translational protein modifications. Glycans are complex carbohydrates attached to the surface of many proteins, but are rarely extensively studied in a high-throughput manner. However, there is an increasing evidence of their involvement in various physiological processes and diseases. Glycosylation of Immunglobulin G was shown to be important in adaptive immunity, where it can act as a "safety switch" for different types of the immune response. Although the main enzymes of the glycosylation pathway are known, little is understood about how this template-independent process is regulated to result in a faithful synthesis of a specific glycoform. This question was previously addressed using genome-wide association studies (GWAS) and 9 loci were identified as being significantly associated with IgG N-glycosylation. Only 4 of these loci were the known glycosylation enzymes. An additional five loci were discovered by applying a newly developed multivariate GWAS method on the same dataset. Here, by performing a GWAS on 77 IgG N-glycan traits measured by ultra-performance liquid chromatography in more than 8000 samples from four European cohorts the number of genome-wide significant (𝑝�≤ 2.4 × 10−9) loci increased to 27, 15 of which are novel, with 6 additional loci being suggestively associated (𝑝� ≤ 2.4 × 10−8). To assess which of the genes from the associated loci are more likely to be regulating IgG glycosylation, different gene prioritising strategies were employed. For 7 loci evidence of a non-synonymous amino acid change was found, two of which were predicted to be deleterious. Evidence of regulation through changes in gene expression levels in B-cells, the cell lineage responsible for production of IgG, was found for 4 genes, with an additional 11 genes exhibiting the same evidence with expression in peripheral blood or other immune cells. For the remaining loci the most likely candidate gene was proposed based on co-expression with genes from the enriched gene-sets or based on a physical proximity to the variant with the strongest association. To narrow down the most important loci for a functional follow-up, the omics nature of this data was used to compare glycome-wide SNP effects and suggest how newly discovered loci form a functional network that regulates the established members of the glycosylation pathway. The potential role of IgG glycosylation in various complex traits and diseases was explored by assessing the pleiotropy of the associated SNPs. The inflation of SNPs related to autoimmune, digestive and neurological diseases was observed in glycosylation SNPs. To assess whether IgG N-glycosylation is likely to share the same causal variant as the identified pleiotropic traits and diseases, regional association patterns were compared using summary data based Mendelian Randomisation analyses. This work demonstrates that an increased sample size empowered the identification of novel loci, enabling further insights into the molecular mechanisms underlying protein glycosylation and its relationship with complex human diseases. It also shows that such analyses of omic traits can assist in creating a functional network of the identified loci, prioritising the most important genes and allowing a more focused approach to future experimental functional follow-up.