Multi-Level Analyses of Genome-Wide Association Study to Reveal Significant Risk Genes and Pathways in Neuromyelitis Optica Spectrum Disorder

Abstract

Background: Neuromyelitis optica spectrum disorder (NMOSD) is an inflammatory disease of the central nervous system and it is understandable that environmental and genetic factors underlie the etiology of NMOSD. However, the susceptibility genes and associated pathways of NMOSD patients who are AQP4-Ab positive and negative have not been elucidated.

Methods: Secondary analysis from a NMOSD Genome-wide association study (GWAS) dataset originally published in 2018 (215 NMOSD cases and 1244 controls) was conducted to identify potential susceptibility genes and associated pathways in AQP4-positive and negative NMOSD patients, respectively (132 AQP4-positive and 83 AQP4-negative).

Results: In AQP4-positive NMOSD cases, five shared risk genes were obtained at chromosome 6 in AQP4-positive NMOSD cases by using more stringent p-Values in both methods (p < 0.05/16,532), comprising CFB, EHMT2, HLA-DQA1, MSH5, and SLC44A4. Fifty potential susceptibility gene sets were determined and 12 significant KEGG pathways were identified. Sixty-seven biological process pathways, 32 cellular-component pathways, and 29 molecular-function pathways with a p-Value of <0.05 were obtained from the GO annotations of the 128 pathways identified. In the AQP4 negative NMOSD group, no significant genes were obtained by using more stringent p-Values in both methods (p < 0.05/16,485). The 22 potential susceptibility gene sets were determined. There were no shared potential susceptibility genes between the AQP4-positive and negative groups, furthermore, four significant KEGG pathways were also identified. Of the GO annotations of the 165 pathways identified, 99 biological process pathways, 37 cellular-component pathways, and 29 molecular-function pathways with a p-Value of <0.05 were obtained.

Conclusion: The potential molecular mechanism underlying NMOSD may be related to proteins encoded by these novel genes in complements, antigen presentation, and immune regulation. The new results may represent an improved comprehension of the genetic and molecular mechanisms underlying NMOSD.

Keywords: Genome-wide association study; Neuromyelitis optica spectrum disorder; gene differential expression; gene sets; pathway.

FIGURE 1

Flow diagram of the three-phase analysis design. In phase I, two new gene-based tests for the GWAS approach, called VEGAS2 and PLINK, were used to conduct a gene-based association study. In phase II, Gene sets identified with meta p < 0.05/n were carried forward to the next phase. In phase III, protein-protein association networks were performed by STRING. Meanwhile, KEGG pathways and GO analysis were carried out for these NMOSD susceptibility genes. VEGAS2 and PLINK are two kinds of new gene-based tests. “n” represents the number of genes shared by the VEGAS2 and PLINK tests.

FIGURE 2

Network of known and predicted interactions between proteins encoded by NMOSD potential susceptibility genes were identified by GWAS of AQP4-positive NMOSD. Network nodes represent the proteins produced by a single, protein-coding gene locus. Edges represent protein-protein associations meant to be specific and meaningful. In STRING, blue edge represents protein-protein associations with known interactions from curated databases, purple edge represents protein-protein associations with known interactions experimentally determined, green edge represents protein-protein associations with predicted interactions with gene neighborhood, red edge represents protein-protein associations with predicted interactions with gene fusions, Navy blue edge represents protein-protein associations with predicted interactions with gene co-occurrence, and other associations with text mining (yellow edge), co-expression (black edge), and protein homology (lavender edge).

FIGURE 3

Network of known and predicted interactions between proteins encoded by NMOSD potential susceptibility genes were identified by GWAS of AQP4-negative NMOSD.

FIGURE 4

The top ten GO annotations of AQP4-positive NMOSD, (A) is for biological process, (B) is for molecular function, and (C) is for cellular component.

FIGURE 5

The top ten GO annotations of AQP4- negative NMOSD, (A) is for biological process, (B) is for molecular function, and (C) is for cellular component.