The causal relationship between neuromyelitis optica spectrum disorder and other autoimmune diseases

Abstract

Objectives: The coexistence of neuromyelitis optica spectrum disorder (NMOSD) with other autoimmune diseases has been well recognized. However, the causal association between these two conditions has not been fully studied. The etiology and therapies of NMOSD coexisting with autoimmune diseases also need to be elucidated.

Methods: We performed two-sample Mendelian randomization (MR) analysis to examine the causality. Genome-wide association (GWAS) summary data from NMOSD, autoimmune thyroid disease (AITD), systemic lupus erythematosus (SLE), and Sjogren's syndrome (SS) were used to identify genetic instruments. Causal single-nucleotide polymorphisms (SNPs) were annotated and searched for cis-expression quantitative trait loci (cis-eQTL) data. Pathway enrichment analysis was performed to identify the mechanism of NMOSD coexisting with AITD, SLE, and SS. Potential therapeutic chemicals were searched using the Comparative Toxicogenomics Database.

Results: The MR analysis found that AITD, SLE, and SS were causally associated with NMOSD susceptibility, but not vice versa. Gene Ontology (GO) enrichment analysis revealed that MHC class I-related biological processes and the interferon-gamma-mediated signaling pathway may be involved in the pathogenesis of NMOSD coexisting with AITD, SLE, and SS. A total of 30 chemicals were found which could inhibit the biological function of cis-eQTL genes.

Conclusions: Our findings could help better understand the etiology of NMOSD and provide potential therapeutic targets for patients with coexisting conditions.

Keywords: Mendelian randomization; autoimmune diseases; causal relation; genome-wide associated study; neuromyelitis optica spectrum disorders.

Figure 1

Flowchart of the study’s design. Summary data of GWAS traits were extracted, then MR analysis was performed to detect the causal relationships between NMOSD and other ADs (AITD, SLE, and SS). The frequencies of causal SNPs were searched in 1000 Genomes. The causal SNPs were annotated, and gene enrichment analysis was performed. Chemicals that may affect the enriched pathways were analyzed.

Figure 2

Forest plots of Mendelian randomization analysis that estimates the risk of NMOSD on the exposure of AITD, SLE, and SS. nsnp, number of causal SNPs; CI, confidence interval.

Figure 3

Forest plots of Mendelian randomization analysis that estimates the risk of AITD, SLE, and SS on the exposure of NMOSD. nsnp, number of causal SNPs; CI, confidence interval.

Figure 4

Gene Ontology enrichment analysis of the cis-eQTL genes. Dots indicate the number of genes that enriched in certain pathways; columns indicate the - log10(p-value) of pathway enrichment analysis.

Figure 5

Chemical–biological process interaction network. Blue boxes indicate significant enriched GO pathways, and yellow boxes indicate potential therapeutic chemicals. Each line indicates an interaction between the GO pathway and chemical. Circles indicate chemicals that serve the similar function.

Figure 6

The mechanism of NMOSD coexisting with other ADs. Both environmental and genetic factors contribute to the pathogenesis of peripheral ADs. After peripheral ADs occur, lymphocytes may become hyperactive and produce high levels of inflammatory cytokines, which would activate the low-response or bystander T cells specific to AQP4. Then, the astrocytes would be attacked by AQP4-specific CD8+ T cells and the AQP4-IgG antibody.