Shared Genetics of Multiple System Atrophy and Inflammatory Bowel Disease

Abstract

Background: Multiple system atrophy (MSA) is a rare neurodegenerative disease characterized by intracellular accumulations of α-synuclein and nerve cell loss in striatonigral and olivopontocerebellar structures. Epidemiological and clinical studies have reported potential involvement of autoimmune mechanisms in MSA pathogenesis. However, genetic etiology of this interaction remains unknown. We aimed to investigate genetic overlap between MSA and 7 autoimmune diseases and to identify shared genetic loci.

Methods: Genome-wide association study summary statistics of MSA and 7 autoimmune diseases were combined in cross-trait conjunctional false discovery rate analysis to explore overlapping genetic background. Expression of selected candidate genes was compared in transgenic MSA mice and wild-type mice. Genetic variability of candidate genes was further investigated using independent whole-exome genotyping data from large cohorts of MSA and autoimmune disease patients and healthy controls.

Results: We observed substantial polygenic overlap between MSA and inflammatory bowel disease and identified 3 shared genetic loci with leading variants upstream of the DENND1B and RSP04 genes, and in intron of the C7 gene. Further, the C7 gene showed significantly dysregulated expression in the degenerating midbrain of transgenic MSA mice compared with wild-type mice and had elevated burden of protein-coding variants in independent MSA and inflammatory bowel disease cohorts.

Conclusion: Our study provides evidence of shared genetic etiology between MSA and inflammatory bowel disease with an important role of the C7 gene in both phenotypes, with the implication of immune and gut dysfunction in MSA pathophysiology. © 2020 The Authors. Movement Disorders published by Wiley Periodicals LLC on behalf of International Parkinson and Movement Disorder Society.

Keywords: conjunctional false discovery rate; genetic overlap; inflammatory bowel disease; multiple system atrophy.

FIG. 1.

Flowchart representing major steps of the study. MSA, multiple system atrophy; CD, Crohn’s disease; IBD, inflammatory bowel disease, including CD, ulcerative colitis, and unclassified IBD cases; UC, ulcerative colitis; T1D, diabetes mellitus type 1; CeD, celiac disease; RA, rheumatoid arthritis; MS, multiple sclerosis.

FIG. 2.

Conditional Q-Q plot showing the relation between expected (x axis) and observed (y axis) significance of SNPs in the primary phenotype when markers are stratified by their P values in the conditional phenotype. A sequence of 4 nested strata is presented: blue, all SNPs; orange, P_{conditional_phenotype} < 0.1; yellow, P_{conditional_phenotype} < 0.01; and purple, P_{conditional_phenotype} < 0.001. Dashed black line demonstrates expected behavior under no association. The increasing degree of leftward deflection from the no-association line for strata of SNPs with higher significance in the conditional phenotype indicates putative polygenic overlap. (A) MSA conditioned on CD; (B) CD conditioned on MSA; (C) MSA conditioned on IBD; (D) IBD conditioned on MSA. [Color figure can be viewed at wileyonlinelibrary.com]

FIG. 3.

Manhattan plot of −log₁₀(conjFDR) for MSA and CD (orange)/IBD (blue). Horizontal dashed black line shows the significance threshold conjFDR = 0.05. For each significant locus, genes within 100 kb of the locus lead SNP are shown. Lead variants at each locus are shown as bold dots with black border. Variants in high LD (r² > 0.6) with the lead variant are shown as bold dots without border. [Color figure can be viewed at wileyonlinelibrary.com]

FIG. 4.

Comparison of expression in the midbrain between wild-type mice (blue bars) and transgenic MSA mice (orange bars) for candidate genes identified in the conjFDR and eQTL analyses. Genes with significantly different expression between wild-type and transgenic mice are marked either with as * (nominal P < 0.05), ** (nominal P < 0.01), or *** (nominal P < 0.005). C7 and MROH2B genes survive multiple testing with Bonferroni-Dunn correction (adjusted P < 0.05). Error bars show standard error of the mean. [Color figure can be viewed at wileyonlinelibrary.com]