Identifying genetic modifiers of age-associated penetrance in X-linked dystonia-parkinsonism

Abstract

X-linked dystonia-parkinsonism is a neurodegenerative disorder caused by a founder retrotransposon insertion, in which a polymorphic hexanucleotide repeat accounts for ~50% of age at onset variability. Employing a genome-wide association study to identify additional factors modifying age at onset, we establish that three independent loci are significantly associated with age at onset (p < 5 × 10^-8). The lead single nucleotide polymorphisms collectively account for 25.6% of the remaining variance not explained by the hexanucleotide repeat and 13.0% of the overall variance in age at onset in X-linked dystonia-parkinsonism with the protective alleles delaying disease onset by seven years. These regions harbor or lie adjacent to MSH3 and PMS2, the genes that were recently implicated in modifying age at onset in Huntington's disease, likely through a common pathway influencing repeat instability. Our work indicates the existence of three modifiers of age at onset in X-linked dystonia-parkinsonism that likely affect the DNA mismatch repair pathway.

Fig. 1. Genome-wide association study (GWAS) on age at onset (AAO) in X-linked dystonia-parkinsonism (XDP) patients.

a Manhattan plot of the XDP AAO GWAS. We estimated a series of linear regression models assuming SNPs with a p-value below 5 × 10⁻⁸ to be significantly associated with AAO adjusted for repeat number. Using this approach, we identified 93 candidate SNPs to cover a potential genetic modifier influencing AAO. b, c Locus zoom plot of the GWAS data. The structure of linkage disequilibrium and –log10 (p-value) of the identified loci on chromosomes 5 and 7, respectively. Genomic coordinates are given according to GRCh37. Red lines: p = 5 × 10⁻⁸; Mb: megabase; each dot: −log10 (p-value) of a SNP; r²: the Pearson coefficient of correlation between any SNP in the graph and the SNP with the lowest p-value in that region. Source data are provided as a Source data file.

Fig. 2. Relationship between age at onset (AAO) and the lead single-nucleotide polymorphisms (SNPs) in the three independent regions identified in our genome-wide association study (GWAS).

a, b Two independent putative genetic loci on chromosome 5, the alternative alleles of which were found to accelerate disease onset. c The alternative allele of the chromosome 7 signal delays AAO in our patients. Box-plot elements: center line: median; box limits: upper and lower quartiles; whiskers: 1.5× interquartile range; single points: outliers. Source data are provided as a Source data file.

Fig. 3. The length polymorphism in exon 1 of MSH3 and its correlation within different genotypes with age at onset (AAO) in X-linked dystonia-parkinsonism (XDP) patients.

a cDNA and amino-acid sequences showing the wild-type sequence, the shortest (c.162_179del;199_207del), and the longest (c.181_189dup) allele of MSH3 exon 1, i.e., 6a, 3a, and 7a alleles, respectively. The sequences are based on the MSH3 transcript ID ENST00000265081.7 (https://www.ensembl.org). Blue shading highlights the region deleted in the shortest allele (stretch of six alanines (A) and three other amino acids (proline, alanine, proline—PAP)). Red shading indicates the region duplicated in the longest allele. M: methionine (start codon); S: serine, R: arginine, AFPPQLPPHI: amino-acid sequence from the length polymorphism to the end of MSH3 exon 1. b Difference between the observed and predicted (based on the hexanucleotide repeat number) AAO negatively correlates with the number of 7a alleles in a genotype. Box-plot elements: center line: median; box limits: upper and lower quartiles; whiskers: 1.5× interquartile range; points: outliers. c Disease onset is delayed by ~4 years in heterozygous carriers of the shortest allele (3a/6a) in comparison to the patients homozygous for the longest MSH3 allele (7a/7a). Red color designates a harmful effect (earlier AAO), while blue depicts a protective effect (later AAO). The effect estimates are based on the linear regression model. Source data are provided as a Source data file.