Rare Variant Burden Analysis of Dystonia Genes in Parkinson's Disease

Background: Dystonia frequently co-exists with Parkinson's disease (PD), yet the extent of genetic overlap remains insufficiently explored. Objective: To examine whether rare variants in dystonia-related genes are associated with PD or early-onset PD (EOPD). Methods: We curated 44 dystonia-related genes using OMIM and the updated Movement Disorder Society report on hereditary dystonia. Whole-genome sequencing data from 5,315 PD patients, including 300 with EOPD, and 36,902 controls across the Accelerating Medicines Partnership-PD and UK Biobank cohorts were analyzed. For each gene, we evaluated rare variants (minor allele frequency <1%) in four pre-specified variant classes: exonic, nonsynonymous, CADD score [&ge;]20 and loss-of-function. For the rare variant burden analysis, SKAT-O was performed, followed by meta-analysis with MetaSKAT. Results: In analyses of all PD cases, several genes showed nominal associations in meta-analysis: SQSTM1 (Ploss-of-function = 5.52 x 10-3), AOPEP (Pexonic = 6.96 x 10-3; Pnonsynonymous = 0.017), KCNA4 (Pexonic = 0.017), SPR (Pexonic = 0.029), SLC30A10 (PCADD[&ge;]20 = 0.046), and ACTB (Pexonic = 0.047). However, none remained significant after multiple-testing correction. In exploratory EOPD analyses, five genes reached significance after multiple test correction (ATP5MC3, DNAJC12, KMT2B, TBC1D24, TMEM151A). These signals were driven by small numbers of variants and were not robust to leave-one-variant-out analyses. GCH1 was nominally significant in the meta-analysis of EOPD (Pnonsynonymous = 4.36 x 10-3, PFDR = 0.062). Conclusions: Rare variants in dystonia-related genes do not appear to make a major contribution to PD risk overall. Signals observed in the EOPD subset were based on small numbers of variant carriers and require replication in larger cohorts.