Identification of common variants influencing risk of the three-repeat tauopathy Pick’s disease: a genome wide association study

Pick's disease (PiD) is a rare cause of sporadic frontotemporal dementia, neuropathologically defined by the presence of Pick bodies consisting of aggregates of 3-repeat tau. Given the genetic aetiology of PiD remains unresolved, we assembled the Pick's disease International Consortium (PIC) to identify susceptibility loci through a genome-wide association study (GWAS). A GWAS was conducted in 294 autopsy confirmed PiD cases and 1,055 controls. Lead variants were annotated using the Functional Mapping and Annotation of GWAS (FUMA) platform, followed by co-localisation analyses using the METABRAIN dataset and statistical finemapping using FINEMAP and SuSiE. After exclusion of 3 cases of MAPT mutations, no variants were associated with risk of PiD at genome-wide significance (<i>p</i> < 5 × 10^-8). The strongest association was on chromosome 4 (4p13, lead SNP rs112161979, OR = 7.53, 95% Confidence Interval (CI) = 3.62-15.65, <i>p</i> = 6.37 × 10^-8) followed by chromosome 11 (11p15.4, lead SNP rs66481907, OR = 2.10, 95% CI = 1.54-2.84, <i>p</i> = 1.83 × 10^-6). rs112161979 is an intronic SNP in the <i>KCDT8</i> gene, encoding a potassium channel tetramerization domain that acts as an auxiliary subunit for GABA_B receptors, whilst rs66481907 is an intronic SNP in <i>TRIM22</i>, encoding an E3 ubiquitin-protein ligase. Our GWAS provides the first evidence of possible genetic risk for PiD that implicate the modulation of GABA_B receptor signalling and inflammation, in disease pathogenesis. Replication of these findings will be important, but our results suggest that, if present, the genetic risk of PiD beyond MAPT mutations is low.