Oligogenic risk model for Gilles de la Tourette syndrome reveals a genetic continuum of tic disorders

Abstract

Gilles de la Tourette syndrome (GTS) and other Tic Disorders (TDs) have a substantial genetic component with their heritability estimated at between 60 and 80%. Here we propose an oligogenic risk model of TDs using whole-genome sequencing (WGS) data from a group of Polish GTS patients, their families, and control samples (n = 278). The model is based on the overrepresentation of coding and non-coding genetic variants in and in the vicinity of genes selected from a set of 84 genes previously indicated as putatively associated with GTS. In the discovery phase, based on a variant burden test between unrelated GTS cases (n = 37) and a database of local allele frequencies 10 genes were selected for the model (CHADL, DRD2, MAOA, PCDH10, HTR2A, SLITRK5, SORCS3, KCNQ5, CDH9, and CHD8). Variants in these genes (n = 7654) with a median minor allele frequency in the non-Finnish European population of 0.02 were integrated into an additive classifier. This risk model was then applied to healthy and GTS-affected individuals from 23 families and 100 unrelated healthy samples from the Polish population (AUC-ROC=0.62, p=0.02). Application of the oligogenic model to a group of patients with other tic disorders revealed a continuous increase of the oligogenic score with healthy individuals with the lowest mean, then patients with other tic disorders, then GTS patients, and finally with severe GTS cases with the highest oligogenic score. Results were also overlapped with Psychiatric Genomics Consortium (PGC) GWAS data and we found no significant overlap between the common variant signal and our oligogenic model (p=0.21). Therefore obtained results were compared with the polygenic risk score built from the PGC GWAS data, which revealed a significant contribution of common variant background in severe GTS cases. Overall, we leveraged WGS data to construct a GTS/TDs risk model based on variants that may cooperatively contribute to the etiology of these disorders. This study provides evidence that typical and severe adult GTS as well as other tic disorders may exist on a single spectrum in terms of their genetic background.