Idiopathic epilepsy in the Border Collie: using genome-wide association study and whole genome sequencing approaches to identify genetic risk factors

Presenter Sally Ricketts
Authors Christopher A. Jenkins (1,2), Luisa De Risio (3), Tiina Heinonen (4), Jim Johnson (1), Lorna J. Kennedy (2), Andrea D. Short (2), Sofie F. M. Bhatti (5), Donna Foster (1), Peter Leegwater (6), Paul Mandigers (6), Tarja S. Jokinen (4), Rowena Packer (7), Koen Santifort (8), Holger Volk (9), Cathryn S. Mellersh (1), Marjo Hytönen (4), Hannes Lohi (4)*, Sally L. Ricketts (1,2)*
Affiliations 1. Kennel Club Genetics Centre, Department of Veterinary Medicine, University of Cambridge, UK; 2. Division of Population Health, Health Services Research & Primary Care, University of Manchester, Manchester, UK; 3. Linnaeus Veterinary Ltd, UK; 4. University of Helsinki, Finland; 5. Ghent University, Faculty of Veterinary Medicine, Belgium; 6. University of Utrecht, The Netherlands; 7. Royal Veterinary College, UK; 8. Evidensia Small Animal Hospital Arnhem, Arnhem, The Netherlands; 9. Department of Small Animal Medicine and Surgery, University of Veterinary Medicine, Hannover, Germany
Presentation Type Poster


Idiopathic epilepsy (IE) is a chronic and life-limiting neurological disorder characterised by recurrent seizures. The Border Collie (BC) is one of the most commonly and severely affected breeds, though little is known regarding the genetic factors that increase this breed’s susceptibility to IE. This large-scale collaborative project aims to identify genetic variants that are consistently associated with IE in the BC. We conducted a genome-wide association study meta-analysis utilising three study sets genotyped using the Illumina canineHD array – two Kennel Club Genetics Centre (KCGC) sets and one from the University of Helsinki. We define cases as BC diagnosed with IE with at least a tier I level of confidence. Controls are BC over the age of 8 years that have been reported by owners to have never had a seizure. After genotyping a subset of the KCGC sets on the Axiom array, we imputed genotypes to increase the resolution of all three datasets. The analysis included 271 BC (104 cases and 167 controls) and 291,450 SNPs. We subsequently imputed genotypes to whole genome level in an additional meta-analysis of 123 cases and 187 controls and 5,993,223 SNPs. We took forward the top tier of 29 SNP associations for replication testing in an independent replication set of BC comprising 288 IE cases and 372 controls. In a separate analysis, we searched for rare, common and fixed risk variants directly by conducting whole genome sequencing (WGS) of 10 BC IE cases and five controls, and compared WGS of cases to 59 dogs of 35 different breeds not known to have a high prevalence of epilepsy from literature searches and breed health reports. Following filtering for candidacy using human gene lists and subsequent follow-up in our extended case-control set, we identified variants that are both common, and rare, in cases. It is likely that IE in the BC is a complex disease, with a polygenic set of risk factors predisposing to greater risk of IE. Improved understanding of the genetic architecture and biological pathways underlying IE in the BC could lead to better treatments and prevention of the disease.