Genome wide association study of insect bite hypersensitivity in two populations of Icelandic horses

Merina Shrestha¹, Lisa S. Andersson¹, Freddy Fikse¹, Tomas Bergström¹, Anouk Schurink², Bart J. Ducro², Susanne Eriksson¹, Gabriella Lindgren¹

¹Swedish University of Agricultural Sciences, Department of Animal Breeding and Genetics, Uppsala, Sweden;

²Wageningen University, Animal Breeding and Genomics Centre, Wageningen, the Netherlands

Equine Insect Bite Hypersensitivity (IBH) is a chronic, pruritic, recurrent seasonal dermatitis. It is caused by an allergic reaction to protein in the saliva of biting midges of Culicoides spp and sometimes Simulium spp. It leads to discomfort and disfigurement that impairs the quality of life of the horse and economical loss for the horse owner. Severely affected horses are sometimes euthanized. The worldwide prevalence ranges from 3% to 60%. There are no effective preventive measures and cures for IBH. Environmental and genetic factors are responsible for its development. Heritability for IBH in Swedish born Icelandic horses is estimated to be around 0.10 (observable scale) and 0.33 (sd=0.19, liability scale); in Dutch Shetland breeding mares around 0.24 (sd=0.06) and 0.16 (sd=0.06, liability scale) for Friesian broodmares. A genome wide association study was performed using Illumina 50K SNP chip data of 209 Icelandic horses. GenABEL package in R was used in analyses fitting a single marker effect at a time.

A significant association was observed on chromosome 23. Odds ratio for IBH development of the unfavorable allele was 23.4 and the unfavorable allele had a frequency of 0.1 in cases and 0.004 in controls. This tentative association will be confirmed by re-genotyping the horses included in the study and also genotyping additional horses. Borderline associations were observed on chromosomes 3, 10, 17, 18 and 32. For further confirmation analysis using multi marker association model, based on a Bayesian variable selection method, will be done. A combined GWAS will also be performed on data of 146 and 209 Icelandic horses from the Netherlands and Sweden respectively, using Bayesian methodology