GWAS in Labrador retrievers identifies novel obesity genes in dogs and humans.

Presenter Natalie Wallis
Authors Natalie J. Wallis (1, 2), Alyce McClellan (2), Justine Chan (2), Sambhavi S. Kumar (2), Ellen Schofield (3), Jacek Mokrosinski (1), Alexander Mörseburg (1), Katherine Kentistou (4), John Perry (1, 4), Giles Yeo (1), Rebecca Bounds (1), Edson Mendes de Oliveira (1), Sadaf Farooqi (1), Stephen O'Rahilly (1), Eleanor Raffan (1, 2)
Affiliations 1. Wellcome-MRC Institute of Metabolic Science, Cambridge, UK, 2. University of Cambridge Department of Physiology Development and Neuroscience, Cambridge, UK, 3. University of Cambridge Department of Veterinary Medicine, Cambridge, UK, 4. MRC Epidemiology Unit, Cambridge, UK
Presentation Type Talk


Obesity is an increasingly prevalent and complex disorder which poses a serious threat to canine and human health. The heritability of obesity is estimated at 40-70%, but the majority of genetic variants responsible are yet to be uncovered. In dogs, selective breeding and population bottlenecks simplify gene mapping for complex disease. Despite this, the genetic basis of canine obesity is poorly understood. Therefore, we investigate the genetic basis of obesity in a population of Labrador retrievers, an obesity prone breed. Similarities between canine and human obesity means genetic associations in dogs can prioritise and identify candidate genes from human genomic studies.

A genome wide association study for obesity in Labrador retrievers identifies multiple obesity associated loci and allows for construction of a canine polygenic score method. Genetic scores for obesity predict phenotypes in related but not unrelated breeds. Polygenic scores also provide insight into known within-breed variation in obesity risk, explaining associations with coat colour. We demonstrate how polygenic background influences penetrance of a well characterised POMC mutation that affects obesity risk by altering hypothalamic leptin-melanocortin signalling. Utilising large scale human cohorts, we find that canine obesity loci of large effect can serve to highlight human loci of small effect, including genes CARD11 and DENND1B. We also identify novel associations between CDH8, CSNK1A1 and DENND1B for monogenic cases of severe, early onset obesity in humans and pursue these carriers through familial investigation and deep clinical phenotyping. Molecular studies reported in an accompanying abstract delineate a role for DENND1B in modifying neural control of energy homeostasis. Further, selective sweep mapping highlights regions under selection in Labrador retrievers which harbour known obesity genes including MC4R.

Conclusion: We have identified novel obesity-related genes in humans and dogs. This demonstrates the benefits of studying complex disease in non-traditional animal models such as the dog, with implications for both human and veterinary medicine.