P0653 Analysis of Normal and Disease Related Genomic Copy Number Variation in Horses

Sharmila Ghosh , Texas A&M University, College Station, TX
Zhipeng Qu , University of Adelaide, Adelaide, SA, Australia
Pranab Jyoti Das , Texas A&M University, College Station, TX
Rytis Juras , Texas A&M University, College Station, TX
Vincent Gerber , 4Equine Clinic, Department of Clinical Veterinary Medicine, Vetsuisse Faculty, University of Bern, Bern, Switzerland
David L. Adelson , University of Adelaide, Adelaide, SA, Australia
Ernest Gus Cothran , Texas A&M University, College Station, TX
Bhanu P. Chowdhary , Texas A&M University, College Station, TX
Terje Raudsepp , Texas A&M University, College Station, TX
Copy number variations (CNVs) constitute 5-12% of the mammalian genomes contributing to normal phenotypic variation and affecting biomedical conditions. Here we used an equine whole genome 400K tiling array to investigate naturally occurring CNVs in diverse horse breeds. The CNVs were studied by array CGH in 13 breeds, two individuals each. Using conservative calling criteria ≥0.5 log2 ratio over 5 probes, 202 CNVs were identified with the highest number present in ponies. Many CNVs were shared yielding in 88 CNV regions, while others were breed-specific. ECA1, 3, 9, 10, 20 and Unknown had the highest and ECA21 and Y the lowest number of CNVs. There were more losses than gains, the ratio of genic and non-genic CNVs was 3:2, and the CNV size ranged between 408 bp-2,556 kb with an average 126 kb. The CNVs involved 718 annotated genes with diverse biological functions. These data form a foundation for the discovery of CNVs associated with complex genetic disorders. Thus, a pilot CNV study was conducted in recurrent airway obstruction (RAO) using affected (n=4) and control (n=4) Swiss Warmblood horses. At ≥0.2 log2 ratio over 3 probes, 319 CNVs were discovered: 137 in controls, 49 in cases, and 133 in both. Losses prevailed over gains in RAO affected horses, with the most functionally relevant losses in CBR3 and CSMD1 gene clusters known to be involved in metabolic responses to drugs and immunity pathways. These preliminary findings indicate that CNV studies might facilitate the identification of genomic regions contributing to complex disease phenotypes.