Canine Genetics The tremendous phenotypic diversity of modern dog breeds represents the end point of a >15,000-year experiment. Each breed has undergone strong artificial selection in which dog fanciers selected for many traits including body size, fur color, skull shape, and even behavior, to create novel breeds. As a result, there is strong phenotypic homogeneity within breeds including breed-associated genetic diseases such as cancer. These traits are explored below (Ostrander et al., 2017). Canine Whole Genome Sequence Drs. Bob Wayne and Elaine Ostrander organized a June 2016 meeting in Beijing with18 other labs to establish a collaboration aimed at generating whole genome sequence (WGS) for 10,000 dogs in the next five years. We assigned participating labs to subcommittees (sampling, processing, QC, analyses, etc.) with a plan to meet again this year. Our goal is to sequence (30x) 10,000 unrelated dogs from each of 300 registered breeds from the U.S., Europe, and Asia. Dogs from non-breed working populations, as well as free breeding, semi-feral populations from East/Central Asia, South America, the Middle East, and Africa will also be sequenced. This effort is a partnership with the Chinese Academy of Sciences. The Ostrander lab has catalogued WGS sequence from over 700 canines thus far. Ostrander lab sequences are uploaded in SRA and variants are being uploaded to Broad-managed browsers. Canine Origins To better understand demographically based evaluation of genomic regions under selection in domestic dogs, we collaborated with several investigators (Freedman et al., 2016) to investigate positive selection on canids early in domestication. The study identified 349 outlier regions consistent with positive selection at a low false discovery rate, which was computed using an inferred demographic model. The strongest signals focused on genes important in brain function and behavior, an expanded the set of neurobehavioral candidate genes, and genes related to lipid metabolism. This type of methods development represents advances in our ability to study selection in human populations (Ostrander et al., 2017). Canine Population Structure Ostrander lab Staff scientist Heidi Parker undertook a large study of canine breed structure (Parker et al., 2017) in order to better inform our disease mapping studies (Davis and Ostrander, 2014). We assembled the most diverse dataset of dog breeds in existence reflecting extensive phenotypic variation and heritage. Combining genetic distance, migration, and genome-wide haplotype sharing analyses we uncovered geographic patterns of development and independent origins of common traits. Our analyses characterized the complexities of breed development, resolving longstanding questions regarding individual breed origination, the effect of migration on geographically distinct breeds and, by inference, transfer of trait and disease alleles among dog breeds. When applied to the traits of blindness and drug tolerance in a subset of breeds, we show we can track the history of variant alleles, offering predictions as to additional breeds where these diseases may appear next. This work is foundational to all studies in disease mapping in dog breeds, as it informs study design by determining how breeds should be partitioned or grouped for analyses. We also performed a study to provide guidelines regarding how breed-based studies can optimally use pedigrees, WGS, and SNP chip data to design the most informative studies (Dreger et al., 2016). Using data from 800 individuals, we show that successful breed mapping studies need to consider the unique demography of each population, including estimation of effective population size and timing of population bottlenecks. The study revealed that for 80 breeds, sequencing of 10 individuals captures most of the variation in a breed (Dreger et al., 2016). Mediterranean Dog Breeds The island inhabitants of Sardinia have long been a focus for studies of complex human traits due to their unique ancestral background and population isolation reflecting geographic and cultural restriction. We studied the genomic architecture of an endemic dog population on Sardinia, the Fonni's dog, which has not been subjected to an intensive system of artificial selection, but rather has developed alongside the human population of Sardinia. Incorporating data from other Mediterranean breeds we show how characteristics of human population isolates are reflected in dog breeds that have undergone artificial selection, and mirrored in the Fonni's Dog through traditional isolating factors that affect human populations. This sets the stage for studies of human migration carried out through genetic studies of regional dog populations. Canine Body Size Large (great Dane) and small (Chihuahua) dog breeds differ in size by nearly 40x, a claim no other land mammal can make. We previously identified six genes that account for about 50% of body size in small and medium dogs (Hoopes et al., 2012; Rimbault et al., 2013; Sutter et al., 2007). This highlights a recurring theme in dog genetics whereby a small number of genes of large effect control many complex phenotypes, as opposed to many genes of small effect, which is typical for humans (over 180 human body size loci identified), and indicating how much easier it will be to find morphology genes in dogs than in humans. This year, we iidentified variants in three new genes on the X chromosome that contribute to large size (Plassais et al., 2017), including IRS4 and IGSF1, both involved in the thyroid hormone pathway and associated with obesity and body mass index. We also identified variants in ACSL4, which control muscling and back-fat thickness. This study provides insight into how pathways controlling growth regulation interact, providing critical data for diseases of growth regulation, such as cancer. Dog Behavior and Human Syndromes In a study led by collaborator Bridget vonHoldt (vonHoldt et al., 2017) we were able to examine the molecular underpinnings of behavioral characteristics associated with domestication. We analyzed a 5Mb genomic region in the dog genome under positive selection in domestication. Deletion of this region is linked to Williams-Beuren syndrome (WBS) in humans, a multi-system congenital disorder characterized by hyper-social behavior. This study associated quantitative data on behavioral phenotypes symptomatic of WBS in humans with structural changes in the WBS locus in dogs, showing that hyper-sociability, a central feature of WBS, is also a core element of domestication that distinguishes dogs from wolves. Structural variants in the GTF2I and GTF2IRD1 genes, previously implicated in the WBS, contribute to extreme sociability in dogs. This finding expands the ever-growing knowledge base between variation in genes controlling behavior in dogs, which are also components of major human syndromes (vonHoldt et al., 2017). Canine Cancer Studies Our studies of histiocytic carcinoma advanced, with additional loci being mapped in the Flat Coated Retriever. This lethal disease is the only model for human non-Langerhans cell sarcomas, a highly lethal disease. We also undertook RNAseq of additional invasive bladder tumors, expanding our understanding of the role of BRAF mutations in tumor progression. Sottish terriers are at a 22-fold increased risk for invasive bladder cancer, offering an excellent naturally occurring model for a common human and canine cancer. Finally, we have done WGS on DNA from 22 dogs from our gastric cancer study, and are now analyzing sequence and SNP chip data between cases and controls in an effort to understand the susceptibility of Chow Chow and Belgian breeds to this lethal cancer (Davis and Ostrander, 2014).
Showing the most recent 10 out of 43 publications
