Elinor K. Karlsson
Broad Institute and FAS Center for Systems Biology at Harvard University, Cambridge, MA, USA
Objectives of the Presentation
Describe genome-wide mapping methods and their application to domesticated species
Discuss how the unique history of a domesticated species is considered in the design of mapping studies.
Describe the development of the necessary tools, including SNP chips, in the domesticated dog.
Present recent work on successfully mapping traits, both single gene and multi-genic, in the domesticated dog
Discuss ongoing work to develop genome-wide mapping tools and resources for the domesticated cat.
Overview of the Issue
We are developing powerful new tools for mapping genes in dogs and cats that can search the entire genome (2-3 billion bases) for the genes underlying a disease using genome-wide association (GWA). In GWA studies, the gene underlying a trait or disease is mapped by examining the genome in individuals with the disease (cases) and those without it (controls), and identifying positions that significantly differ between the cases and controls. It requires a genome-wide mapping array, or "SNP chip," as well as blood samples from 10-100s of carefully phenotyped, unrelated cases and controls. Before developing the tools for mapping, we need to elucidate the history of each domesticated species and incorporate it into the design of mapping studies. Recent work in dogs shows that using this approach, we have exceptional power for mapping genes. The domestic cat has similar potential, and the tools for mapping are under development. In coming years, cat, along with horse, cow, chicken, and many other species, will join dog as a model organism for mapping the genes underlying diseases, morphologies and behaviors.
With the sequencing of genomes for human, mouse, dog, cat and dozens of other mammals, the search for the causes of heritable traits and diseases is changing radically. Based on the genome sequences, we are developing powerful new tools that can search the entire genome (2-3 billion bases) for the genes underlying a disease using genome-wide association (GWA). In GWA studies, the gene underlying a trait or disease is mapped by examining the genome in individuals with the disease (cases) and those without it (controls), and identifying positions that significantly differ between the cases and controls. It requires a genome-wide mapping array, or "SNP chip," as well as blood samples from 10-100s of carefully phenotyped, unrelated cases and controls. Here, I will review the development and application of SNP chips for mapping traits in dogs and cats, and discuss how the unique history of a domesticated species is considered in the design of mapping studies. I will describe recent successes mapping traits with both single gene and complex, multi-genic inheritance in the domesticated dog, one of the first fully sequenced mammals and exceptionally powerful species for mapping traits.[2,3] Finally, I will discuss ongoing work to develop similar tools and resources for the domesticated cat.
Click on the image to see a larger view.
History of Dog Breeds. Two population bottlenecks in dog population history, one old and one recent, shaped genome structure in modern dog breeds. First, the domestic dog diverged from wolves ~15,000 years ago, probably through multiple domestication events. Within the past few hundred years, modern dog breeds were created.
As one of the very first mammals to have its genome sequenced, the domesticated dog, Canis familiaris, occupies a special niche in genomics. The genetic history of dogs has been closely intertwined with our own for at least 15,000 years, since they were domesticated from the gray wolf (Figure 1). Dogs evolved in a mutually beneficial relationship with humans, sharing living space and food sources. In recent centuries, we selectively bred dogs that excel at herding, hunting, and obedience, and, in the process, created breeds rich in behaviors that both mimic our own and support our needs. Similarly, dogs were bred for desired physical characteristics such as size, skull shape, coat color and texture producing breeds with closely delineated morphologies. This evolutionary experiment produced the most diverse domestic species, but had unintended consequences on the health of purebred dogs. Many of the more than 400 modern dog breeds exhibit high prevalence of specific diseases, including cancers, autoimmune diseases, diabetes, blindness, heart disease, cataracts, epilepsy, hip dysplasia and deafness.[5,6] The very high rates of specific diseases within certain breeds points to a genetic cause, and suggests that a limited number of genes underlie each disease.
First Two Traits Mapped With SNP Chip in Dogs. Two traits were mapped as a proof of principle test of the canine SNP chip. (a) White coat color in boxers was mapped to the gene MITF, a developmental gene involved in pigmentary and auditory disorders, using 10 white dogs and 9 pigmented dogs. (b) The dorsal hair ridge in the Rhodesian ridgeback breed was mapped to a duplication of 4 genes, including three fibroblast growth factor genes, using 12 ridged dogs and 9 ridgeless dogs
The dog genome sequencing project, published in 2005, included a careful analysis of the genetics and history of dog breeds.[2,3] While the genome sequence itself is from a single purebred dog (a female Boxer), we also partially sequenced many individuals from nine additional breeds and four canid species. Using this data, we identified positions in the genome that varied between dogs (single nucleotide polymorphisms, or SNPs) and built a 2.5 million SNP map of the genome. We also examined the genetic diversity and evolutionary history of the domesticated dog population and within breeds. We found that the dog genome does indeed reflect two population bottlenecks, one early one, when dogs were domesticated from wolves, and a second, much more recent one, when breeds were created.
Based on this work, we developed a genome-wide SNP chip for gene mapping in dog breeds. Today, three versions are available from two companies: a 26,578 SNP Affymetrix array, a 49,663 SNP Affymetrix array and a 22,362 SNP Illumina array. A fourth, much denser Illumina array with ~150,000 SNPs is in production. In 2007, using the initial version of the Affymetrix chip, we used the canine SNP array to find the genes for two traits with mendelian inheritance: white coat color in boxers and bull terriers and the dorsal hair ridge in ridgeback breeds (Figure 2).[7,8] In each case, we needed just a few dozen dogs to search the genome and successful identify the underlying gene. We then used a two stage mapping strategy to narrow the region and identify the causative mutations (Figure 3). In both cases, the mutations were non-coding, not directly changing the protein made by the gene, and would have been extremely difficult to map by more traditional mapping methods.
Since we complete those initial studies, researchers have identified candidate genes for a range of dog traits and diseases. These include diseases with complex, multi-genic inheritance, which are much more difficult to map. For example, recent work mapping autoimmune diseases in the Nova Scotia Duck Tolling Retriever shows that with good sample sets and well described phenotypes, these genes can be found using just a few hundred individuals (compared to the thousands required for human studies).
Numerous other studies are underway across the canine genetics community. The American Kennel Club Canine Health Foundation and Morris Animal Foundation have funded mapping studies of cancers, autoimmune diseases, kidney diseases, liver diseases, epilepsy, other neurological diseases, orthopedic diseases, endocrine disorders, behavior disorders, eye diseases, gastrointestinal diseases and reproductive disorders. The European Union funded LUPA consortium aims to find genes responsible for at least 18 diseases, including four cancers, four inflammatory disorders, and three heart diseases in the next 4 years. To do so, they are collecting DNA, health information and GWA data for 8,000 dogs, thus creating a data resource with enormous future potential. As researchers find the suspected genetic causes for cancers, the newly established Canine Comparative Oncology & Genomics Consortium (CCOGC) aims to provide samples for functional studies. It recently announced funding for six institutions to begin collecting tumor and normal tissue for a shared repository focusing on cancers that have clear human relevance, including osteosarcoma, lymphoma and melanoma.
The domesticated cat, Felis catus, has a breed structure and history similar to dogs, and carries over ~200 documented genetic diseases. Currently, work is underway to build the tools required for WGA in cats. The current cat genome, released in 2007, is assembled from light sequencing of an Abyssinian cat, and captures just 65% of the genome. The Genome Center at Washington University in collaboration with the Cat Genetics Community is currently working to bring the cat genome up to a sequence density and quality comparable to dog. In addition, they are lightly sequencing individuals from multiple cat breeds. With this data, they will compile a SNP map for cats, a resource the feline research community will use to design a SNP chip with ~100,000 SNPs. They will also investigate the history of cat breeds in more detail, information that is critical for implementing genetic mapping studies in cats. For breeds that are similar in diversity and age to dog breeds, a similar number of cases and controls will be sufficient for mapping traits, and the work in dogs can be used as a guide. However, some cat breeds may have higher levels of diversity and more distant origins, like most horse breeds, and thus may require large sample sets and more SNPs for mapping.
The domestic dog is now a full-fledged genetic model ideally suited to trait mapping, and work in the domestic cat is well underway. In dogs, monogenic traits can be mapped with remarkable power using just a few dozen samples. With a few hundred samples, genes for complex traits such as autoimmune diseases can be found. Scientists have also begun developing cross-breed mapping strategies to find the loci controlling the remarkable phenotypic variation between breeds. Whatever the approach, mapping the associated genomic loci is just the first step-finding the genes and exact mutations responsible will be very challenging, especially for polygenic traits caused by regulatory mutations.
Click on the image to see a larger view.
The domestic cat has significant trait mapping potential, but awaits development of the necessary resources. The work in dogs clearly shows that by considering the unique population history and biology of each species, tools of exceptional power and efficiency can be designed. In coming years, cat, along with horse, cow, chicken, and many other species, will join dog as a model organism for mapping the genes underlying diseases, morphologies and behaviors.
The domestic dog is now a full-fledged genetic model ideally suited to trait mapping, and work in the domestic cat is well underway. In dogs, monogenic traits can be mapped with remarkable power using just a few dozen samples. With a few hundred samples, genes for complex traits such as autoimmune diseases can be found. The domestic cat has similar trait mapping potential, but awaits development of the necessary resources, including a denser coverage genome and a SNP map. The work in dogs clearly shows that by considering the unique population history and biology of each species, tools of exceptional power and efficiency can be designed. In coming years, cat, along with horse, cow, chicken, and many other species, will join dog as model organisms for mapping the genes underlying diseases, morphologies and behaviors.
1. Wade CM, Karlsson EK, Mikkelsen TS, Zody MC, Lindblad-Toh K (2006) The Dog Genome: Sequence, Evolution and Haplotype Structure. In: Ostrander EA, Giger U, Lindblad-Toh K, editors. The dog and its genome. Cold Spring Harbor, N.Y.: Cold Spring Harbor Laboratory Press. pp. xix, 584 p.
2. Lindblad-Toh K, Wade CM, Mikkelsen TS, Karlsson EK, Jaffe DB, et al. (2005) Genome sequence, comparative analysis and haplotype structure of the domestic dog. Nature 438: 803-819.
3. Karlsson EK, Lindblad-Toh K (2008) Leader of the pack: gene mapping in dogs and other model organisms. Nat Rev Genet 9: 713-725.
4. Sablin MV, Khlopachev GA (2002) The Earliest Ice Age Dogs: Evidence from Eliseevichi 1. Current Anthropology 43: 795-799.
5. Ostrander EA, Galibert F, Patterson DF (2000) Canine genetics comes of age. Trends Genet 16: 117-123.
6. Patterson D (2000) Companion animal medicine in the age of medical genetics. J Vet Intern Med 14: 1-9.
7. Karlsson EK, Baranowska I, Wade CM, Salmon Hillbertz NH, Zody MC, et al. (2007) Efficient mapping of mendelian traits in dogs through genome-wide association. Nat Genet 39: 1321-1328.
8. Salmon Hillbertz NH, Isaksson M, Karlsson EK, Hellmen E, Pielberg GR, et al. (2007) Duplication of FGF3, FGF4, FGF19 and ORAOV1 causes hair ridge and predisposition to dermoid sinus in Ridgeback dogs. Nat Genet 39: 1318-1320.