GNN - Genome News Network  
  Home | About | Topics
Mapping the Dog
What the human genome sequence can do for dogs. And vice versa.
Edward R. Winstead

Featured article.

We are not so different, dogs and people. We love treats and hugs and naps; we feel better after playing outdoors; and we stick our noses where they do not belong.

Tess, a Border Collie, goes to work with Elaine Ostrander. Courtesy Elaine Ostrander.
Tess, a Border Collie, goes to work with Elaine Ostrander.

We are not so different when it comes to genes either. The dog genome is basically the human genome divided into about 70 different pieces and rearranged on a greater number of chromosomes, according to a new map of the dog genome. While making the map, the researchers lined up each chromosome segment in the dog with its corresponding region of the human genome.

"The new map represents a real triumph," says Elaine Ostrander, of the Fred Hutchinson Cancer Research Center in Seattle, Washington, who heads one of three groups behind the Dog Genome Project. "The genomes of humans and dogs are very, very similar, and any disease that humans get, dogs get also. Now we have a map with nearly complete coverage of the canine genome, and this will allow us to take advantage of the dog as a model system."

Matthew Breen at the Animal Health Trust in the U.K. and Francis Galibert at the University of Rennes in France head the other two groups in the genome project. Since the project was launched in the mid-nineties, each group has used different strategies to identify landmarks on dog chromosomes. In the last three years, these landmarks were assembled onto a single map. It has over 1,800 DNA markers, including 320 dog genes.

‘Our goal is to help dogs. But we are learning things that will help humans.’

For the first time researchers have a map that covers 90 percent of the dog genome and has landmarks on every chromosome. This means that hunting for genes in the dog can be done with more precision because the regions containing genes of interest will be smaller. Dog breeders, among others, have provided financial support for the new map in the hope that it will lead to new therapies and tests to reduce genetic disease in the dog.

More than 350 genetic diseases have been described, and half of these have known counterparts in humans. Cancer, blindness, deafness, and congenital heart disease, for example. A few dozen, such as muscular dystrophy and the bleeding disorder von Willebrand's disease, involve the same gene in both humans and dogs.

"The value and the power of the dog system is that it allows us to dig much deeper than we can in humans," says Ostrander. Human populations are genetically diverse, and identifying disease genes can be like searching for a needle in a haystack. By contrast, there are 300 pure breeds of dog, each of which constitutes a genetically homogenous population. Researchers have access to multiple generations of a dog family and can do non-invasive studies using DNA from cheek swabs. Pinpointing disease genes in these families is relatively straightforward.

Shadow next to a DNA analyzer.

Dog researchers, of course, would like to have the complete canine genome sequence, and one day they will get their wish. Preliminary work on the dog genome has been done at Celera Genomics in Rockville, Maryland. After sequencing the genomes of five humans and several mice last year, Celera scientists have begun to analyze the DNA of a seven-year-old Standard Poodle named Shadow and have completed an initial sequence.

Shadow's genome project could take years, but for now researchers have the map of the dog genome. "The map is a resource that researchers can use to undertake genome-wide scans for genes of interest with a high degree of confidence," says Ostrander. It will allow investigators to study "the nitty-gritty of how gene families are arranged in the dog." The map was published in Genome Research.

As part of a long-term project to characterize gene families in the dog, Keith E. Murphy, of the College of Veterinary Medicine at Texas A&M University, and colleagues have been studying the keratin gene family. The keratins are structural proteins found in hair, nails and skin cells, and mutations in these genes cause skin diseases in humans.

"Several diseases in the dog look similar to keratin-associated diseases in humans, and we think they may involve the same genes," says Murphy. His team recently sequenced a keratin gene, KRT2p, and mapped it to chromosome 27 of the dog. The study appeared in Functional and Integrative Genomics.

"We dog researchers are not working in the dark," says Murphy. "The great progress in human genetics means that if you have a disease that presents the same way in humans and dogs, you can almost go straight to the gene in the dog."

The DNA sequences for thousands of genes in hundreds of species—everything from humans to jellyfish to tomatoes—are available in scientific databases. It takes seconds to search the database for a match to a dog DNA sequence. Then the real work begins.

"The whole purpose of having the genome sequence of any species is to help solve questions related to biology," says Robert Dunstan, an expert on skin diseases in the dog and a colleague of Murphy's at Texas A&M University.

Detail of multicolor fluorescence in situ hybridization of canine clones to dog chromosome 5. View larger

"Our first goal is to help the dog," he continues, "but we are increasingly finding that by studying skin diseases in the dog we are learning things that improve our understanding of the pathogenesis of disease. And that will help humans."

Dunstan used to apply for grants to support research on dogs by saying that the dog is a great model for studying human disease. Now, he uses a different tactic. Investigating diseases in the dog, he argues, is valuable because the dog is a different species with its own genes: "We study them not because they're similar to humans but because they're different from humans."

Hair growth is one example. Humans need regular haircuts and most dogs do not; and dog hair does not fall out during chemotherapy while human hair does. The researchers are investigating hair genes in the dog as a way to help answer some fundamental questions about hair in humans.

When researchers want to find a gene in dogs—any gene—they go to an online database of all publicly available DNA sequences and get the sequence of a human gene and the sequence of the corresponding mouse gene. They align the two genes, and the overlapping regions are the basis for a 'probe' that can fish out a dog version of the two genes, if one exists.

"This strategy has allowed us to play catch up very quickly in a molecular sense and become a player in cell biology," says Dunstan. "In a very true sense, we can piggyback all existing knowledge in humans and in mice."

Riley, a Jack Russell Terrier with ichthyosis.

Dunstan and a colleague, Kelly M. Credille, founded the Comparative Dermatology Laboratory at Texas A&M University to diagnose skin diseases in pets by looking at biopsies under the microscope. The laboratory's mascot is Riley, a Jack Russell Terrier who was born with a skin disease, ichthyosis. A similar disease occurs with less frequency in humans.

"These types of skin diseases in humans are rare," says Credille. "But there are scientific papers on them every month because, by studying them, you learn so much about how the skin acts normally."

Riley's disease does not appear to be caused by the same defect as the human disease. Credille screened the dog version of the human gene, but did not find any mutations. "I was really disappointed," she says, "but I'm looking at this gene in other affected dogs to be sure I haven't missed a mutation."

Although the comparative approach can speed the discovery of homologous genes between species, it will not work for diseases that are specific to the dog. This is where the map becomes even more valuable. Mapping genes to chromosomes is a little like finding a house on a street when you have the name of the street but not the address of the house. The more familiar landmarks there are, the better.

Dunstan and Credille are collaborating with Mathew Breen and others to localize genes involved in Sebaceous Adenitis, a hereditary skin disease that occurs primarily in Standard Poodles but also in other breeds. "Nothing like this disease has been reported in humans, but we hope to get a better understanding of acne in humans by studying it," says Dunstan.

"We study dogs to help dogs," he adds, "but this research helps humans and, we think, also biomedical science generally. This, in turn, comes back to help dogs because it's all part of one big circle."

The dog is currently helping researchers annotate the human genome, which, again, will benefit canine research. Identifying human genes and the DNA sequences that regulate genes is a challenge because they comprise a tiny portion of the genome. Detecting elusive sequences can be done efficiently by comparing the content of mammalian genomes, according to a new study of human chromosome 21, which appeared in the same issue of Genome Research as the dog map.

Detail from schematic of detecting evolutionarily conserved sequences on human chromosome 21. View larger

Kelly A. Frazer, of Perlegen Sciences in Santa Clara, California, and colleagues screened human chromosome 21 against the corresponding regions in dogs and mice. They identified a high degree of similarity between DNA sequences shared by humans and dogs and those shared by humans and mice. Even Frazer was surprised. "I was amazed by how well the patterns of conserved sequences among the three species overlaid each other," she recalls.

Frazer's team used high-density human microarrays, which are wafers that contain millions of chemical bases of DNA, or nucleotides. The sequences on the wafers bind corresponding sequences in the cells of other species, which, in effect, pulls out the common sequences.

The new study is the first to identify conserved sequences among the three species. The dog was included because of public interest in the animal but also because the researchers needed a third mammal to increase the statistical power of the analysis. On the evolutionary tree, the dog is somewhere between humans and mice. Sequences that are conserved in all three species are likely to be in the human genome for a reason.

The dog map will benefit from the new study by getting more DNA markers. And because Frazer and her colleagues succeeded in identifying new elements in the human genome, they have undertaken similar studies using five mammalian species.

"We've looked at conserved regions in six different species, and some are conserved only between humans and dogs and cats, while others are only conserved between cats and cows—and some are present in all species," says Frazer. There is no way to know which species will be most helpful for annotating different parts of the human genome, which is why, she adds, scientists need high-throughput screening methods.

Retford, who is featured on the cover of Genome Research, is one of many Bernese Mountain dogs that have died from a form of cancer that appears to be familial.

No single species will be perfect for annotating the human genome, says Frazer. Genomes evolve at different rates, and various mammalian species will be needed to pick up human genes of interest. The same gene might be present in both humans and mice, but in mice it may have mutated or evolved to the point that its DNA sequence no longer closely matches the human version. The statistical tools researchers use to find matches across species would not recognize the mouse gene.

"The work in human and mouse genomics is a breakthrough for all mammalian species, including dog," says Frazer.

. . .

Breen, M. et al. Chromosome-specific single-locus FISH probes allow anchorage of an 1800-marker integrated radiation-hybrid/linkage map of the domestic dog genome to all chromosomes. Genome Res 11, 1784-1795 (October 2001).
Frazer, K.A. et al. Evolutionarily conserved sequences on human chromosome 21. Genome Res 11, 1651-1659 (October 2001).
Miller, A.B. et al. Cloning, sequence analysis and radiation hybrid mapping of a mammalian KRT2p gene. Funct Integr Genomics 1, 305-311 (2001). Published online July 5, 2001.
Ostrander, E.A. et al. Canine genetics comes of age. Trends Genet 16, 117-124 (March 2000).

Back to GNN Home Page