In Dog Genome, Scientists See Man's Best Hope

All dogs are descended from gray wolves, which were originally domesticated in East Asia. Some breeds, such as the Akita, have existed for more than 1,000 years. Most, though, are the product of selective breeding in the past 400 years to create specific characteristics.

That breeding has, in effect, concentrated specific versions of specific genes in specific populations of dogs. The result is a breed with physical and behavioral traits that existed in ancestral dogs but are now greatly magnified.

This results in animals that can look and act very different even though they scarcely differ from one another in their genetic identity. On a genetic level, breeds differ from one another only about as much as humans do. Gray wolves have more in common with Mexican hairless Chihuahuas than with coyotes, which they more closely resemble.

As traits have coalesced in breeds, so have specific diseases. That happened because genes involved in the diseases are physical neighbors of the trait genes; they move together in long stretches of DNA called haplotype blocks.

The blocks are 50 times as long in dogs as in people. This is because most breeds are only 30 to 90 generations old, much younger than human populations.

By studying the genomes of dogs of the same breed that have the same disease -- for example, a group of German shepherds with kidney cancer -- scientists can identify the genes responsible. Because the blocks are so large, researchers can narrow the possible location of disease genes to a few "neighborhoods" in the genome. They can then search in those stretches of DNA for the culprit genes.

"Breed-creation gives us a genome structure that makes it very easy to find disease genes," said Kerstin Lindblad-Toh, the lead author of the paper, which appears in the journal Nature. "We now have the tools; we have actually started cancer studies."

Similarly, scientists may be able to figure out what genes contribute to complex behaviors such as retrieving and pointing. They will look for haplotype blocks shared by different breeds with the same trait -- all retrievers or all pointers -- but that aren't found in breeds lacking the trait.

That work will be harder, Lindblad-Toh said. Nevertheless, it may provide insights into behavioral genetics, which historically have been the murkiest and most controversial part of the field.

Other less practical but equally interesting insights are already arising from the dog genome.

As with human DNA, only about 5 percent of dog DNA carries genetic information. The function, if any, of the rest (often termed "junk DNA") is unknown.

The exact number of genes in people, chimpanzees, mice, rats and dogs is uncertain, but it is now clear that of the 5 percent of genetic information in the genome, less than half that represents "classical" genes that encode the instructions to make proteins. The rest is DNA that regulates the activity of those genes.

The dog researchers identified 0.2 percent of the genome that comprises "highly conserved" DNA -- the stretches that are essentially identical in all individuals and across species -- but that do not code for proteins. They found that half of this favored DNA resides in the neighborhood of only 200 genes, some of which are involved in embryonic development and nerve growth. It appears to be crucial in the regulation of those genes.

"Which is to say, those genes must be very special. You just don't want to mess with them," Collins said. "This will undoubtedly cause other investigators to look at those 200 genes and ask why they are so important."