Dog genome exposed

Right off the presses: The dog officially joined the Age of the Genome in the guise of Tasha, an inbred boxer, representatives of nine other breeds—German shepherd, Rottweiler, Bedlington terrier, beagle, Labrador retriever, English shepherd, Italian greyhound, Alaskan malamute and Portuguese water dog -- grey wolves from China, Alaska, India and Spain, and a California coyote. Researchers sequenced virtually all—98 to 99 percent—of Tasha's genome, the 2.4 billion base pairs of DNA (A [adenine], which binds with T [thymine] and C [cytosine] which binds with G [guanine]) that form the genetic code for her in particular and the dog in general.

The goal, of course, is to use the genome to find genes responsible for the more than 450 genetic diseases in dogs, some 350 of them shared with people, and also to study the genetics of morphology and behavior. The genome should also provide a tool for exploring the genetic basis for the dog's evolution from the wolf—for domestication. Since much in the dog sorts by breed and since breeds retain a fair amount of genetic variability, despite high levels of inbreeding, they are ideal populations for such studies.

A detailed description and analysis of the sequence is published in this week's, December 8, issue of the science journal, Nature (subscription probably required). A press conference today in Boston, where members of the sequencing team from the Broad Institute at MIT and Harvard and Agencourt Bioscience, Corp., of Beverley, Massachusetts, led by Kerstin Lindblad-Toh, will doubtless generate many stories a well, but we're not linked to them because they're not out yet. Behind its wobbly editor-at-large The Bark is chugging along at the head of the pack. We'll link to the best stories later, and The Bark will carry a complete report in January. Here are some highlights.

Basic science:

With approximately 2.4-billion base pairs of DNA and 19,300 genes, the dog genome is slightly “smaller” and “cleaner” than the human genome with its 2.9-billion base pairs of DNA and approximately 22,000. The dog genome appears to have fewer duplicated elements and to have lost some sequences belonging to the “common mammal ancestor.”

About 72 percent of the genes are orthologous between dogs and humans—meaning they look the same, even if their function is different. These genes make up 1 to 2 percent of the genome and are part of a larger block covering approximately 5.3 percent of the genome. In addition to genes that block contains RNA, regulatory elements, structural elements, and “other stuff.” Researchers believe it is inherited in all mammals from the “common mammal ancestor” and is involved in development of neural networks and other fundamental processes and organs. It is also subject to intensive “negative selection,” meaning that mutations are weeded out.

--That said, the research team found tantalizing evidence that some sets of functional genes, like those involved in brain development, have evolved in dogs and humans in similar ways.

Dog stuff:

Researchers chose Tasha because she was the most inbred of 120 dogs from 60 breeds they sampled. Because her genome had less variability, it was easier to sequence. It then had to be assigned to its proper chromosomes—the dog has 39 pairs, including the sex, compared with 23 for humans.

To navigate this genetic code, the scientists used changes in a single base or letter, called “single nucleotide polymorphisms, SNPs (pronounced “snips”), which occur randomly throughout the genome—in genes they can cause lethal diseases.--to create a SNPs map. They drew SNPs from their nine breeds, the wolves, the coyote, and the poodle, Shadow, the majority of whose genome was sequenced several years ago by a team working for genome entrepreneur Craig Venter.

In vast non-coding portions of the genome, SNPs are used to identify sections of code, called “haplotypes,” that are used to study inheritance, disease, and in the dog, morphology and behavior.

Here's what they found:

--The dog has passed through two major genetic bottlenecks. The first was at its origin some 9,000 generations ago from perhaps as few as two wolves. It could be more; it could have happened multiple places, Lindblad-Toh said, which leaves multiple questions unanswered. Her team calculates a generation at 3 years, putting the origin of the dog at 27,000 years ago. Three years seems wrong to me, but I've been unable to find a generation time for dogs and welcome suggestions, either here or at [email protected].

Hans Ellegren, an ecologist at Uppsala University, Sweden, in a commentary on the genome in the same Nature, suggests that Lindblad-Toh and her colleagues failed to account for what might have been frequent back-crosses to wolves over the dog's long history. If they had allowed for them, their date might have been different, he says.

Archaeologists object because there are no dogs in the fossil record at 27,000 years.

Clearly, the question of the origin of the dog is far from settled.

--The dog passed through another bottleneck 30 to 90 generations ago with the creation of modern breeds through inbreeding and the overuse of favored sires. At that time tremendous pressure was placed on just a few, overlapping chromosomes, carrying the genes underlying the desired traits, as well as those nasty diseases.

--As a result of this recent bottleneck, most purebred dogs, like Tasha, are a mix of homogeneity and heterogeneity at a ration of 62 percent to 38 percent, respectively.

--In testing out their SNPs map, they scanned the genomes of 20 dogs from breeds and one dog each from 24 breeds, from the Glen of Imaal terrier to the Irish wolfhound. While the overall pattern remained nearly the same and while there was sharing of haplotypes among different breeds, there are also differences between breeds in terms of which haplotypes appear in the genome and how frequently they do so.  Of a total of 7 to 10,  4 are most commonly found in each breed.

--This structure will help find those genes, especially once a promised SNP chip becomes available to researchers and vets.

These are Western breeds, of course, and more work needs to be done with dogs from elsewhere and our own curs and feists and purpose bred dogs. Indeed, the completed sequence raises more questions than it answers, as it should -- Mark Derr