When humans began intensively breeding dogs in 19th century Great Britain, they created a cornucopia of canines seemingly out of whole cloth: hulking bullmastiffs, graceful golden retrievers, and pint-size Yorkshire terriers. But the real key to their success, a new study reveals, was taking advantage of two tiny but powerful genetic ratchets that have controlled the size of canines for ages—including one gene variant that nearly vanished in wolves around the end of the last ice age.
“It’s a great paper,” says Adam Boyko, an expert on canine genetics at Cornell University who was not involved with the work. The study, he says, doesn’t just question origin myths about modern dog breeding—it challenges a hypothesis about the very beginnings of canine domestication. Many scientists believe that early humans, via intensive early selection for smaller, more docile wolves, helped lock in a new genetic signature that made dogs smaller—and thus less threatening and hungry—than their gray wolf ancestors, he says. “This research suggests that we didn’t create it—it was already there.”
The new study traces its beginnings to 2007, when scientists led by Elaine Ostrander, a geneticist at the U.S. National Human Genome Research Institute, discovered a major genetic player in doggy dimensions. When the team scanned the DNA of thousands of dogs, it hit on a gene called insulin-like growth factor 1 (IGF1), which was responsible for 15% of the size difference between breeds. That may not sound like much, Ostrander says, but that’s a lot of work for a single gene to do. “In humans, the difference between being 5’6” and 6’6” is hundreds of genes.” It was unclear, however, exactly which genetic changes caused IGF1 to turn some dogs into giants and others into pipsqueaks.
In the new work, Ostrander’s postdoc, Jocelyn Plassais, now at the University of Rennes, widened the genetic net. He scanned nearly 1300 genomes from 230 modern breeds, indigenous and village dogs, and one dingo (considered more ancient than today’s dogs). IGF1 came up again as a key predictor of body size.
When Plassais zoomed in on this genetic region, he discovered the IGF1 gene itself was not the source of the variation, however. Rather, the key was a gene called IGF1-AS that lies next to and partially overlaps it, but does not code for any proteins. Unlike most RNA transcribed from genes, which is then used to make proteins, IGF1-AS’s RNA lines up “backward” with the RNA from the IGF1 gene, binding to it and altering how much IGF1 RNA is translated into growth hormone. “This affects how much of the hormone the dog’s body makes,” Plassais says, “but we still don’t know how.”
Most importantly, Plassais found two tiny gene variations—or alleles—in IGF1-AS with a huge impact on dog body size, he and colleagues report today in Current Biology. Three-quarters of dogs with two copies of the C allele were small breeds weighing less than 15 kilograms, such as pugs and Chihuahuas. Conversely, three-quarters of dogs with two copies of the T variant were large breeds averaging more than 25 kilograms, such as Irish wolfhounds and great danes. Medium-size dogs like border collies tended to have one C and one T.
To trace the evolutionary history of the alleles, Plassais and his colleagues scanned 33 ancient dog genomes dating back approximately 11,000 years. Even the oldest dogs had varying proportions of C and T alleles—which, as in modern dogs, matched their probable body sizes. The find reveals Victorian breeders did not introduce these genetic variants.
Indeed, when Plassais peered even further back, he found evidence of the T allele 53,000 years ago in a Pleistocene Siberian wolf, which had both C and T variants. But the team found nothing but Cs before that in other canids, including coyotes, jackals, and foxes. All of this suggests the C variant arose first and the T evolved relatively recently in arctic wolves, the researchers say, potentially as a way to help them bulk up and survive in their challenging ice age environment.
The big question is why C stuck around, not just in wolves, but in the dogs they eventually gave rise to. Ostrander suspects it’s because these wolves would sometimes have to adapt to changing environmental conditions—such as warmer temperatures and smaller prey—that made huge bodies disadvantageous. “The small allele is in nature’s back pocket until she needs it.”
Xiaoming Wang, a vertebrate paleontologist and expert on ancient canids at the Natural History Museum of Los Angeles County, agrees. “Animals don’t foresee what their descendants need,” he says. But variation often persists in creatures that carry one copy of each allele, providing “protection for the future.” It’s still important that researchers look beyond IGF1-AS, he adds, as there are likely other genetic variants that control canid body size.
Ostrander is on it. Her team has already uncovered 24 other genes that play a role in pumping up pups. “Who knows,” she says, “which of them may also have secrets hiding out in ancient DNA?”