FIDO loves to cuddle on the sofa, Spot’s a cheeky hound with a good eye for an unguarded slice of roast lamb, and Rover is just the best, most caring babysitter. As dog owners and lovers everywhere will tell you, pooches have personality. Some of that – quite a bit of it, in fact – is down to breeds. Terriers are notorious for their boundless energy. Border collies like to chase things. Pointers point. German shepherds and poodles learn quickly, bulldogs and beagles not so much.
All this makes dogs more than just great pets. They’re also ideal subjects for geneticists eager to understand one of the great mysteries still remaining in biology: how complex behaviours are encoded in genes. If scientists can crack the code for dogs, we may be able to understand – and perhaps better manage – not only our loyal companions’ behaviour, but our own as well.
Geneticists have known for decades that dogs are perfect for gene prospecting. The comparison may seem a touch unkind, but dog breeds aren’t unlike strains of lab rats: each one is essentially an inbred line reared over generations to be nearly identical clones of each other. That makes both types of animal well suited to studies seeking to understand which gene does what. If you can identify subtle genetic differences between almost identical individuals, you can try to link those to variations in physical appearance, health or behaviour. It’s a far cry from us humans, who tend to be very outbred and unsuited to genetic studies, which is why human genetic studies often focus on small, isolated populations such as Icelanders or the Amish.
Advertisement
Dog geneticists knew their day had come when new genetic tools came online in the early 2000s that made scanning entire genomes faster. This opened up the potential for revolutionary genetic searches called genome-wide association studies. The concept is simple: you compare the genomes of dogs with a particular trait – say, compulsive licking – to the genomes of dogs without it, and look for regions of DNA where the two groups differ. If a gene is involved in the trait, it will lie somewhere within these regions.
Back in 2008, Elaine Ostrander, a geneticist at the US National Human Genome Research Institute, and her colleagues did this with 148 dog breeds. The team searched for bits of the genome linked to breed-specific physical traits, and located several genetic variants that code for body shapes and sizes. They also included a few crude behavioural measures in their study: does the breed herd? Does it point? Is it highly trainable or resistant to teaching? Is it bold or timid?
The work kicked up . A number of genetic variants seemed to show up only with specific behaviours, indicating those behaviours were down to nature, not just nurture. For instance, one particular bit of DNA on chromosome 8 tended to be present only if the breed was a pointing one. This suggests that dogs that point are genetically programmed to do so.
Pointing wasn’t the only behaviour for which the group found genetic signals. There were also markers associated with breeds that are skilled herders, ones that are easy to train and others that are particularly bold. The study showed that some behaviours could be traced back to genetics, but it was limited by its use of genetic markers known as SNPs. These are scattered more or less evenly across the genome, so finding one that is linked to a specific behaviour only tells you that a gene involved in the behaviour must lie nearby. Without being able to single the genes out, that was as far as the early studies could go.
That barrier has now fallen with the advent of faster, cheaper genome sequencing. Last year, a team led by Carlos Alvarez at Nationwide Children’s Hospital in Columbus, Ohio, took pooch genetics one step further. The team completed by dog owners about their pets’ behaviour, especially fear and aggression. They selected the chilled-out breeds and compared their genomes to those of more aggressive breeds. Sure enough, they found several gene variants that increase the risk of fearful or aggressive behaviours. Two of those are active in the amygdala, a key brain region for processing fear.
Alvarez also found an explanation for the old cliché that small dogs are yappy and have more bite. Breeds like Dachshunds and Yorkshire terriers carry a version of at least one gene, IGF1, that makes them more aggressive towards their owners and other familiar dogs, and contributes to their diminutive stature. The questionnaire showed that small dogs carrying this variant of IGF1 weren’t more aggressive towards strangers, which suggests that it isn’t just a case of them feeling more vulnerable. IGF1 performs a range of physiological functions throughout the body, and Alvarez speculates that it may affect both size and behaviour, perhaps through different pathways.
The findings may have implications that go far beyond the vague annoyance of snappish lapdogs. “The genetics and physiology of fear and aggression are very ancient,” says Alvarez, meaning they are shared by many different animals – including humans. If the same genetic pathways can be muted with drugs in humans, then snappy Spot’s purebred lineage could end up shedding light on his owner’s unfortunate tendency to fly off the handle.
Those conclusions are a bit of a stretch for now. Genes could have a big or a small role in human aggression – we don’t yet know. Other “nurture” factors like upbringing and nutrition no doubt also play a part. But comparisons are already being drawn between some dog behaviours and human obsessive compulsive disorder.
“If you saw a dog with a compulsive disorder for 5 minutes, you wouldn’t know there was anything wrong,” says Elinor Karlsson, a geneticist at the University of Massachusetts Medical School. “But if you lived with it, you’d see that it did the same behaviour over and over for hours.” Think hour upon hour of tail-chasing – bull terriers and German shepherds are particularly prone to this. Dobermans, meanwhile, compulsively suck their flanks and lick their paws. It can be quite distressing. A quick search online will kick up countless pictures of Dobermans that have licked themselves to the point of bleeding.
Dogs and their masters
Karlsson and her colleagues have identified four gene variants that occur far more often in dogs with compulsive behaviour than in their unaffected peers. All four play a part in communication between nerve cells, which makes sense for a trait involving behaviour. . And at least one drug that treats the human condition, an antidepressant called clomipramine, also works in dogs – evidence that the shared genetic underpinnings can be clinically important.
That said, Karlsson warns that genes don’t tell the whole story. “It’s not like coat colour, where you get the gene and you get the colour,” she says. Even dogs with the OCD-prone gene variants are only slightly more likely to develop compulsive disorders, so other factors are important as well.
Some of the unexplained variation in compulsive behaviour is probably the result of other genes whose importance isn’t yet recognised. In 2016, a team led by Nick Dodman, a veterinary behaviourist at Tufts University in North Grafton, Massachusetts, dug a little deeper by comparing Dobermans with severe compulsive behaviour to others with milder forms. They found several other genes – some related to the neurotransmitter serotonin and others involved in stress pathways – that appear to modify the intensity of the behaviour.
The genetic approach is beginning to pay off even in more complex social behaviours. For example, Per Jensen and his colleagues at Linköping University in Sweden studied the willingness of dogs to solicit help from people to solve difficult problems. The team gave a group of beagles an impossible problem: a box containing a food treat, but with a lid that the dog couldn’t open. Faced with this, most dogs make eye contact with a nearby human. Their closest wild relatives, wolves, show no such tendency, which suggests that this willingness to interact with humans may lie at the heart of domestication. Jensen’s team measured how quickly each dog turned to a human, how close it approached and how often it solicited help, and then scanned their DNA for associated genes.
The experiment pinpointed variants of five genes that were linked to how vigorously the dogs sought help from humans. Changes in three of the five genes are linked to autism in humans. This suggests that the propensity of Jensen’s beagles to interact with humans could involve the same genes and pathways that can incline people with autism to avoid human interactions. “If that were true – and obviously, we have to look at this in other breeds and other contexts – then I think dogs may prove to be interesting models for autism disorders in humans,” says Jensen.
Behaviours such as these are just the beginning. Dogs have another enormous advantage for behavioural biologists. “There are millions of pet dogs in the world, and they live with humans who spend a lot of time watching their behaviours,” says Karlsson. So the pets don’t need to undergo hours, weeks or months of tedious and costly observations.
Sequence your pet
At least two major research efforts are now beginning to tap into all that knowledge. Karlsson is spearheading one, known as Darwin’s Dogs. Dog owners – more than 11,000 to date – volunteer to fill out detailed behavioural questionnaires on their pets, and some also provide genetic samples. A second programme, headed by Adam Boyko at Cornell University in Ithaca, New York, offers what is essentially a canine version of the personal-genomics service 23andMe. Dog owners pay Boyko’s company, Embark, to analyse their pet’s genome. They get a glimpse of its ancestry and genetic risk factors; Boyko gets the dog’s genetic profile. In late 2016, Embark teamed up with Dognition, a company that uses games to measure dogs’ cognitive skills. By combining cognitive and genetic profiles, Boyko and Dognition’s Brian Hare plan to work out the genes behind intelligence, trainability and sociability.
Both Darwin’s Dogs and the Embark/Dognition collaboration are in their infancy, but the researchers behind them expect rapid progress. “A year from now, or two years from now, we’re going to have answers to some really big questions,” says Hare.
The hope is that these projects can shine a light on how genes control traits we associate with personality – an idea science-fiction films have toyed with in the past. “If you were to ask me 10 years ago if it was doable, I would have said absolutely, positively no. No way,” says Ostrander. “If you ask me now, I say absolutely, positively yes.”
If Ostrander is right, what might the future hold? At the simplest level, knowing the genes that influence behaviour in dogs might help breeders, trainers and owners understand the best way to handle a particular dog. An unusually nervous one might need a gentler hand in training, or be a poor choice as a service dog.
Of course, you could argue that experienced trainers don’t need fancy genetics to know which dog has a gentle or jumpy temperament. Could there be a more drastic use for all this information? Instead of making allowances for our pets’ adorable but occasionally infuriating personalities, could we simply correct genetic “flaws”? A kennel owner in the US has already proposed doing this, albeit not for a behavioural trait. David Ishee suggests using CRISPR gene-editing technology to swap out a defective gene that causes severe urinary stones in Dalmatians. “If it could be done safely, why wouldn’t you do it? It’ll make the dogs healthier,” says Karlsson. However, she notes, most behaviours are likely to be harder to modify in this way. For one thing, they often depend on several different genes, rather than just one. And each of those may have several unrelated functions – IGF1 is a good example of this – so the likelihood of unwanted side effects is high. Finally, nurture is clearly important as well as nature: even a dog that was genetically modified to be less aggressive is likely to bite if mistreated. All this means that behavioural genetic modification may not yield enough benefit to be worth confronting the sticky ethical issues.
“Dachshunds and yorkshire terriers have a gene that makes them both aggressive and small”
In the long run, the biggest questions – and also the biggest ethical minefield – have to do not with our best friends but with ourselves. Do some humans share the same genetic predispositions to aggressive behaviour that Alvarez finds in dogs? If so, should society treat such people differently from their more placid kin, or even edit aggression out of future generations? The ethical issues this raises make dealing with dogs seem easy by comparison.
Sooner or later, we will need to grapple with these questions and our growing ability to meddle in genetic destiny. It may seem cruel, but if we can tackle them in dogs first, that might help. “We can’t stop it,” says Karlsson, “so I think the more we talk about it, the better off we are in the long run.”
This article appeared in print under the headline “Not just a hound dog”
