Elinor Karlsson: From wolves to dogs and back again

Dexter. Image Credit: Sarah Hazell, All rights reserved.

Click here to listen to the full podcast episode

Kat: First, Georgia caught up with Elinor Karlsson at the University of Massachusetts Medical School and the Broad Institute. She and her team are looking into all aspects of dog genetics, from discovering what gives different breeds their distinctive traits to finding out how genetic variations contribute to psychological characteristics. 

They’re the lab behind Darwin’s Dogs - part of the bigger Darwin’s Ark project - which is working with pet owners and gathering doggie DNA samples from all over the world to understand what makes dogs tick and get insights into animal and human health. And, as she tells Georgia, she’s also interested in the question of how we got from wolves to dogs, and what happens when wolves and dogs start interbreeding again - plus, solving the mystery of the missing ancient American doggos.

But before Georgia got to asking her about her work, she had to address a confusing issue - that Elinor had appeared with a cat on the video call, rather than a dog…

Elinor: Yes. You know, it turns out cats also like novelty so I don't usually sit in this particular room and so she thinks it's pretty much the coolest thing that's happened probably this week is me sitting in this room to do this podcast.

Georgia: I have to say, when I saw the cats in your Zoom background I was like, hang on, wrong animal.

Elinor: Yeah, I know. Everybody always thinks I must be like a super big dog person because I studied dogs and it's like, I mean, I like dogs, but I actually got into it because they're an incredible genetic model system rather than because I'm all about dogs.

Georgia: Nice, and do you have a favourite dog?

Elinor: It's probably my sister's dog Besco. She's a total nutcase of a dog. Super high energy, anxious about everything and really sweet. She's a wonderful dog.

Georgia: What breed of dog is she?

Elinor: I actually do know what her genetics are because we have sequenced Besco so many times at this point in time. So Besco became our baseline test dog. So every time we were trying out a new technology, and in science when you're trying something new the first thing you want to do is look at something you already know about so that you can make sure the new thing you're doing is actually working, and so every time we switched the genetic technology we were doing, I would send my sister a text message and be like, "Hey, can I come get us another saliva swab from Besco?". Because we want to test her DNA again and see if this thing is actually working. 

So Besco is a total mutt. She's a mix of a lot of different breeds and I'd have to check exactly what she is. But she, like a lot of the different mixed breed dogs that we looked at, is many different breeds not just one or two or three

Georgia: We'll probably get into this later, but can you just take a dog's genome and then be like, Ooh, that's a bit of Dalmatian. Ooh, that's a bit of Terrier. Is it clear what they are from their genetic profile?

Elinor: Yes, you can. There's a couple of things about it that make it a little bit difficult. Obviously there's algorithms and fancy technology involved and kind of changing a whole string of A's, C's, G's and T's into; "This is ancestry from this breed, and then this other piece is ancestry from this other breed". So that's the technical challenge, but there's a logistics challenge as well, which is that it's not like when you're looking in the genome and it's written in there that, this is German Shepherd DNA. The only reason you call that DNA German Shepherd DNA is because you have a set of German Shepherds that you've compared it to and realise that the DNA matches German Shepherd DNA closer than any of the other breeds you looked at. 

But that means that if you don't have any reference data from a breed, you don't have it in what we call our breed panel, we can't ever find it. And we might end up getting it wrong just because it's trying to match it to something and it'll find out whatever the closest match is and if the actual breed isn't in there, then you can't match to it and it might make a guess at something else.

Georgia: So let's talk a bit about the origin story of dogs. So this is a fascinating topic for loads of people, but the genetics must give a whole different angle. So where did dogs actually come from?

Elinor: You're just going to start with all the really hard questions. So this has ended up being a much harder question than a lot of people anticipated it would be for a couple of different reasons. So the short answer is we don't know yet, and it's not like we haven't been trying to figure this out. There's a lot of scientists that are really interested in this question. Partly because we all love dogs, but also because dogs are kind of part of the story of human history as well. And so learning something about dogs actually tells us something about the people too. Because the whole point of domestication is that they're living with or near people.

So what we know right now is that dogs were probably domesticated around 15,000 years ago, but it might've been longer ago. It's not very clear when, because it's thousands of years plus or minus. And we're not really sure where at this point in time. And I think we will get better at figuring that out over time, but it's just ended up being a really challenging question and, in order to answer it, we basically need more DNA from ancient dogs.

So for a long time people tried to figure out this question by looking at modern dogs and that is really hard because there's this really interesting feature to dog history. So there's two really interesting parts of dog history that made this really hard. The first is that the wolves that dogs were domesticated from don't appear to be around anymore. And so when people first started trying to figure out when dogs were domesticated, they compared them to modern wolves and they got these really ridiculously long ago dates, 60,000 years ago, which is well before there were any humans around to domesticate them so that would have been really strange. And the reason that happened was that they were actually probably picking the split between those two lineages of wolves rather than the split between dogs and wolves. And so they'd found something that just didn't find the thing they thought they were looking at.

The other reason why it's really complicated is actually the breeds. And so what happens when you create a dog breed, and the dog breeds for the most part are only a few hundred years old so mid 1800s Victorian England got very excited about dog breeds. And that's when the idea of a dog breed, which is a group of dogs that are reproductively isolated so those dogs breed with each other but they don't ever breed with dogs outside of that population. 

And what happens is you lose all the genetic diversity outside of that and after that you only have the genetic diversity that was in those 20 dogs or 30 dogs or however many it was, and so that kind of almost removes the history before that. And so in a lot of ways, the information we need to figure out the history of dogs got taken out of the population when they created the breeds. And so now that we're actually looking at a lot of breeds, but even more importantly, we're looking at the non breed dogs. I think we're going to be able to get a lot more information about what happened in the time before the breeds and dog history.

Georgia: Wow. So there's almost a two stage story of domestication. You've got this long history of these mysterious dogs that we had for thousands of years, then the Victorians come up and muddle everything together and that's the genetics we see today is sort of hiding the original story.

Elinor: Yeah, exactly. And it's one of those interesting kind of questions of scientific bias that comes from "it's not science, it's our cultures" and we don't realise what we're doing because a lot of the early dog genetics was done by people in the United States, Canada, and Europe, and in those places, they don't really have non breed dogs. Most of the dogs are either pure bred dogs or they're mixes of pure bred dogs and so I don't think it really occurred to us when we were starting our research that, by studying the breeds, we were missing most of the dogs on the planet because those weren't the dogs that were in our world and we just didn't think about it.

Georgia: Right, and the fact that all these dogs we have are a small proportion, we've lost a load, did they just get bred out these other types of dogs? Did they just die off? Because a large amount of dogs can't have been under sort of Victorian pet owner control. So what happened to those dogs?

Elinor: You know, I find this to be a really interesting question because of exactly what you said. We got so focused on the breeds, but the idea that all the dogs on the planet could possibly be descended from these small number of dogs and just a few places on the planet where breeds were created is just kind of silly when I started to think about it. And so there's two different parts of this. The first is just that there are lots of dogs that we have never even looked at the genetics of that are still around and I think now dog geneticists are making a much more concerted effort to go out and find those dogs and sequence them and get their genetic information into the databases we're using so that we're not so biased towards the pure bred dogs. But I think the other thing that's really interesting is that some of them did seem to vanish.

Elinor: So there's this really interesting story that came out a few years ago about the ancient American dogs and this was one of those classic "once you think about it" it is a huge puzzle because there were always dogs in the Americas. For pretty much as long as there are people in the Americas, there were dogs in the Americas. And there were millions of them and they were everywhere. Yet, somehow, when the Europeans showed up, they seem to have vanished. And we don't know how that happened. Vanished in the sense that if they had been there and the Europeans had just brought dogs in and they'd mixed with the local dogs, you would expect that dogs in the United States today, when we looked at their genetics, in fact dogs anywhere in the Americas, when we looked at their genetics, we would expect them to be a mix of ancient American dog and European ancestry dogs.

And that was, I think, what we thought we would find when we started looking into this. But the study a few years ago, which I wasn't actually involved in doing the work for but I just found really interesting so I wrote a little editorial about it for the journal, found that if there is ancient American DNA in American dogs today, it's vanishingly small percentage, we're kind of talking about 2%, 3%. It's just not very much, and so there's this kind of big question now as to what happened to all of those ancient American dogs, how did they disappear so quickly once the Europeans showed up?

Georgia: Wow, and are there any theories at the moment?

Elinor: Oh, there are all sorts of theories out there. I have to say that as a scientist, my bias is towards the same thing that happened to the people in the Americas. I mean, we know that when the Europeans showed up they brought all of these terrible diseases with them that the people that were living in the Americas hadn't been exposed to before and they died in absolutely devastatingly high numbers from those diseases. And so I wonder whether the same thing happened to the dogs, with the European dogs bringing something in, or multiple things, and maybe they were very dangerous to them. And if the people were kind of going through this absolutely terrible time, it might be that no records of the fact that the dogs were dying as well really got left behind.

Georgia: Wow, and going back to the domestication of dogs and specifically wolves. I know comparisons, in behaviour terms, between dogs and wolves has led us down loads of blind alleys and we've made loads of mistakes based on thinking they're the same. When we look at the genetics, apart from it telling us they split a while ago, is there anything we can tell about the genetics of their domestication or anything Interesting that's come up between comparing dogs with modern wolves,

Elinor: Dogs and wolves are very different animals and it's one of the things in the dog/wolf genetics role that really fascinates me the most. There's some very, very clear differences between dogs and wolves. So you can compare wolf genomes to dog genomes and look for where the differences are. The problem with this is that they're two completely separate populations and most of the genome isn't really functionally important, but it acquires mutations over time.

So because they're two separate populations, when you compare them, you see lots and lots millions of differences between their genomes. Most of which don't have anything to do with why a dog is a dog and a wolf is a wolf. It's just these random changes in the DNA. So what we're trying to do, and we have a really exciting project right now, is we're working with sanctuaries that house wolf/dog mixes. 

So these are animals where their genomes are part wolf and part dog. And so we're studying their behaviour to try and figure out whether they behave more like a wolf or more like a dog, and then we're looking in their genomes. And basically we want to find places in their genomes genes where the wolf/dog hybrids that act more like dogs look, doggy and ones that act more like wolves look wolfie, and that would suggested thats a gene that's involved in the behavioural differences between those two populations.

Georgia: Oh wow, and I've got to ask, getting DNA from a wolf, or a wolfie dog, that's got to be a bit tricky?

Elinor: Yeah, so I should just point out here. The wolf dog hybrid project, the entire thing has been a learning experience for me. I did a PhD and a postdoc that mostly involved analysing data on computers and there was nothing in my world that I touched because it was all virtual, to running my own research lab where all of a sudden we're studying wolf/dog hybrids, which I just thought was the coolest thing, and so the actual logistics of studying wolf/dog hybrids was entirely organised by Kathryn Lord because she was the one that actually has experience and expertise in this.

So the system she came up with, so the ones where the people that owned them were comfortable handling them, could do saliva swabs on them, we never did them, but if the person that had the animal wanted to do a saliva swab, they could do that and some of the wolf dog hybrids are incredibly doggy and are totally able to be handled, but there are some of them that you can't do a saliva swab on for obvious reasons. Saliva swab requires sticking a swab into their mouth, and that's not always an easy thing with an animal that you don't completely trust. And so we came up with this alternate system.

So the other, another thing that you can get DNA out of, in addition to saliva is hair, and so we came up with this fantastic system where the wolf dog hybrids live in these enclosures that are usually very large enclosures that have a chain link fence around them, so that's the fence that has those kinds of diamond holes in it, and so what we would do is we would get one of those, people have probably seen them they have them for dogs, prong collars that are very controversial where they have the metal poky sticks on the inside of them. But what we would do is we would actually take one of those prong collars and we would unclip it so that it was a long, thin thing and feed it through the chain link fence so that the prongs were poking out. And then we would spray it apparently either with deer urine or Chanel No. 5, either worked equally well, which was one of the biggest surprises in my research life I have to say, and the wolf/dog hybrids would get incredibly excited about this and come running over and rub endlessly against this prong collar because they thought it just felt so good to scratch themselves on that. They thought it was wonderful. And then we would just pull it back out again because it was thin and we had that chain link fence and just pull these like huge clumps of hair off it. So once we figured out that system, we had no trouble getting DNA from those hybrids that we couldn't actually get saliva swabs from.

Georgia: Wow, Chanel number five. You Never know who you'll attract.

Elinor: I mean, if you're going to hang out with a wolf/dog hybrid, apparently that's really what you want to be wearing.

Kat: Elinor Karrlson from the Broad Institute in Massachusetts. You can find out more about her work at karlssonlab.org, that's Karlsson with two S's, and the Darwin Dog's project, including the awesomely cute dog of the month gallery is at Darwinsark.org. 

Image Credit: Dexter, by Sarah Hazell, all rights reserved