MICHAEL KOTLIKOFF: Hello my name is Michael Kotlikoff, and I'm the Dean of the Veterinary College at Cornell University. I'm here today with Adam Boyko, Assistant Professor of Genetics in the college. Adam is one of the world's foremost experts on canine genetics. This is an exciting area of understanding those genes that make this wonderful companion for humans what in fact it is. From media reports, we know that DNA has begun to inform a lot of our understanding about who we are, how we're made, how we behave, and what diseases we're susceptible to.
What you may not know is that the dog is emerging as a superb model system for understanding complex traits in both the dog population and the human population. By complex traits, I mean those traits, whether it's our height, our appearance, our skin color-- in the case of dogs, coat color or coat length-- that are encoded not by a single gene, but by the interaction of a collection of genes. The dog is uniquely set up in that it has been bred by humans to achieve certain, what we call, complex traits or phenotypes that are desirable.
And that breeding process has provided us with an opportunity to really give us insight about the way that genes determine these complex traits. So Adam, it has now been six years since the publication of the dog genome. How has that information begun to be used by scientists? And how is that providing us understanding about the origin of the dog and about some of these aspects of the diversity of dogs that we see today?
ADAM BOYKO: Hi, Mike. Thanks. Yeah. The publication of the dog genome has been a huge boon for research into dog genetics. It essentially gave us a roadmap for what's going on in the dog. We can now line up sections of the dog genome. We can match it up with the human genome or the mouse genome, and we can see for the first time that dogs are actually more genetically similar to people than mice are. We're able to compare different dog breeds to figure out where in the genome might the differences for these breeds lie.
And we're able to develop new technology that allow us to interrogate the entire genome at once. So we have been able now to map several different diseases-- actually over 100 diseases in dogs-- by looking at dogs that have certain traits or conditions and other dogs that don't and figure out where in the genome these differences lie. So it's really accelerated our ability to tell what mutations are underlying disease, as well as what mutations have affected the course of dog evolution, adding different morphological features, a lot of fur phenotypes, coat color phenotypes, body size, body proportions. And not just in dogs, I might add, but also to the sister species for dogs. Wolves, coyotes, jackals. This has been a huge bonus for conservation efforts and what not, being able to understand the history of these species as well.
MICHAEL KOTLIKOFF: So I should say before we go further that all of this information arises from samples, whether it's blood or saliva from dogs. So not invasive in any way, and not harming dogs in any way, but providing some tissue for analysis of DNA. You mentioned dog evolution. When I was younger, I think the thoughts about the origin of dogs was that dogs may have arisen from jackals or species-- I think there was a controversy around jackals versus wolves. Has that controversy been settled, and do we know for sure the origin of the dog?
ADAM BOYKO: Yes. So that controversy has been around for a long time, starting with Darwin, who proposed that dogs were probably a hybrid cross between wolves and jackals, or possibly some dogs were related more to wolves and other dogs more to jackals. And this idea was actually exponent by Konrad Lorenz early in his career. And the idea simply was there's so much diversity in dogs these days.
It's almost impossible to think that they could have come from a single wild species that just lacks that diversity. However, if you have multiple species contributing to the ancestor of dogs, then you've got many more variance at the disposal that you can shuffle together and get this huge acceleration of canine forms. But interestingly, when scientists started developing genetic methods and looking at them, it became very clear that dogs did only have one ancestor and that is gray wolves. And in fact, the gray wolves had to have come from somewhere in Eurasia as well. So they have that much resolution for it.
Native Americans, of course, did have dogs. But if you look at the DNA from the ancient burials there, it's clear that those dogs carry the Eurasian gray wolf signatures. So it's not that there was a second domestication event in the New World. So geneticists are getting closer and closer to pinning down exactly how and when dogs originated. But at least we know who the parents of dogs are and that those are gray wolves.
MICHAEL KOTLIKOFF: So tell us a little bit about the village dog project, your project trying to understand some of this diversity of dog types and some of the original ways in which these dogs evolved with communities.
ADAM BOYKO: Right. So I got my start in dog genetics in 2008 and 2009 working as a researcher in Carlos Bustamante's lab. They're on a project that we called the Can Map Project, for canine mapping project. And I was impressed at the 1,000 different samples of dogs that we had. All of them came from purebred populations. And of course, there's a huge variety of purebred populations. They're extremely interesting to geneticists.
But I was really interested in the origin and evolution of dogs. And most of the world's dogs, of course, are not part of these purebred populations but actually live as what we call village dogs. So these are free ranging dogs living throughout the developing world and even in some other countries. And for the most part, a lot of these dogs live probably how dogs have lived for thousands of years. So behaviorally, they're very close to what early dog populations were like.
And the question is, genetically, are they just an amalgamation of different breeds that have sort of resumed a feral lifestyle? Or do they actually carry unique signatures from their population history which is different than the signatures that we see in the modern breed populations? And so I embarked on a project to start collecting samples from what I felt were these potentially very interesting populations and analyzing the DNA with some of the latest genomic technologies that have been available for dogs. And we've gotten some very interesting answers. And we see many populations that have very interesting genetic signatures which point us in the direction of where dogs originated from and how they're related to other populations.
MICHAEL KOTLIKOFF: So tell us a little bit about some of the exotic places that you've collected dog DNA.
ADAM BOYKO: So myself and other researchers on the project have gone to several different countries. I haven't done all of the sampling myself. But we've collected now from almost 30 different countries. I got to go on an expedition in Peru last year, which was extremely interesting, sampling dogs there throughout the Sacred Valley, and throughout the Amazonian jungle, and on the coasts of Peru looking there for the ancestors of Native American dogs to see if we see any dogs today that can trace their ancestry from early Native American dog populations versus the majority of dogs that we see running around the streets of Central and South America now, which seem to be a mix of various European breed dogs.
Our latest sampling expedition was to Western Africa. So we just got samples in here from Liberia and the Congo. And we're looking to understand more of the African population history of dogs. There's some very interesting African breeds. There aren't any local wolves in Africa that can complicate the relationships between the populations through interbreeding. And so it's sort of an interesting ancient reservoir of dog diversity that is yet to be fully analyzed.
MICHAEL KOTLIKOFF: So we've heard recently some very interesting information, again from DNA, around very early human-like species, non-hominids that interacted and exchanged DNA with the ultimate human population. It sounds as though the evolutionary history of the dog is somewhat similar in that it occurred in a similar region-- this region of the Fertile Crescent, Africa, Northern Africa, Asia-- and that domestic medication occurred with the origin of human civilization around that time-- domestication of species and plants. Can you say anything about that, the emergence of the dog with human populations?
ADAM BOYKO: Yes. So the amalgamation of early populations of humans and other hominids has been a tale that only recently was resolved with genetic analysis. And now that we have the dog genome, and we have samples from very interesting village dog populations, we can start to address the same sorts of questions looking for the same signatures in dogs. And there's a lot of reason to suspect we might find this because dogs and wolves are so closely related.
We know that they interbreed freely, at least potentially. And we know that dogs and coyotes even can interbreed. And not only do we know they have the potential for interbreeding, we know they do. There are wolf-dog hybrids that we see in the wild living as wolves, that we see people trying to make as pets. There's even evidence that some of the traits in wolves that we see now in America, for instance, the black melanin pattern that we see in many wolf populations, actually originated in dogs and was introgressed into wolves through hybridization.
So the question is whether any interesting traits have gone the other way. And so, was there an early domestication event in dogs followed by subsequent introgression of new alleles from different wolf populations that perhaps gave new traits into dogs and helped them proliferate and become the highly successful species that they are today? And that's where our research is right now. Humans, of course, originated in Africa well before the origin of dogs and spread out through Eurasia. Dogs originated somewhere in Eurasia. And the timing of it is really interesting because they are the first species that we can tell that were domesticated by humans.
And it actually predates, by at least a couple of thousand years, the beginning of agriculture in the archaeological record. So it's not that dogs became domesticated once we started raising large amounts of food and having large trash heaps with vermin and stuff like that, which certainly explains cat domestication, but dogs actually came somewhat before then. So there must have been some other trigger, perhaps, that caused this event. And we're still not sure how much further beforehand. There's some evidence of burials from 30,000 years ago. But that's not as conclusive as more recent archaeological evidence around 12,000 to 13,000 years ago and genetic evidence pointing at a domestication around 15,000 years ago.
MICHAEL KOTLIKOFF: And so moving forward, can you say something about this marvelous expansion of dog breeds that occurred that gives us the breeds that we see today and the diversity in size, and hair color, and coat thickness? When did that occur? Where did that occur? And how is that providing us an opportunity?
ADAM BOYKO: Well, some of the first archaeological specimens that are clearly dogs are rather small dogs. And so we know that there was morphological distinctiveness early on. And in ancient times, there were warrior dogs. There were Hunter dogs. That were lap dogs. There were a whole wide range of dogs-- shapes and sizes. But for the most part, the breeds that we see today are all from this expansion around 200-250 years ago during the Victorian era when people really took an interest in proliferating-- not only dog breeds-- but also chickens, and pigeons, and all sorts of other species.
And of the hundreds of breeds around today, certainly the majority of them have much more modern than ancient roots from these animals. And so this huge diversification that happened owes a lot to this later breeding. And in fact, that makes dogs a very great system for mapping the genetic basis for a lot of this variation. So a lot of these extremely large and extremely small breeds-- it was a very recent selection event, which makes it a bit easier to find in the genome. And so we can learn things about morphology, and even behavior, looking at the genes of dogs in a much simpler way than would be possible in a lot of other species.
MICHAEL KOTLIKOFF: So tell us a little bit more about that, about the project that involves the DNA bank and this effort to try and understand specific interactions of genes that result in diseases or propensity for diseases.
ADAM BOYKO: Right. So as these breeds were being formed, they were generally formed with a relatively small number of individuals. And then they were made into a closed population. So there was no new genes being added to the population. And so because of that, there's less diversity within a single breed than you see within the species as a whole. Much more so than is the case in humans where populations tend to have thousands of individuals that are more or less freely interbreeding.
And so much of the diversity is maintained within populations. In dogs within an individual breed, you have less diversity. And because you're strongly selecting for certain phenotypes around the genes that are causing the traits for large size, let's say, or for black coat color, if that's what's being selected for, you reduce the diversity even further because just one type of that gene is going to come to represent the whole population.
MICHAEL KOTLIKOFF: So by phenotype you mean a particular manifestation of a trait like coat color, hair coat thickness, size, that sort of thing.
ADAM BOYKO: That's right. Now, when you put on top of this the idea of chance, what happens is there's going to be some traits that aren't selected for-- let's say, a propensity for a certain kind of heart disease or a certain kind of cancer-- that by chance are going to reach a high frequency within this relatively small population. And it'll be lost in other populations. And we see that in modern breeds today that the disease propensities for different breeds can vary dramatically, sometimes even with closely related breeds.
So that leads us to believe there's a strong genetic component to these diseases and that by comparing individuals within a breed that have a disease or not, or between breeds where you have highly different disease propensities, you might be able to find the genes that underlie these diseases much more easily than you would be able to find, let's say, in humans where there could be dozens and dozens of different genes that are interacting to cause this disease.
MICHAEL KOTLIKOFF: So on this area of traits, it's really not just diseases, but also things like body size. The dog varies from the Chihuahua to the Great Dane. Some of your early work with Carlos Bustamante, Nate Sutter, and others worked on this size variation. How is that important in understanding the things that control the shape of our body?
ADAM BOYKO: Well, body size is a fantastic model trait. So we know that it is a complex interaction of several different genes. And humans, in fact, doing genome wide studies, we can count over 200 different genetic regions and mutations that affect human height variation, human weight variation, and, in fact, cumulatively these hundreds of mutations still only account for about 20% of the variation that we see between humans.
With dogs, again, we think it's a complex trait. We think that there's lots of genes involved. But because of the way the breeds have been stratified and selected, what we see is that with just a handful of genes, in fact, with six genes we're able to explain over 70% of the variation. So when we're trying to do complex trait mapping based on our results with body size, we think that dogs are going to be a much more amenable system in order to do that.
So with a sample size of just a few hundred dogs, we can explain the majority of variation in body size, whereas with humans these studies involve tens of thousands of individuals. And even though cumulatively they found many different mutations-- or at least genes underlying the variation-- they still could only account for a small proportion of the variation. So if complex disease behaves similarly as body size, that really makes dogs a tremendous model species for discovering some genes that influence these diseases and that are likely to also influence similar traits that we see in humans.
MICHAEL KOTLIKOFF: Terrific. So we have this marvelous model system with all this diversity, a genetic structure that's very similar within these breeds, almost inbred families if you will. And how is the Cornell program, the DNA bank, going about exploiting this system?
ADAM BOYKO: So the DNA bank is taking advantage now of all the genomic tools for mapping and is amassing a huge repository of DNA from dogs that have been well characterized for which diseases they have, which diseases they don't have, entire medical histories there. And so this is a resource now that researchers can tap into and that they can use to map diseases of interest, as well as traits, and phenotypes, and population history. It now contains over 9,000 different canine samples, primarily, and is being used to map traits anywhere from cancer, to autoimmune disease, to heart disease. And I think it's going to be a really valuable resource as genomic technology becomes more advanced in the dog, and cheaper of course.
MICHAEL KOTLIKOFF: So I think we have some advantages of some early success. I'm thinking of Rory Todhunter's work in hip dysplasia. Can you describe a little bit about how that works practically, how we make a prediction about the chances that an individual dog will later in life develop hip dysplasia and what that means for dog disease, how we would go about and perhaps select breeding partners based on the likelihood that a dog would get the painful arthritis in the future?
ADAM BOYKO: Right. So canine hip dysplasia, of course, is a very serious disease. It affects many breeds. Some breeds, it has a very high prevalence. And so by taking individuals that are affected with severe forms of hip dysplasia and comparing them to individuals that are free of--
MICHAEL KOTLIKOFF: Osteoarthritis.
ADAM BOYKO: Osteoarthritis and other predisposing conditions. We can actually use the genome of an individual dog to predict whether it is at an elevated or decreased risk of hip dysplasia, and in fact, even take a male and female dog and predict what the distribution of hip dysplasia prevalences would be in the offspring. Essentially a breeding value. So even though the exact mutations underlying the susceptibility to hip dysplasia haven't been found yet, we can still use genomic information to try to reduce the prevalence of the disease within certain pedigrees and breeds.
MICHAEL KOTLIKOFF: So on a practical basis we may take a German Shepherd litter and provide some information about what dogs are likely to be susceptible to hip displeasure in the future. Of course, that's the genetic component only, not the environmental component. But then perhaps to make a decision that that dog will be a pet, but perhaps not a sire and breeder, and thereby improve the breed in the future in terms of lessening arthritis in that breed.
ADAM BOYKO: Right. And if the genetic information is found early enough, of course, there's interventions that can be done surgically to also ensure that the individual is not going to develop hip dysplasia later on. Things that you wouldn't want to do for a dog that doesn't have a genetic predisposition for the disease.
MICHAEL KOTLIKOFF: And then finally, Adam, how would this information, whether it's cancer, the specific genes that collectively together their variations result in susceptibility to cancer, or the set of genes knowing the different set of genes whose variations together result in susceptibility to osteoarthritis, how do you see this in forming our understanding of human disease, and can the dog be a model system? You mentioned the diversity within human populations. Can this be a way to understand those variations in the human population?
ADAM BOYKO: Sure. There's multiple diseases in humans that we don't understand all the genetic pathways that are involved. Humans get various diseases for dozens of different reasons. And we don't always know where in the genome to look for defects or to look for ways that we can intervene and provide effective care for different diseases. And dogs have been crucial in some diseases for allowing researchers to map a gene that is predisposing individuals to the disease and then look at humans and realize that same gene is playing an important role, and a previously unnoticed role, in the human condition.
And so because of the relative ease of finding genetic variants in the dog genome underlying these phenotypes, it is really a powerful system, not just for understanding canine disease, but also for understanding human disease and understanding disease progression in general.
MICHAEL KOTLIKOFF: So the potential then is that the dog, who has been such a tremendous companion to human population, can really serve as a genetic model to inform us about diseases of the human population. Really, man's best friend giving back in a way in which we improve the understanding of dog disease, and we improve the understanding of human disease.
ADAM BOYKO: Of course. I mean, being a human companion from a geneticist standpoint is terrific because it means that they're sharing the same environment that we are. And because of the genetic similarity, it really does make it a powerful system for helping us understand, again, not just canine disease, but human disease as well.
MICHAEL KOTLIKOFF: Terrific. Well, thank you very much. I want to thank you for taking this time to explain to a very complex situation. And it really is a very exciting area of research.
ADAM BOYKO: Thanks, Mike.
We've received your request
You will be notified by email when the transcript and captions are available. The process may take up to 5 business days. Please contact firstname.lastname@example.org if you have any questions about this request.
A Q&A with Dr. Michael Kotlikoff, the Austin O. Hooey Dean in the College of Veterinary Medicine, and one of the world's foremost experts in canine genetics, Adam Boyko, about how dogs are helping to unlock the mysteries of disease and what it means for you and your animals.