You are here


Podcast / 02.6.17

Revealing the genetic basis of ecological adaptation, with Hopi Hoekstra

Susanna M. Hamilton, Broad Communications
Credit : Susanna M. Hamilton, Broad Communications
By Karen Zusi
Hopi Hoekstra, an institute member at the Broad Institute, a Howard Hughes Medical Investigator at Harvard University, and the curator of mammals at Harvard’s Museum of Comparative Zoology, tells us how her lab combines the genetic, environmental, and morphological traits of wild deer mice to answer questions about evolution and behavior.



Hi. I’m Lisa Girard, director of scientific communications for the Broad Institute, and this is Biologic, the logic behind the science: conversations with Broad researchers exploring what they do and why they do it.

In this episode, we talk with Hopi Hoekstra about her work on adaptive evolution. Hopi is an institute member at Broad, a Howard Hughes Medical Institute Investigator at Harvard University, and the curator of mammals at Harvard’s Museum of Comparative Zoology. Her lab examines how species adapt to their environments, how these traits — both physical and behavioral — evolved, and what their molecular basis really is.

But how do you draw connections between these complicated areas? To start with, Hopi uses what many scientists would consider a nontraditional laboratory model: wild deer mice, of the genus Peromyscus. I asked her to tell us what makes this model so special.


These wild mice have a lot of the advantages of laboratory mice, so we can bring them into the lab, they can breed, we can do controlled experiments in laboratory conditions. Because they’re closely related to laboratory mice, we can borrow all those wonderful tools that have been developed in laboratory mice over the last, you know, several decades.

But they’re unique from laboratory mice in the sense that they’re not human commensals outside of the laboratory, and they are found in a wide diversity of habitats where they’re often very abundant, so you can find them on the top of the Rocky Mountains, on the coast of Maine, the beaches of Florida, the deserts of Arizona, the cornfields in Nebraska. And because they occupy so many different habitats, there’s a lot of opportunity for local adaptation — so, changes that make them better suited for survival and/or reproduction in that habitat type. So they’re both sort of phenotypically diverse, they have lots of different traits depending on what population or species you’re looking at, and they’re also genetically diverse.

And that last point is really important because it’s the genetic diversity that makes them really interesting and a nice model for human populations. Because while laboratory mice certainly have advantages to being genetically homogenous — being inbred — we as humans, of course, are not inbred, and there’s a lot of diversity and there’s a lot of structure between populations. And so our mice sort of capture that same level of diversity that we see in human populations.


Adaption is a pretty broad topic, and Hopi’s lab is interested in a broad range of traits. She got started by linking DNA mutations to pigmentation a decade and a half ago, examining the genetic basis of different coat color patterns that help wild mice to thrive in certain environments. And over the last several years, she’s gone after genes related to other types of variation.


We’ve been looking at some skeletal traits that are, of course, relevant in some sense to human populations — we vary a lot in bone lengths, for example, that give rise to differences in human height. Reproductive traits — so, for example, we have a project that’s been looking at morphology of sperm. And the shape and size of those sperm affects their ability to swim and ultimately compete to fertilize eggs, so that research has some nice implications for human fertility.

And then probably what’s of most interest, and is really sort of expanding in our lab, is trying to identify genes underlying behavior. We study a wealth of behaviors, some of which have nice analogues to human behaviors and others which are a little harder to make the connection.


Hopi noted that behavior is notoriously messy and difficult to measure. A mouse’s actions can be affected by the time of day, how hungry it is, or any number of other factors. So her lab has tried to standardize one approach by looking at burrowing behavior, or the way mice dig their homes. She explained how she studies the burrows of two different species.


One of the exciting sort of species pairs that we’ve been focusing on are called Peromyscus polionotus and Peromyscus maniculatus. The names are not so important; the important thing to note is that they’re two sister species. So, they’re closely related, and even though they don’t overlap in the wild, if we bring them into the lab and don’t give them any choice, they’ll happily mate and produce viable and fertile offspring.

So what that means is that we can use forward genetics, or classical crosses between these two species, to try to pinpoint the regions of the genome that are associated with any trait differences we measure in those two species — one which makes little small burrows, and one which makes these nice long complex burrows, because they have an entrance tunnel followed by a nest chamber and then an escape tunnel that allows them sort of a backdoor “secret escape” if some sort of predator comes in the front door.

And the architecture of that is really intriguing in the sense that the escape tunnel doesn’t penetrate the surface, but gets very close. So nothing can come in the back door, but it’s shallow enough that mice can go out very quickly. It’s really an intricate, sort of clever architecture.

And when we bring them into the lab, they’ll still build those species-specific burrows, which allows us to test them in the lab. And so what we can do is use this variation that just sort of naturally evolved, and try to pinpoint the underlying genes involved.


Hopi uses the burrow as what’s called an “extended phenotype,” a term coined by evolutionary biologist Richard Dawkins. Rather than try to quantify the digging behavior itself, her team measures the physical burrow to correlate its shape with a mouse’s genetic makeup.


One of the nice things about burrows is that we can put a mouse in a box full of, ah, dirt, let it build its burrow, trap that mouse out of the box in the morning, and then that burrow represents the behavior that that mouse has performed over the previous night. And what we do is just make a cast of that burrow, and then we treat that like any other morphological trait and we measure it like you would measure the length of a femur. So it’s a way to circumvent the challenge of measuring behavior because we can transform the behavior into a morphology.


In addition to using extended phenotypes such as burrows to understand the genetic circuitry associated with behavior, Hopi’s lab is exploring the basis of other behaviors with these same two species.

For an example of particular interest to those listeners who might have already gone a round or two with their partner today divvying up child care responsibilities, she’s investigating the way different mating systems actually give rise to differences in parental care. I asked her to tell us a bit more about this.


So one of the exciting things about these two species, in addition to differences in burrowing behavior, is that they have different mating systems. So, one is highly promiscuous — what that means is basically males and females will mate with whoever they run into. And they’re one of the most promiscuous mice, so, for example, if you genotype a litter in a wild population, most likely that litter is sired by more than one male.

And then the sister species, by contrast, is monogamous both genetically and socially. So what that means is that if you genotype their litters, they’re always sired by the male who’s there in the burrow with the female. And these are really extremes of what we see among mammals. Most mammals are promiscuous, and monogamy is the sort of rare exception.

Now, folks have found an association between the mating system and the underlying parental care that’s given to those pups. So the most extreme is paternal care, or the care that’s given by the dad. So if you’re promiscuous, the dad generally will run into a female, they’ll mate, have offspring, but the dad is off finding another female and contributes very little to parental care. Whereas, in the monogamous species, the male and the female will be there with the young and they’ll both contribute to nest building, huddling, licking, retrieving pups, et cetera.

And so what we can do is a cross between these two species and measure the parental care as the trait, and then identify the genetic regions — and ultimately the genes — that give rise to this difference that we see in parental care. And using this approach, the same approach we use to find genes involved in pigmentation or nest-building or burrow-building, we’re starting to identify and pinpoint those genes involved in differences in parental care.

And there’s so much variation in these natural populations of mice. It’s been really fun, because almost any trait we look at, we find variation and then can capitalize on that variation and use that as a way to find its underlying genetic basis.


While her work is focused on wild mice right now, Hopi is also looking for opportunities to carefully hypothesize beyond rodents.


We think about this a lot related to mice, but it’s hard not to then extrapolate to “what could this mean to human behavior?” Um, where I think we have to be really careful, but it certainly is fun to think about that. And for us, you know, it may not be the same genes that we find in these mice that contribute to, let’s say, variation in male care of young in humans — assuming there is a genetic component. But it could be that, using these genes, we find the neural pathways that are involved, and that gives hints to, you know, differences, not at the genetic level, but let’s say at the neurobiological level.

So these mice, I think, are going to open some really interesting doors into how we think about or even study human behavior.


Hopi’s own scientific behavior didn’t begin with deer mice. Instead, she started on the West coast with something that had a few more legs. For the end of this podcast, she shared some of her experiences as a young researcher.


When I was an undergrad, I went to University of California, Berkeley, and I started as a political science major. And then quickly realized that that just wasn’t what got me excited, so I started in on biology, and I think the thing that got me hooked was starting in a research lab.

And it wasn’t the most glamorous work. So, I was running cockroaches on treadmills and measuring their oxygen consumption. It was a really exciting biomechanics lab that I was working in, so I got hooked on research but not necessarily on cockroaches. So I took a year off, trying to figure out what exactly I wanted to do, and I ended up working at Yellowstone National Park with the bear management team. So I learned how to, you know, shoot tranquilizer guns out of helicopters to tranquilize grizzly bears and take samples and so forth, so that was the other extreme.

So I went from cockroaches to grizzly bears and then settled on mice as a graduate student, which seemed like a perfect in-between. They were, ah, you can get lots of samples, you can manipulate them in the lab, et cetera, but they weren’t maybe as slimy as cockroaches. No offense.

So when I first started graduate school, it was a time when molecular biology and organismal biology were sort of really starting to meet. And so some of the first classes I took were about “how do you use molecular tools to address questions in organismal biology?” And that’s really what laid the foundation for the work that I’ve done since.

And it was really when I started my own lab that I started to think about deer mice as a nice model system. And there is, you know, a hundred years’ worth of natural history literature on these mice. Because they’re so abundant, lots of people have studied their ecology or their reproduction or their behavior in the wild, and so that’s really the literature that we read that gives us ideas about “maybe these behaviors differ in a robust way,” and we’ve really taken those, that natural history data, to then apply modern genomics, genetics, molecular biology, neurobiology, to questions that some of these early naturalists wondered about. Like, “I wonder what genes are causing this difference,” or “I wonder what differences in the brain are occurring and why they’re behaving this way,” but just weren’t able to answer those with the sophistication and tools that we have now.


Hopi’s truly interdisciplinary approach combines molecular techniques with behavioral observations and field-based experiments to find these answers. You can read more about the latest questions she’s pursuing, and her team’s discoveries in behavioral genetics, at For the Broad, I’m Lisa Girard. Thanks for listening.

Biologic on SoundCloud

BioLogic on iTunes