In the latest installment of our occasional conversations with Fort Worth newsmakers, Todd Castoe, professor of biology at The University of Texas at Arlington, discusses how rattlesnake venom defies scientists’ previous understanding of why snake venoms are so variable, how evolution drives this variation, and what that means for treating snakebites.
Castoe is also the senior author on a related study that published July 18 in Nature Ecology and Evolution.
This conversation has been edited for length and clarity. For an unabridged version — and more about snake venom — please listen to the audio file attached to this article.
Alexis Allison: Dr. Castoe, why the world of snake venom? How did you get into this field?
Todd Castoe: I’ve been fascinated with snakes since I was a kid. Of all organisms around me growing up, snakes just stood out as really different and interesting and a little scary, to be honest. I started out working on them before I left college, working for The Nature Conservancy, went to graduate school to understand relationships of venomous snakes and how many species there were (and) things like that. (I’d) go around Latin America, collecting snakes and then sequencing genes.
My interest in that faded pretty quickly. It seemed pretty superficial. Snakes are unique and fascinating vertebrates, super different from any other vertebrates. I kept wondering just how they work, why they’re so different and that’s kind of led me to eventually going on and getting a PhD in biomolecular science and a postdoc at a med school in computational genomics. And then here I am now.
We sort of do all kinds of genomics on snakes as model systems, everything from understanding how they regenerate organs after they feed, which is a whole other line of work that the lab does, to studying venom and to using snakes as a model to understand how the process of speciation really works. It turns out that they’re this little Swiss Army knife of a model system that is really seeded with tons of pretty extreme adaptations and innovations like that.
What is genomics?
The study of genes, including interactions of those genes with each other and with the environment.
Allison: I know we need to move on to your research, but I’m wondering, how do you navigate conversations around snakes with people who are afraid of snakes?
Castoe: (Laughs) I kind of started naively assuming everybody was as silly excited as I was. I remember a particular faculty member when I gave my job talk saying, “I think I really liked the talk. But please, next time show less pictures of snakes. It really makes me anxious.” Yes, I have to admit, I haven’t always been sensitive to the fact that not everybody feels the way I do. But yeah, certainly talking about them doesn’t skeeve too many people.
Allison: Interesting, what a cool path. And I know that you and your colleagues at UT Arlington recently published a study — I guess it came out (last) week — in Nature Ecology and Evolution. First, the study involved 68 rattlesnakes. Can you tell us about how you procured them?
Castoe: Yeah, so this study is part of a really big project that we’ve been working on for about a decade, really. There’s this group of rattlesnakes in the western U.S., we kind of called Western rattlesnakes, things that range basically from the front range of the Rockies, from Texas to Canada, including from Baja, Mexico, up to Vancouver. And everything in between. This group is particularly interesting because they recently radiate into, you know, about a dozen very, very distinct forms. Nobody knows how many species we should call them.
They are all different shapes. They’re all different colors. They eat all different things. They have really different venoms. And so for about a decade, we’ve been using this as a model system to explore lots of evolutionary questions. Myself and collaborators, my graduate students, have been spending mostly Mays and Junes in the field in the West, sometimes camping in remote places, looking for rattlesnakes, and that’s that time of year they kind of emerge from their dens from the winter. And yeah, we sample them in the field while we’re there.
Allison: How do you know where exactly to find a rattlesnake?
Castoe: (Laughs) That’s exactly the problem. Sometimes we don’t. I mean, over the years, you kind of get a sense of what you’re looking for. A lot of times these guys like to hibernate in the big rock outcrops where they can get way back in them in the winter. And they face south, so they get really nice sun exposure as it starts to warm up. So that’s one way we kind of point to a few places on a map. But otherwise, we ask researchers locally, we sometimes strike out. And, yeah, it is not much of a science. It’s more of a guess.
Allison: Well, can you tell us what the study was about and some of your findings?
Castoe: Maybe I should set snake venom up a little bit so it’s clear why it’s kind of an interesting system. Snake venom is made up of proteins as many as 50 to 100 different proteins that are toxins that have different activities, are all combined in a single snake’s venom. These proteins are not all just one type. We’re looking at a really diverse — what they call a cocktail — of these proteins.
These proteins have hugely different biological effects, when they’re injected into a prey or injected into a human in case of snakebite. Some of them have very broad effects, like they disrupt tissue and cause tissue to break down or break apart red blood cells. Other ones have very targeted effects. Some of these target other specific proteins in prey bodies or human bodies, like, basically, neuromuscular junctions, things that control your muscle activity and your breathing. And that’s how snakes use them to kill their prey.
So picture this cocktail of tons of really diverse proteins, these proteins are encoded by genes in snakes. And a couple of things have been striking. One, it’s been discovered over decades and decades of research that different species of snakes — so there are about 600 venomous snake species, about 200 or so are venomous enough to be of concern for maybe killing — each species of these venomous snakes have really different venom, which means different levels and different forms of these proteins. And they have really different effects.
Basically, some snakes focus on different strategies for killing prey. It also depends on whether they’re in the desert or the forest, maybe they’re eating animals, maybe they’re eating lizards, maybe they’re eating birds, maybe they’re eating insects. This could change throughout their life. When they’re little, they may eat one thing. And so, we see snake venom varies a ton across species. It varies even within an individual as it ages. It varies within individuals in a population, and it varies between populations.
When the first snake genomes came out, it was really interesting to look at the genes that make these proteins and understand what kind of evolutionary forces were driving these things to evolve to make proteins that seemed to vary so much between species and even individuals. And all of the data at the time was analyzed in a very particular way. Most of these comparisons were comparing very, very distantly related species of snakes, or would compare these proteins in a venomous snake to something that wasn’t venomous or in lizard or in a human.
The relevance of this is, all the comparisons really had this particular context that it was looking at extremely deep evolutionary divergences. So looking at proteins that diverged from each other maybe 100 million years ago and using that to make inferences about what process was shaping this. And all that research really returned one conclusion: Directional selection (was driving these changes). Directional selection is kind of really easy to envision. It’s saying that, let’s say there’s a single trait — let’s say, wing size. And the bigger your wings get, the more mates you can find, the more you can reproduce, and therefore bigger wings are selected, and it goes in that direction. And so when we think of directional selection, we think of a single linear direction that evolution wants to take something that makes something survive better.
And what that suggests, also, in terms of the process, is that — versions of a gene that make different versions of a protein are called alleles — so these different versions of genes, once we get a version that makes bigger wings, or better wings, we’ll just throw that old version out, keep the new version, and in the process, we’re constantly purging all these prior versions that are no longer the best one. And that results in really characteristic signatures of variation in the genome.
Specifically, what it does is, it throws out variation. So in areas that encode these genes, we end up seeing very few alternate versions, just the one that was great. And everybody else got thrown away. And so that leaves very clear predictions of how to find that pattern when we look at genomes and compare genomes from multiple individuals in a population.
And so that’s what was the state of the art, this was the quintessential textbook example of directional selection being so strong to make the single best toxin to kill the lizard the fastest or to immobilize the mammal quickest so it wouldn’t bite the snake. This idea that there was one perfect version and all the other versions didn’t matter, so we threw them out, is the paradigm for how we thought selection drove this great diversity of snake venom. Which would also mean that, boy, every different snake species must have a different, perfect thing they’re trying to figure out, because it’s driving all this variation.
Only recently was it practically affordable and possible to sequence multiple genomes for individual snakes. What that allows us to do is really test this idea of, hey, if it’s directional selection, did we indeed throw out all those other versions? Turns out the answer was no.
There’s other theories for how selection can operate in different ways. And another alternative to this directional selection is balancing selection. Balancing selection works differently because, what we’re saying is that there’s never one permanent, perfect version of a protein. Instead, we might have a new version that works really well for now, and it might not work right later. And over time, it might work less well and less well and less, and you would say, ‘Well, why the heck would a venom work less and less well over time?’
Well, turns out a whole other suite of literature has shown all of this evidence for all different kinds of prey, whether it’s lizards or rats or mice or whatever, squirrels, that they tend to evolve resistance to snake venoms really, really quick, so much so that when you take two populations that might be 20 miles apart of snakes and squirrels, and you mix them — so basically use snake venom against a squirrel population that’s from the next door town — it wipes them out, whereas the ones from their town are super resistant.
So what that really suggests is a very rapid, ever changing, geographically structured landscape where prey are evolving resistance. Snakes are having to get around that somehow because they have to eat, and there’s this constant, what we call an ‘arms race.’ What that arms race suggests, interestingly, is that there will never be a perfect protein version, there will never be a perfect venom that’s perfect forever. Because the longer it’s there, the more time prey has to find a solution to stop it.
And so this rapid evolutionary race between predator and prey or venom and prey specifically creates a very different and much more interesting evolutionary scenario where no one version of a venom protein is the perfect one for very long. And what that means is, before one can replace all the others, it starts losing its efficacy and other ones become better. And then another one becomes better and another one becomes better. And it does it for a very, very, very long time.
So what that means is, rather than throwing out a version of your favorite venom protein every 100 generations, or something, we have some of these alleles, or versions of proteins that have been maintained that are probably 2 or 3 million years old. And we have ones that evolved recently, they’re all still bouncing around in these populations of snakes, because you never know when one might be valuable. And turns out, the rarer it is, the better the chance the prey have never seen it. And so then the really rare ones become really common. And once they become really common, they’re no longer useful. And then a new really rare one becomes common. And you can see how this creates a juggling act.
That does a couple cool things. One, it means that snake populations have humongously broad reservoirs of versions of these venom proteins, some of them very low frequency. But that could change real quick if selection changes with prey. The other thing that means is some of these versions are ancient. So ancient that species that diverged from one another on other sides of the Rockies, 3 or 4 million years ago, have the same copies of these ancient protein versions still circulating around despite that they haven’t basically hybridized or had any mating between them in millions of years. And that’s pretty rare.
The other thing that this led us to is a really fundamentally different story and appreciation for the process that drives this extraordinary variation in snake venom across species, across individuals, and that has relevance for really understanding why snake bite is so hard to treat.
Allison: I am curious about snake bites and how your research could impact human health care. Can you break that down for me as well?
Castoe: Yeah, so it turns out it’s a time where a lot of people are thinking about how this impacts human health care. Specifically a few years ago and 2017, the World Health Organization, the WHO, identified snakebite in humans as a priority neglected tropical disease that really needed global attention. It affects an estimated 5 million people globally a year, causes over 100,000 deaths, but importantly also leaves a very, very large number of people mostly in low-earning tropical countries permanently disabled.
And so the reason snake bite is so difficult to treat is because making anti-venom is pretty hard. Anti-venom is made by making antibodies in another animal. They do this by going out in the field, collecting snakes, bringing them back to the lab and extracting venom from them by basically squeezing their head very carefully and collecting venom proteins from a bunch of different snakes from all over the place and putting that together and injecting it into large mammals in small doses.
That doesn’t harm them, but they have a reaction. And in that reaction, they make antibodies, just like vaccines would make antibodies. Those antibodies are harvested from those animals. And then we take those antibodies and inject them into humans and try to have those antibodies neutralize, i.e., stick to and stop the activity of venom proteins when you get injected by them.
The problem is, those antibodies are only as good as they are specific to what you get bit by. So if you make antibodies from a snake with a particular kind of venom, it will probably only be useful for that particular region. Or maybe even that part of a country, right. And you have to do that for every snake species that’s in that country. What that does is create this problem where it’s financially not tractable. So, one, anti-venom is outrageously expensive, because it’s really expensive to produce, you have to deal with huge risks having these highly venomous snakes around, you have to deal with inoculating these horses, or sheep, and all this other biochemistry down the line to to clean it up.
And what that results in is, there’s huge parts of the world where anti-venom is either non-existent, or the one that is available is so low in its efficacy, it’s not really very effective. There’re parts of the U.S. that you’d be in rough shape, because not many of the available anti-venoms would help you.
What this means then, if you zoom out a little bit farther, is that the primary reason that snakebite is such a global problem to treat is directly related to this extreme amount of variation we see at all levels — between species, within species, among individuals.
So, first, what’s the relevance of our study? Well, this is the first step. At least we now perhaps understand the process that’s generating this. We may then perhaps be able to predict what populations of snakes might be more variable than others. We also know that it’s incredibly important to not just sample one snake from a population but that different individuals in a population might have very medically relevant variation in their venoms.
Other groups that are not us, a number of groups from Liverpool in particular, a colleague has been doing other work to try to think of alternative approaches rather than this clunky, expensive, animal-oriented production of anti-venoms. Can we use any existing drugs or small molecules that are available from pharmaceutical companies? Do any of these stop any venoms worldwide? So that’s been a really big contributor.
There have been other proposals to use a new technology called organoid technology, this is pretty neat. This would take the highly venomous snake part out of the picture. What this is, is the ability to take cells from an organism, treat them with some chemicals that basically reverse development to make them into stem cells and then treat them with a few other chemicals that get them to undergo development again in a petri dish. The result of this, is that, with these approaches that we’ve developed, for studying lots of different things in medicine, they’ve also been able to grow little tiny, miniature versions of snake venom glands in a petri dish. Why the heck would we want to do that? Well, we could use a petri dish, not a roomful of venomous snakes to produce boatloads of these proteins that we could then make antibodies from. So that might just automate and therefore hopefully, maybe, make cheaper this process.
I mean, the other thing to mention is we mentioned all the bad things about venom. Turns out that because these proteins are so perfectly tuned as targeting mechanisms. In some cases, they target single molecules. In mammals, for example, they are perfect pharmaceutical tools for modifying slightly to make very, very targeted drugs that might have very few side effects, things like pain medications. So there’s also a lot of upside to venom.
Allison: What’s next for you and snake venom research?
Castoe: That’s a fun question. While this was a pretty large study, this is one species. To be really curious to see if this is the rule in snake venom. I think it makes a lot of sense that it is.
Another really fun side of this is, we’re looking only at the predator side, only at the venom side. What this suggests are lots of different mammals and lizards and things like that probably have a whole other set of protein variants that are trying to keep up the other end of this arms race, which would be really insightful to understand.
Other work for us on venom. A few weeks before this paper, we had a big paper come out in the Journal of Genome Research. That paper was the first to take a stab at trying to understand how did snakes ever beg, borrow and steal all these genes from the vertebrate genome and repurpose them to be these gnarly toxins? And in the process of doing that, how do they get all these really dangerous proteins regulated in a very precise way to all 100 of them be produced in the venom gland and coordinated?
That is a broader question about, how do you evolve such complex adaptations where it seems like a more random evolutionary process would have a hard time explaining. And it turns out that snake venom genes are a super cool example that illustrates how precise evolution can be in affecting and basically cobbling together bits and pieces of existing regulatory sequences, existing genes and repurposing them for very fundamentally different use and coordinating them in something like venom. Right, so, that’s a whole other fun story that’s just at the beginning for us.
Allison: Thank you so much for this. Dr. Castoe. Is there anything else that you’d like to share?
Castoe: I think the average layperson might say, ‘Who cares about snakes? Why are you studying snakes? Why do we care about that?’ I think that’s always a valid question. And I think the answer is, we use model systems in science to understand basic universal processes that shape everything. The processes that we’re talking about processes of selection, how it works in the genome, are some of the same processes that, for example, are relevant when understanding why we still have bits of Neanderthal genomes in modern humans and the role that plays in disease and the roles that played in human survival.
And so I think that it’s easy to dismiss research that’s not on humans as, who cares? This is esoteric. I think that it’s important then to appreciate that, we gain insight into how everything works by figuring out how something works first. So it’s a funny example of how, sometimes, these very different extreme model systems end up being the thing that shines light into darker, more difficult-to-understand corners of biology. And, yeah, sometimes we get some gems, and snakes are full of those gems.
Alexis Allison is the health reporter at the Fort Worth Report. Her position is supported by a grant from Texas Health Resources. Contact her by email at firstname.lastname@example.org or via Twitter. At the Fort Worth Report, news decisions are made independently of our board members and financial supporters. Read more about our editorial independence policy here.