Om nom newtRichard Greene

In the mountains of Oregon, there are newts with so much poison in their skin that each could kill a roomful of people. There are also snakes that eat those newts; they’re completely resistant to the toxins. The two are locked in an evolutionary arms race. As the newts become more toxic, the snakes become more resistant. One team of scientists has been studying this evolutionary conflict for five decades, and they’ve now shown that its seeds were planted 170 million years ago—before either snakes or newts even existed.

We know about this ancient conflict because of a young undergraduate student named Edmund “Butch” Brodie Jr. In the early 1960s, he heard a local legend about three hunters who were found dead at their campsite, with no sign of theft, struggle, or foul play. The only thing amiss at the scene was a dead roughskin newt, which the hunters had accidentally boiled in their coffee pot. These dark-backed amphibians have vibrant yellow-orange bellies, which they display to predators by arching their heads and tails over their backs—a clear sign that they’re poisonous. Perhaps those poisons killed the hunters.  

Butch tested this idea by collecting newts, grinding up tiny amounts of their skin, and feeding the extracts to other animals. Everything died. The newts proved to be absurdly lethal. Another team of chemists showed that they carry tetrodotoxin (TTX)— the same poison found in the skins and livers of pufferfish. It’s ten thousand times more toxic than cyanide, and among the deadliest substances in nature. Each newt seemed to carry enough to kill any predator hundreds of times over. Why would they be so ludicrously toxic?

Butch found a clue when he checked one of his traps and found a common garter snake devouring a newt. Overcoming his mild phobia of snakes, he collected some and found that they resisted amounts of tetrodotoxin that would kill far larger animals.

While Butch focused on the newts, his son, Edmund Brodie III, became fascinated by the snakes. Together, they showed that throughout western America, places with mildly toxic newts also had mildly resistant snakes. Meanwhile, hotspots with unusually lethal newts also had snakes that withstood staggering levels of tetrodotoxin. The two species were locked in a beautifully coordinated arms race of toxicity and resistance.

But what set off the starting pistol? How did the first snakes survive their encounters with the first newts? To find out, the team needed to understand how the snakes came to shrug off the poison.

Tetrodotoxin kills by corking molecular pores on the surface of nerve and muscle cells, which act as channels for sodium ions. If these ions can’t traverse the channels, muscles can’t contract and nerves can’t fire. Paralysis ensues, breathing ceases, and death follows. In 2005, the Brodies found that garter snakes avoid this fate by changing the shape of their sodium channels, so that tetrodotoxin no longer plugs them.

At the time, they only focused on one sodium channel called 1.4, which is found in muscles. The snakes have eight others. Three are unknown. Another three are irrelevant—they’re found only in the central nervous system, which is protected from tetrodotoxin by an impermeable barrier. The final two—1.6 and 1.7—are more vulnerable: They’re found in peripheral nerves that connect the brain and spine to limbs and other organs.

When Joel McGlothlin started working with the Brodies, he showed that 1.6 and 1.7 are also resistant to tetrodotoxin, having acquired some of the very same mutations that protect 1.4. But in 1.4, those mutations are found in some snakes but not others, which makes some populations a thousand times more resistant to newts than their peers. By contrast, the resistance mutations in 1.6 and 1.7 showed up in all garter snakes. They looked like much older innovations.

To find out when they arose, McGlothlin, now head of his own lab at Virginia Tech, sequenced the genes that encode the sodium channels of 78 species of snake. And to his surprise, he found that one mutation, which makes channel 1.7 thirty times more resistant to tetrodotoxin, is at least 170 million years old. That makes it older than both newts (which arose between 40 to 50 million years ago) and snakes (which arose 140 million years ago). It originated in lizards—the group from which snakes arose.  

“We think this evolutionary change happened for some other reason,” says McGlothlin. It could have altered how quickly or readily sodium ions pass through the channel, and thus how responsive or excitable an animal’s nerves are. And by total coincidence, it also made snakes partially pre-resistant to tetrodotoxin, right from their very beginnings. “Snakes were predisposed to getting into these co-evolutionary arms races,” says McGlothlin. “They had a baseline resistance, so if they ate something with a little bit of tetrodotoxin, they survived.”

Picture the scene 45 million years ago, when tetrodotoxin-bearing newts first appeared. Their poison payloads were small, but enough to block the peripheral nerves of any predator, numbing their mouths. Nothing ate them—except snakes, protected by their partially resistant nerves. That provided the evolutionary pressure for the newts to become even more toxic. The arms race began to escalate. The snakes responded by building up even more protective mutations in channel 1.7, and in the other peripheral nerve channel, 1.6. And only groups with resistant versions of both these channels went on to develop resistance in 1.4—the one in their muscles. That’s when the arms race really took off.

Today, five species of snake have the full gamut of impervious channels. All of them eat toxic amphibians. All of them can thank the same genetic legacy that they inherited from their lizard ancestors. And all of them elaborated on that legacy on their own, often evolving the same life-saving mutations in the same sodium channels.

“This latest piece of the puzzle is really exciting,” says Ashlee Rowe from Michigan State University. It’s a great example of the importance of historical contingency in evolution, where past changes—even apparently innocuous ones—can set the stage for later adaptations.

These sequences of events can be hard to piece together, but there are a few other examples. In 2007, people were suddenly infected by strains of seasonal flu that resisted the antiviral drug Tamiflu. Within a year, the drug became almost completely useless. The cause of its downfall was a mutation called H274Y, which allowed flu viruses to resist Tamiflu. Scientists had ignored this mutation but they knew that it makes flu viruses less infectious, and so they thought it shouldn’t have been able to spread. They didn’t realize that some viruses already carried a pair of mutations that do nothing on their own, but that compensate for the negative effects of H274Y. These strains were pre-adapted to resisting Tamiflu thanks to historical quirks, just as snakes were pre-adapted to resisting tetrodotoxin.

But while Tamiflu resistance is a simple trait controlled by just one gene, tetrodoxin resistance is a more complicated one, controlled by three or more. Still, it’s less complex than, say, height or intelligence, which are influenced by thousands of genes. It sits in a sweet Goldilocks zone.

“If we want to see how things evolve, we first have to understand the full complexity of a trait, which is difficult for very complex ones. Conversely, if we only understand very simple traits, we are missing a lot, as most traits are not simple. This example provides a nice middle ground,” says Danielle Drabeck from the University of Minnesota. “We understand very few traits at the level of detail.”

We want to hear what you think about this article. Submit a letter to the editor or write to