DNA that was frozen for 2 million years has been sequenced

After an eight-year effort to recover DNA from Greenland’s frozen interior, researchers say they’ve managed to sequence gene fragments from ancient fish, plants, and even a mastodon that lived 2 million years ago.

It’s the oldest DNA ever recovered, beating the mark set only last year when a different team recovered genetic material from a million-year-old mammoth tooth. 

The new effort looked at genetic material that was left behind by dozens of species and washed into sediment layers long ago when Greenland was much warmer than today.

“Here you are getting the whole ecosystem,” says Eske Willerslev of the University of Copenhagen, who led the effort. “You know exactly that at this time, and this place, these organisms were together.”

The genetic findings, which paint a picture of an era when Greenland was covered with flowering plants and cottonwood trees, could provide clues to how ecosystems adapted to warmer climates in the past.

“Here you have a map of where and how to edit the genetics of plants to make them resilient to climate change,” says Willerslev. He adds that the ancient DNA could provide a “road map” to help plant species adapt to a climate that’s warming very quickly.

Speaking at an online press conference organized by the journal Nature, which also published the report, Willerslev said the forested ecosystem revealed by the gene fragments included flowering plants and trees, species currently absent from the area, where nothing much lives except lichen and some musk ox.

“This is an ecosystem with no modern analogue. It’s a mixture between arctic species and temperate species,” says Willerslev. “It’s a climate similar to what we expect to face on Earth due to global warming, and it gives us some idea how nature can respond to increasing temperatures.”

Some researchers have proposed using findings about ancient DNA to re-create extinct mammals like woolly mammoths, but Willerslev says plants “will be much more important” even though they are “not as sexy” as a pachyderm.

Research on old DNA began in 1984, when scientists recovered readable genes from a dried-out quagga, a type of extinct zebra. Since then, new methods and specialized gene-sequencing machines have allowed them to probe deeper and deeper into the past.

DNA breaks apart with time, so the older it is, the smaller the pieces become—until there’s nothing left to detect. And the shorter the fragments are, the trickier it is to assign them to a specific groups of plants or animals.

“The huge damage pattern made it very clear it was ancient DNA,” says Willerslev, who says he and his colleagues began working with the Greenland samples in 2006. “When it’s 2 million years, there has been so much evolutionary time, that whatever [species] you are finding are not necessarily very similar to what you see today.”

The Danish team says the DNA they found was preserved by freezing temperatures and because it was bound to clay and quartz, which also slows down the process of degradation.

Exactly how far back in time researchers will be able to see remains an open question. “Probably we are close to the limit, but who knows,” says Tyler Murchie, a postdoctoral fellow at McMaster University who develops methods for studying ancient DNA. He notes that the Dutch researchers were successful in combining several techniques to “create a robust reconstruction of this ecosystem.”

Willerslev once predicted it would be impossible to recover DNA from anything that lived more than a million years ago. Now that he’s broken the record, he is reluctant to say where the limit lies. “I wouldn’t be surprised if…we could go back twice as far,” he says. “But I wouldn’t guarantee it.”

Here’s my guess: Neuralink will unveil a vision implant at today’s “show and tell”

Elon Musk’s brain-computer interface company Neuralink likes to give progress reports via theatrically staged events that it livestreams.

Its next event, scheduled for tonight at 6 pm Pacific time, was announced by the company via a brief video invitation in which the words “please join us for show and tell” appeared as if they were being typed in green letters on a screen.

The mysterious message immediately had Neuralink fans guessing what it could mean.

Here at MIT Technology Review, we consider such teases an invitation to make hard predictions about what Neuralink will show, relying on our understanding of brain-interface research, Neuralink’s capabilities, and in this case, a timely bit of information from a tipster.

With that in in mind, I predict Neuralink will announce it is not only reading brains with its electronic interface, but is now writing information into them, something it could demonstrate with a “vision prosthetic” that generates images inside an animal’s brain.

That’s possible because electrically stimulating the visual cortex, which lies at the back of the head, produces flashes of lights called “phosphenes” that an animal or person can perceive.

I think the demonstration could work like this: Researchers will send stimulation into a monkey’s visual cortex, creating spots of light arranged into, say, the shape of the letter “A.” Imagine, furthermore, that the monkey is trained to tell you what it sees, for instance by typing the letter A on a keyboard.

That could be the “show” and the “tell” hinted at by Neuralink in its announcement.

Such a vision prosthetic would not only be cool, but it’s totally feasible, even relatively easy to do—and certain to draw oohs and ahhs from Musk’s crowd of fans and followers.

Here’s how it would work

The first demonstration that you could make people see spots of light by stimulating people’s brains dates way back to the 1970s. The idea now is to use more electrodes at once to create more phosphenes, then arrange these into a kind of very crude display, like an old-fashioned ballpark scoreboard.

Recently, a group in Spain, using an implant called the Utah array, which has 96 electrodes, found that that a blind woman could use such a system attached to her brain to make out letters.

In its prior events, Neuralink has followed somewhat cautiously in the footsteps of other neuroscientists. For instance, in 2021, it showed a video of a monkey playing the video game Pong with its brain. However, a human with a brain implant had already played the game 15 years before.

Instead of entirely new applications, what’s actually important about Neuralink is that it has developed a sophisticated type of brain implant using thin wires studded with electrodes. It implants the wires into animal brains using a neural “sewing machine” robot that used optics to avoid blood vessels. The device is wireless, too, transmitting information out from under the skull, making it more practical.

When Musk launched Neuralink in 2017, he outlined plans for “a high-bandwidth, long-lasting, biocompatible, bidirectional” brain implant. This brain modem, or “wizard’s hat,” he believed, would somehow allow humans to keep pace with artificial intelligence.

Despite the grandiose vision, Neuralink has more recently been talking about practical aims, like helping paralyzed people control a computer.

That is why the company’s earlier demonstrations involved implanting its electrodes in the motor cortex of the brains of pigs or monkeys. That allows researchers to read movements and transmit these to a computer, like when that monkey used its brain signals to move the Pong paddle.

Guessing game

After Neuralink’s latest invitation came out, some company fans guessed that it was a signal that today’s demonstration would involve an animal typing with its brain.

That certainly is one possibility. Krishna Shenoy, a Stanford researcher and adviser to Neuralink, is working with human patients who’ve set world records at brain typing. Shenoy’s human subjects use an older type of implant, called the Utah Array.  

As far as we know, Neuralink’s implant is not yet being tested in humans, although monkeys can be taught to brain-type, too. In 2016, Shenoy showed that monkeys could use their motor cortex brain signals to move a cursor and transcribe text from the New York Times and Hamlet.

Despite that, we still think today’s demo will involve a vision implant. But how will the monkey “tell” what it sees? The monkey could type the answer with its fingers, or even use a second brain implant to type with its thoughts.

Another reason to believe it’s a vision system is that Musk has claimed brain implants can cure a huge range of diseases. And showing a possible technique for treating blindness is one that the multibillionaire has hinted at himself.

During an interview with podcaster Joe Rogan, Musk claimed that a Neuralink device “could fix almost anything that is wrong with the brain. So it could be something that … returns your eyesight, even if you have lost your optic nerve.”

“Really?” Rogan asked. “Yeah, yeah, absolutely,” Musk replied.

To treat a neuropsychiatric disorder, like depression, it’s probably necessary to put electrodes deep in the brain. But the visual cortex is conveniently located right at the back of the brain, near the skull, where it’s easy to get at.

Neuralink’s implant could make a good vision prosthetic. Its electrodes go inside the brain, near neurons, to they don’t use much electricity, making it safer. Their system of thin wires could also reach more locations, which means more phosphenes, and possibly a higher-resolution “display.”

In studies carried out at the University of Utah years ago, researchers covered students’ eyes with a screen into which holes had been poked. They found with about 700 holes, they could still navigate obstacle courses and even read books.

That means about 700 phosphenes could make a useful form of vision, even though it’s nothing like natural sight, which is incredibly rich, in color, and travels along a million axons from the eye to the brain.

Exactly how the demo will be carried off is anyone’s guess. Although Musk is a drama-magnet, Neuralink has played it pretty conservative in the past, so we don’t expect anything too crazy. For example, while Neuralink once brought a pig onstage, we wouldn’t expect any kind of live demonstration involving a rhesus monkey. They’re a bit aggressive, and animal rights campaigners are already criticizing the company and its university partners for hurting animals. But there could still be a video demonstration involving a primate, as was the case with the monkey that played Pong.

Certainly, Neuralink could go wild if they wanted to. It’s entirely possible for Musk to type letters live, right into a monkey’s brain, and have the animal transcribe his message.

After taking over Twitter, Musk has been acting like a court jester, making rash but, in his view, true comments about free speech and other matters.

I think it would be pretty funny if all Musk’s recent tweets had been transmitted by him and typed by a monkey.

But that is my stretch goal. Everything about Neuralink tells me they will be fairly cautious and stick to a basic demonstration of implant-driven vision. The company wants to show that its device is useful, could help people, and outline a cool future that will attract talented engineers to apply for jobs there.

I could be wrong about all this. So I am definitely tuning in to Neuralink’s event. And so should you. Even though Neuralink’s demonstrations of brain-interfaces aren’t really new, it’s always a great show.

We thank the following researchers for sharing insights and background on vision implants in the preparation of this item: Eduardo Fernandez, Universidad Miguel Hernandez; Konrad Kording, University of Pennsylvania; Richard A. Normann, University of Utah; Sumner Norman, AE studio; and Anonymous.

How much would you pay to see a woolly mammoth?

Sara Ord spent her week talking to scientists about skin cells from a mouse-size marsupial called the dunnart. The cells were sent to the “de-extinction” company where she works, Colossal Biosciences, from collaborators in Australia.

Ord’s job is to lead a team that’s figuring out how to use gene editing to gradually change the DNA of those cells so that it begins to resemble that of a distantly related animal, the thylacine, a striped marsupial predator also known as the Tasmanian tiger that went extinct in 1936. 

If they can make a dunnart cell with enough thylacine DNA, the next step is to use cloning to try to create an embryo—and, eventually, an animal. Another project involves trying to turn Asian elephants into something resembling a woolly mammoth, by adding genes for cold resistance and thick red hair.

Sara

COLOSSAL

There are no resurrected species yet, of course. Ord’s job as “director, species restoration” is really about an imagined future, in which a high-tech combination of DNA technology, stem-cell research, gene editing, and artificial wombs could lead not just to the resurrection of lost species, but also to the preservation of those close to disappearing.

Ord got into the job after trying her hand at lab research, a job in a hospital, and work for a software company. She says it’s a natural fit. She grew up with many pets and watched a lot of Discovery Channel and National Geographic programs. “I have always loved animals,” she says.

It’s certain Colossal is as much Hollywood production as it is hard science. Its financial backers include investor and entertainment mogul Thomas Tull and Tony Robbins, the motivational speaker, and its ideas originate in the laboratory of the outspoken gene scientist George Church, who has been promoting mammoth resurrection in the media since 2013, though with few results yet.

Ord’s job is similarly composed: part communication, part science, and part futurism. And what if the company succeeds in re-creating the thylacine—or something close to it? Ord says Colossal might turn a profit by selling tickets to see it.

In an interview with MIT Technology Review, Ord says the company hopes to produce a thylacine in just two years, by 2025, and a mammoth by 2027.

This interview has been edited for length and clarity.

You have one of the more futuristic job titles I have seen. 

I was one of the first employees here at Colossal. I was with the CEO, Ben [Lamm], and we were brainstorming what my title should be. We came up with “director of species restoration.” The second I heard it, I was like, yeah, that is the one. 

I would have gone with “director, resurrection technology.”

But that can be scary. Right? And so it’s trying to take what we’re doing and making it very digestible for everybody.

How much of your job is communication?

I would say it’s probably a third of my job. The most fun thing to explain is the thylacine project, which I lead. Why bring back the thylacine? The thylacine was an apex predator in the Tasmanian ecosystem. And when you remove an apex predator, you see a lot of negative effects. You end up with a ton of prey in an environment, and they wreak havoc because there’s no population control. Bringing back the thylacine to the Tasmanian ecosystem will hold tremendous value.

The thylacine is a marsupial, but it’s also a carnivore. So something fluffy could get chomped if this works. Are there animal lovers who oppose this plan?

We had an overwhelmingly positive reaction. I think more than anything, it’s because this animal was hunted to extinction. And this is our opportunity to fix that.

What is the science part of your job?

I have a team of 12 genome engineers and phenotype engineers. We also have collaborations with some of our embryologists and our computational biologists. It is reading as many papers as I can, getting my hands in the lab, and pushing the science forward. And then it’s being a part of conversations about—once we have a thylacine, once we have a mammoth, where do we put it? What does that look like? What is the ecological impact of bringing the species back, and how will this help currently endangered species?

You’ve blogged about how bringing back a species involves quite a few steps, including editing genes in the cells of a related species, cloning an embryo, and then bringing an animal into the world. Which of these is the most speculative?

It’s really about understanding how many genes you need to edit. The thylacine is related to the whole family of dasyurids, which includes the dunnart, the quoll, and the Tasmanian devil. But it’s still about 70 million years of [evolutionary] divergence—an extreme amount of divergence. So what do you have to edit in a dunnart or an Asian elephant in order to create a phenotype of a species that will fill the same ecological niche the thylacine or woolly mammoth filled?

Do you have a stuffed thylacine to work from? What’s the starting point for the project?

There was a pup that was preserved in ethanol in the early 1900s—it’s called the “miracle pup.” Our collaborators at the University of Melbourne have been able to extract DNA from that sample and generate a really high-[accuracy] genome sequence from this. In addition to that, there are a lot of pelts in circulation, as well as museum samples, and we’re getting these and generating sequences from them.

Do you have a timeline for when the first extinct species is going to roam again?

Absolutely. For the mammoth, we are projecting a 2027 timeline, and for the thylacine, 2025. The key difference here is the gestation time. Elephants take around 18 to 22 months to gestate, whereas marsupials—and especially the dunnart, which will be our surrogate species for the thylacine—are anywhere between 12 and 14 days. After that, it matures in the pouch.

There have been studies showing that marsupials can be transferred from one species’ pouch to another species’ pouch and grow just fine. But we also have a team working [on] an “exo pouch.” This will be an artificial pouch that the pups can go in and have all the same nutrition, the same environment, same kind of light exposure that it would inside the pouch of a marsupial mom.

Colossal makes a point of saying it’s a for-profit company. What is the product, exactly? What will you sell?

I think there’s a couple of different ways that Colossal will profit. One of our products is the story. Right? We’re going to have a lot of partners in the media who are helping tell our story. Another is that as we develop new technologies along the way, these can be licensed or spun out. We had a first spinout called FormBio [a biology software company], and we also have a large staff of genome editors. 

And then we get to the real meat, which is the species: the thylacine or the mammoth. We are looking to partner with zoos. I think that there’s a world where we create rewilding habitats and sell tickets to go see these species in their natural area.

How much would you pay to see a thylacine?

Well, I’m putting hours and hours of my life into this. So I would honestly pay all the money in the world.