̽����ѡ

The silent black-and-white footage opens on a seemingly tranquil setting: a burrow where an Atlantic Petrel tends to its chick. Then mice begin scurrying in and out of frame. The dark blurs jostle the adult, darting up to the exposed chick and tearing off bloody bites. They’re eating it alive.

The horrific scene is captured by nest cams on Gough (rhymes with “off”), a rugged volcanic island about 1,700 miles west of South Africa. It has one of the world’s largest seabird nesting colonies, with millions of birds representing 22 species. It’s also home to hundreds of thousands of mice, descendants of stowaways on 19th-century seal-hunting ships. The tiny predators devour some 900,000 chicks a year and threaten to decimate the island’s and , which breed here almost exclusively. In a rodent-free landscape, more than two-thirds of the albatross chicks should make it to adulthood; on Gough, mice cut survival to as low as 10 percent.

To stop the carnage, the U.K.’s Royal Society for the Protection of Birds is in 2019. Helicopters will navigate the steep terrain, strategically dropping 180 tons of food pellets laced with the anticoagulant brodifacoum; mice that consume the spiked pellets will hemorrhage to death. It’s a grisly solution to an intractable problem that extends far beyond Gough. Rodents have voyaged with humans to an estimated 80 percent of the world’s islands, where nesting seabirds were long safe from predators. Poison is the state-of-the-art method for wiping out the voracious pests. Yet it has serious drawbacks. It’s costly. It’s imperfect, failing in roughly a quarter of mouse-eradication efforts. (The survival of just one pregnant mouse can repopulate an island.) It’s indiscriminate, so inadvertent casualties are unavoidable. (That’s why zoologists will bring endemic Gough Moorhens—which may scavenge dosed mice—into captivity before they start the campaign.) It’s also the only option.

Frustrated with the macabre, limited approach, scientists have begun exploring a groundbreaking new tool that might one day replace it. The technology, called gene drive, offers the unprecedented ability to ensure that organisms pass specific traits to their progeny, like eye color or, in the case of invasive mice, sex. Island Conservation, that has undertaken dozens of poison campaigns across the globe, is part of an international consortium trying to harness technology to engineer mice that produce offspring of only one sex. If it works, conservationists could skip the poison and instead breed invasive mice out of existence.

Island Conservation and like-minded groups think the technology could transform seabird conservation. Other researchers are considering whether it could combat agricultural pests, or if sterile gene drive mosquitoes could save native Hawaiian birds from avian malaria or protect people from scourges like Zika virus. But tinkering with organisms could also go terribly awry, critics warn, permanently altering ecosystems in unintended ways. No gene drive animals have been released in the wild—yet. But the technology’s potential use is sparking fierce debate about whether to wield such a powerful tool.

When Charles Darwin visited the Galápagos in 1835, he encountered finches with an array of beak shapes. Each unique shape, he later posited, was adapted for a particular niche in foraging seeds, insects, and nectar across the varying landscapes. The observation helped solidify his theory of natural selection: Organisms that are good at navigating their environment will likely survive and reproduce, passing their traits to the next generation. These traits are linked to genes, and offspring have a 50/50 chance of inheriting a given gene from either parent.

But that’s not always how it works. Gene drives, sometimes called “selfish genetic elements,” hijack heredity. They push the inheritance of a specific��gene higher—sometimes to nearly 100 percent—even if it provides no evolutionary edge. (It isn’t entirely clear why this happens.)

Scientists contemplated leveraging this capability as early as 1940, when a Russian researcher suggested an approach to wipe out troublesome insects. Since then, scientists have mulled ways to manipulate a natural gene drive to do their bidding. In 2003, evolutionary biologist Austin Burt proposed using genetic engineering to build an artificial drive—one that could pinpoint a precise location in an organism’s DNA and insert a desired gene.

Biotechnology caught up in 2012, when an international team of scientists unveiled a revolutionary new way to alter genes. The technology, , is a set of biological editing tools: It can find a specific stretch of DNA, cut it like scissors, and add, delete, or replace genetic information. CRISPR makes it possible to harness natural gene drives, a process akin to building a car from spare parts. Scientists can also employ it to create artificial drives, like building a car with custom parts.

In 2015 geneticists used a CRISPR gene drive to change the color of lab-raised fruit flies from brown to yellow in one generation. Gene drive yeast and mosquitoes quickly followed, spurring public concern over the potential ecological effects engineered organisms could have in the wild.

With all the excitement, and despite the trepidation, it was only a matter of time before someone turned to more complex creatures. The mouse was an obvious choice. It’s among the most common lab animals, and scientists mapped nearly all of its genes in 2002. Island Conservation and its partners are exploring CRISPR for both natural and artificial gene drives. Each approach has an upside. Natural gene drives may be more palatable to the public and obtain federal approval more easily, since they already exist in mice. Artificial gene drives are more flexible because they can be designed from scratch. Both are in the proof-of-concept stage, and as yet there’s no clear winner.

Getting the gene drive to work is just one hurdle. Attraction is another. Engineered mice will have to mate with wild ones in order to pass on the drive. If wild mice rebuff lab mice, the technology won’t save any seabirds. Scientists need a sexy vessel to deliver a gene to the wild.

There is, in fact, a scientist devoted to making mice irresistible. John Godwin, a , spent most of his career studying fish and lizard sexual behavior. Then, in 2011, he read about the Farallons, rocky islands off the coast of San Francisco that have up to an astounding 500 invasive mice per acre in the summer. They also support one of the world’s largest breeding colonies of Ashy Storm-Petrels. The abundant rodents attract Burrowing Owls, which historically used the islands as a pit stop during fall migration but now overwinter there. When mouse numbers dip during winter food shortages and bad weather, owls dine on the storm-petrels.

CRISPR wouldn’t be unveiled for another year, but Godwin’s colleagues were already working on a natural gene drive in mosquitoes to curb dengue fever. Could a similar approach, he wondered, combat invasive mice? Given his expertise, Godwin was especially curious whether an engineered mouse would stand a chance wooing wild females. He convinced entomologist Fred Gould, co-director of NC State’s Center on Genetic Engineering and Society, and David Threadgill, a mouse geneticist with a focus on biomedicine, to take on Farallon mice as a test case. They arranged a conference call with the U.S. Fish and Wildlife Service, which manages the islands.

The agency also invited Karl Campbell, Island Conservation’s project director, on the call. Campbell was deep into a search for new conservation technologies; he’d heard of gene drives but didn’t know anyone in the field. The FWS gave Godwin the go-ahead to collect Farallon mice but ultimately didn’t consider a gene drive solution for its mouse problem. (It opted for poison, as it was already available, but public criticism has put the deployment in limbo.) Campbell, however, wanted in. The scientists were surprised; fairly or not, genetic technology isn’t always popular in conservation circles. But not engaging around gene drives would hardly ensure Campbell’s popularity; Island Conservation often faces protests and lawsuits from animal activists and environmentalists who object to its poison campaigns. “You’re a crazy man,” Gould recalls thinking, recognizing the backlash Campbell might experience. “And it takes a crazy man to do this.”

Island Conservation and NC State struck up a partnership: , or GBIRd. In late 2013, Godwin dispatched grad student Megan Serr to collect mice. She hitched a ride on a sailboat to Southeast Farallon. The rocky island has no dock, so she transferred in choppy, shark-filled waters to a dinghy, which a crane lifted to shore. Back home, while the 17 mice stayed in quarantine to be stripped of parasites, Serr, her husband, and Godwin constructed a 500-square-foot enclosure for the captives inside a university greenhouse. The Mouse Barn, as they call it, took about a year to complete and cost the amateur builders countless trips to Home Depot, quarts of sweat, and “a little bit of blood,” says Godwin.

Today around 100 Farallon mice inhabit the Mouse Barn, whose walls, ceiling, and door help prevent their escape. To enter, you walk through double doors, sign a log, and pull protective booties over your shoes to comply with university regulations. Immediately inside are rows of shoebox-size cages with wild mice. They flit through their bedding, peeking at visitors with bright eyes, and hang from the wire lids like kids on monkey bars.

When a lab mouse is placed in a cage with a wild mouse, they mate. But Godwin wants to see what happens when wild mice have options. At the Mouse Barn’s other end are five walled pens, each about the size of a pool table and scattered with wood chips, sand, cardboard tubes, and plastic nest boxes. They look like giant hamster cages, but the point is to offer an approximation of nature with places to hide and play.

Two males—one wild and one lab-raised, but not yet engineered with a gene drive—are put in a pen with four wild females. While blood tests will reveal paternity, Godwin’s team is interested in more than who the daddies are. Each mouse has unique ear tags, and radiofrequency ID tags will monitor who visits whom, and for how long. Spying on the critters should help reveal what, exactly, a mouse finds attractive—size, for example, or aggressiveness—and ultimately help scientists breed the sexiest lab mice.

While Godwin tackles mouse attraction, other GBIRd members have been investigating gene drives. Threadgill, who’s now at Texas A&M, designed a natural drive using CRISPR to bind two pieces of existing mouse DNA. The first is the T-complex, a group of genes that impairs sperm lacking those same genes, thus ensuring they’re passed to offspring. The second is the Sry gene, whose presence guarantees a mouse will be male. If the combination works, over time, every mouse on an island would be male, causing the population to die out. Paul Thomas, a mouse geneticist at University of Adelaide in Australia, is taking a different tack. He’s working on two types of CRISPR-based artificial gene drives: One forces all offspring to be female, and another makes embryos unviable, so females produce no pups. (Both teams work in biosecure labs and employ extra layers of protection so the drives aren’t likely to spread even if the mice escape: Wild mice in the Lower 48 have an apparent resistance to Threadgill’s natural gene drive, while Thomas’s gene drive only works on a snippet of DNA unique to lab mice.)

When engineered mice like those in Texas or Australia are ready, they’ll be tested with wild mice at a more secure venue. Another GBIRd partner, the USDA Animal and Plant Health Inspection Service, has a large biosafety lab in Fort Collins, Colorado, with secure enclosures that are far closer imitations of nature than the Mouse Barn’s giant hamster cages.

If the gene drive mouse is indeed alluring to potential mates—and if it passes a litany of safety tests and ethical considerations—GBIRd will advance to the next phase. They’ll collect mice from an as-yet-undetermined island, bring them to a biosafe facility, and start the attraction tests anew. Assuming they’re successful, only then will they deploy gene drive mice. The exact number they’ll need isn’t clear: There’s a delicate balance between releasing enough so the drive works, but not so many that the mice wreak environmental havoc and die off from food shortages before they take hold and start reproducing.

Island Conservation initially estimated that it would release gene drives by 2020. It was, they figured, enough time to engineer the mouse, obtain permission to legally and safely release it in the wild, and address critics’ concerns. They expected the timeline would generate excitement. It did the opposite.

On a balmy December day in Cancún, Mexico, staff members of Island Conservation met with the most vehement critics of biotechnology in the conservation movement. They were attending the 2016 UN Biodiversity Conference, and they proposed a neutral place—the convention center lobby—to plead the case for gene drive research. Island Conservation’s audience included representatives from Friends of the Earth (FOE), an environmental advocacy group, and ETC Group, a Canadian biotechnology watchdog. Both groups call gene drives “genetic extinction technology” and are actively trying to block GBIRd’s efforts. From GBIRd’s earliest days, Campbell expected regulatory hurdles and public concern about releasing engineered mice. No legal framework exists for the tech, and no previous engineered organism—such as GMO crops—can potentially spread as rapidly and intentionally through an ecosystem. From the beginning GBIRd has sought help for navigating those uncertain waters. Partners at the USDA are working with federal regulators to understand what laws may apply in the United States. Landcare Research, a New Zealand federal science institute, is exploring regulations abroad, and the Commonwealth Scientific��and Industrial Research Organisation (CSIRO), Australia’s federal scientific research institute, is studying potential risks. Landcare, CSIRO, and NC State are all committed to engaging with communities on or near islands where the mice may be released. NC State also has an interdisciplinary team of graduate and postdoctoral students who study biology, math, social science, and communication analyzing practical and ethical implications. It’s about more than mice, says Elizabeth Pitts, a postdoc in rhetoric at NC State’s Center on Genetic Engineering and Society: Will meddling in places we’ve already altered dramatically help or harm? How do you obtain consent from people who may be affected by the tech? And if we hit the reset button on an island—is the ideal we’re trying to reset to ten years ago? A hundred? A thousand? Or as Pitts puts it: “How do you like your nature?”

In Cancún, Island Conservation’s team and its detractors agreed on some points. Seabird conservation, and island conservation more broadly, is important. Gene drive is a powerful tool that requires great care. But the cordial dialogue didn’t resolve a key conflict: whether it’s appropriate even to study gene drives.

The divide extends beyond these groups. A June 2016 National Academies of Sciences, Engineering, and Medicine report concluded gene drives are promising, but too new to be released yet—if ever—and called for more lab research and public engagement (as GBIRd is doing). In September, the International Union for Conservation of Nature (IUCN) voted to develop policy on gene drives and related technologies and to refrain from endorsing lab research and field tests until it finalizes��its assessment in 2020. Also in September, 30 activists, including environmental heavyweights Jane Goodall and David Suzuki, endorsed a call “for a halt to all proposals for the use of gene drive technologies, but especially in conservation.” And at the United Nations conference in Cancún, FOE and other opponents tried—and failed—to convince world governments to temporarily ban gene drive research.

Because gene drive animals could spread across borders, Dana Perls, a FOE senior campaigner who met with Island Conservation, wants research to pause until international regulations are in place. “[The technology] has the potential to cause incredibly significant damage to ecosystems and to species,” she says. “We need to slow way down and look at the potential risks.”

GBIRd partners agree that a cautious approach is needed, but they also believe that a moratorium could seriously delay the science. Members have largely footed their own bills so far, but the group will need to raise up to $12 million per year to conduct advanced trials and push the project past proof-of-concept. A ban could scare off funders, setting the research back years.

Other scientists exploring the technology fret about the unknowns. “I’m deeply concerned about how well we can predict how [gene drives] evolve outside of our control,” says Kevin Esvelt, an evolutionary engineer at the Massachusetts Institute of Technology who is considering whether gene drive��mice could help in other conservation projects. While mice released on an island are less likely to spread globally than, say, winged insects, Esvelt worries people may steal and release them in untested areas—on the mainland, for instance, in an attempt to check rapacious rodents that diminish crop yields. “My concern is, what would that do to public confidence in scientists and in this technology?” he says. “Are we really willing to risk an accident that might damage the chances that we might be able to use gene drive technology against malaria?”

̽����ѡ hasn’t taken an official stance on the technology. “I personally wouldn’t want to call an end to research, because you don’t really know what will come out of it,” says Steve Kress, vice president for bird conservation. “But I also think there is a place for voices of concern, which will help guide the research.”

Waiting for cohesive international rules could mean waiting forever. Regulations on existing genetic engineering technology are messy, with rules ranging from country to country, and gene drives won’t make regulatory questions any easier. In the United States, for example, the relevant laws are three decades old and therefore ill fit for new biotechnology. To address the shortcoming, in the Obama administration’s final days the U.S. Food and Drug Administration—which regulates biotech, along with the USDA and EPA—proposed changes that might cover the development and release of CRISPR animals. So far the Trump administration hasn’t given any indication about what action, if any, it will take.

No one is dropping mutant mice on a remote island anytime soon, if ever. Late last year Island Conservation nixed its 2020 timeline—or any timeline. “We have no plans for release,” says Heath Packard, the group’s communications director. Still, they’re forging ahead with a set of criteria to define the ideal field test location. It’ll likely be small and simple—not a mountainous island like Gough, with its sweeping cliff coasts and complex landscape.

GBIRd is also forming an independent ethics panel to provide feedback at each major milestone, including mating a gene drive mouse and wild mouse and selecting the pilot study island. The group has to be willing to pause, halt, or change course if the ethics or public perception demands it, says Jennifer Kuzma, co-director of the Center on Genetic Engineering and Society who advised GBIRd on convening the panel. “We don’t want the public engagement and the ethics to just be window dressing.”

And they certainly don’t want to repeat the Monsanto effect. In the 1990s, when the biotech giant and competitors sold their first GMO crops, they didn’t seek much public input. Instead, they identified an agricultural problem, built a solution, and released it. The backlash was fierce and sustained: 21 years later, activists and many consumers are still wary about the products, and a non-GMO food industry has emerged in response. Gene drive researchers wield an even more powerful biotechnology and have learned from Big Ag’s mistakes. Given the technology’s promise, and the fact that it’s already being used in labs around the world, research will likely continue—with a lot of stops along the way.

For many seabirds, this progress may be too slow. Invasive animals and plants threaten around 387 island avian species, according to BirdLife International. The house mouse is still spreading, and many birds teeter on the edge of survival: If poison fails on Gough, for example, the Tristan Albatross could go extinct by 2040. And so conservationists continue to fight back with the tools at hand, even as they wish for something better.

Correction: This story previously stated that��Kevin Esvelt is considering��whether gene drive mice could curb Lyme disease on Nantucket; that��project uses a different type of genetic pest control.