In September 2016, Jennifer Doudna called a new colleague named Kyle Watters to her office. By then, the University of California, Berkeley, biochemist was famous as the coinventor of CRISPR. The invention of the fast and versatile tool to edit genes had vaulted her to global notoriety and to considerable wealth. She was the founder of several startup companies and had collected millions in science-prize money.
Ominously, though, as Doudna has recounted, she was haunted by a dream in which Adolf Hitler appeared, holding a pen and paper, requesting a copy of the CRISPR recipe. What horrible purpose could Hitler have? Doudna, in her retellings of her dream, didn’t say.
Now Doudna’s question was, would Watters like to work on a way to stop it? Stop CRISPR.
CRISPR is found inside bacteria. It’s a billion-year-old defense against marauding viruses that spots their DNA and uses a scissors-like protein to chop it up. Doudna played a key role in transforming the find into a revolutionary gene-editing tool that’s been taken up worldwide, propelling a wave of new research and potential cures.
But if scientists learn to deliver gene editors inside people’s bodies, what’s to stop a madman, terrorist, or state from employing CRISPR to cause harm? People imagine personalized attacks that would strike only at certain ethnic groups or super soldiers edited to feel no pain. Doudna was well familiar with the dilemma. In her book A Crack in Creation, she wrote that she feared gene editing could come to the world’s attention, as atomic power did, in a mushroom cloud. “Could I and other concerned scientists save CRISPR from itself … before a cataclysm occurred?”
Now she would have a chance. Earlier in 2016, the US intelligence agencies had designated gene editing as a potential weapon of mass destruction. That September, the Defense Advanced Research Projects Agency (DARPA) had jumped in, putting out a call for new ways to control or reverse the effects of gene-editing technology. The program, called Safe Genes, would end up with a budget of more than $65 million, making it one of the largest sources of cash for CRISPR research, aside from biotech startups developing new genetic treatments.
One problem, as DARPA saw it, was the lack of any easy-to-use countermeasure, undo button, or antidote for CRISPR. And the more powerful gene editing becomes, the more we might need one—in case of a lab accident, or worse. As UC Berkeley put it in a 2017 press release after Doudna, with Watters’s help, claimed part of the big DARPA contract, the university intended to build tools to counter bioterrorism threats including “weapons employing CRISPR itself.”
CRISPR weapons? We’ll leave it to your imagination exactly what one could look like. What is safe to say, though, is that DARPA has asked Doudna and others to start looking into prophylactic treatments or even pills you could take to stop gene editing, just the way you can swallow antibiotics if you’ve gotten an anthrax letter in the mail. Scientists under Doudna’s project say they are set to begin initial tests on mice to see if the rodents can be made immune to CRISPR editors.
“Can we shut off CRISPR?” asks Joseph S. Schoeniger, who leads one arm of the defense effort at Sandia National Laboratories, in Livermore, California. “That is what we are looking at. The basic concept is that this technology is coming along, [so] wouldn’t it be nice to have an ‘off’ switch.”
Alexander Heinl/picture-alliance/dpa/AP Images
By the time Doudna drafted her proposal to DARPA, other scientists already had one big clue for how to stop CRISPR. In the ancient struggle between bacteria and the viruses called phage that infect them, phage had developed their own antidotes to CRISPR. In fact, their genomes, it’s been found, harbor the ability to produce what is essentially CRISPR kryptonite—small proteins exquisitely tuned by evolution to disable the gene-editing tool. Scientists call these molecules “anti-CRISPRs.”
The first anti-CRISPRs were discovered in 2013 by a student at the University of Toronto named Joseph Bondy-Denomy. “It was serendipity. We stumbled onto the fact that some phages seemed to be resistant to CRISPR. When we put the phage into a cell, the bacteria couldn’t protect itself,” says Bondy-Denomy, now a professor at the University of California, San Francisco. He quickly zeroed in on one of the virus’s 50 or so genes as the reason. “We thought, wow, maybe this is turning off CRISPR.”
The number of labs studying such defenses is smaller than the number working with CRISPR. But anti-CRISPR is becoming a booming field in its own right. More than 40 anti-CRISPR proteins have already been found, many by Doudna’s lab. Other teams are having early success locating conventional chemicals that can inhibit CRISPR as well. Today, Amit Choudhary of Harvard Medical School, in Boston, also with funding from DARPA, reported he had found two drugs that prevent gene-editing when mixed with human cells. “The hallmark of any powerful technology is control,” says Choudhary. “It’s that simple.”
Researchers like Bondy-Denomy believe anti-CRISPRs could have a role in improving future gene-editing treatments, by giving researchers more precise control. For instance, a team in Germany recently showed if they combined CRISPR and anti-CRISPR, they could create an editor that will change DNA only in liver cells, not neurons or muscle.
Another application being studied is whether anti-CRISPR could create a safeguard against “gene drives.” The Bill & Melinda Gates Foundation is backing the development of a CRISPR tool that will spread though wild mosquitoes, causing their populations to crash, with the idea of preventing malaria. Others want to develop such gene drives in mice, so they can eradicate the rodents from islands without using poison.
But what if these experiments go haywire and lead to an extinction? Researchers think they can create organisms with anti-CRISPR programmed into their genomes so they’re immune. In an initial proof of principle, scientists in Kansas last year engineered yeast cells with anti-CRISPR to resist a gene drive. “If some North Korean lab comes at you with a gene drive to wipe out an economically important crop, you could have a transgenic crop that makes it [resistant]. That is the drawing board scenario,” says Erik Sontheimer of the University of Massachusetts Medical School.
The advent of the CRISPR tool starting in mid-2012 surprised scientists. Essentially overnight, ham-fisted ways of genetic engineering were replaced by a cheap, versatile, and programmable means of changing DNA inside any living thing. Forecasters whose job was to anticipate new dangers “totally missed” CRISPR, says Renee Wegrzyn, the biodefense scientist who runs DARPA’s program. The humbling failure to see the future quickly morphed into a “critical urgent issue for national security.”
That’s because researchers, doctors, and startups backed by venture capitalists began a race to learn how to deploy CRISPR inside plants, animals, and humans, using viruses, injections, nanoparticles, or electrical shocks. And the better they got at it, the more realistic some sort of novel biothreat could become.
By early 2015, Doudna had also started to question how CRISPR was being used in more-routine laboratory research settings. Some of the experiments looked dangerous—what if a graduate student was hurt? “We are pushing these technologies out into the world, and we are not accompanying them with the safety measures that should be in place,” Wegyrzyn told a gathering of the Long Now Foundation, in 2017, in San Francisco. “I really started to feel this sense of urgency that someone needed to do something about this.”
In her talk, Wegyrzyn said the danger of CRISPR was obvious from how scientists were already using gene-editing to make mice sick by snipping important genes. “I don’t think you need to be a biosecurity expert to recognize that there is a need for scrutiny when you look at a tool that can both cure and cause disease,” she told the California gathering. “If we need to shut down a gene editor immediately, we just don’t know how to do this.”
There’s still no agreement about how dangerous CRISPR could be in the wrong hands. “Red team” exercises sponsored by the Central Intelligence Agency over the summer of 2016, where a group of analysts called the Jasons were asked to dream up their worst ideas, didn’t settle the question. Later, the National Academies of Sciences, Engineering and Medicine, at the request of the Department of Defense, produced an entire ranking of possible threats from synthetic biology, putting CRISPR weapons toward the middle of the pack. The military said it saw no imminent danger to soldiers.
Doudna agrees that CRISPR’s dangers should not be overblown. “I get these questions a lot about CRISPR systems and nefarious uses, and my feeling is that I am no more or less worried about CRISPR than other things. Someone could synthesize the smallpox virus,” she says. Similarly, while her research may lead to an eventual gene-editing antidote, her lab’s work with anti-CRISPRs is mainly addressing fundamental biological questions. “I am still at step one,” she says. “How do these work?”
Others, though, worry the risks are already apparent and that antidotes can’t come soon enough. For instance, some scientists have sought to prevent public discussion of specific CRISPR studies, or even delete mention of them from the internet, presumably to allow scientists more time to develop countermeasures. “The general prevailing attitude is not to give people nightmare fuel while we are actively looking for answers. There’s always a concern about an early freak-out,” says Doudna’s former collaborator Watters, who in 2018 authored a review of gene editing’s implications for biosecurity.
A video showing CRISPR editing DNA in real time
Osamu Nureki, Nature Communications
This year, as part of Doudna’s DARPA project, teams of scientists plan to begin their first experiments—in mice—to determine if it’s possible to protect them from CRISPR. One lab involved in the work is at Sandia National Laboratories, which will employ mice primed for editing because they are engineered to be born with CRISPR’s molecular scissors, a protein called Cas9, in every cell.
Schoeniger, who leads the Sandia effort, says soon his lab will instruct the mice to edit themselves but will also give them a shot of anti-CRISPR molecules, to see if the process is blocked. “Anti-CRISPR works well in nature, and we are trying to see if it works well in animals,” he says.
Schoeniger believes there is a “significant risk of accidental exposure” to CRISPR agents. As a large industry leaps up around the editing tool, CRISPR is being formulated into gene therapies, injections, ointments, and food, which raises the chance of a laboratory accident. Even a secret bioweapons program is more likely to release a designer germ by accident than it is to launch an attack. “As people use this in bigger and bigger amounts, there is an increased chance of people coming into contact, of getting stabbed or sprayed,” he says. “And if I get a mutagen sprayed in my eyes, it would be nice to stop it.”
Work on an antidote might also be helpful just as public relations. It could, at the very least, “tamp down the mental accessibility to a malign personality,” Schoeniger says. “If you can turn it off, maybe they won’t bother. From a psychological point of view, it’s nice to have an ‘off’ button. It’s nice for positioning that technology in society.”
Schoeniger isn’t under an illusion that an antidote to CRISPR will make threats go away. In fact, the security problem is growing, as laboratories improve the tool and invent related ones, each with different implications for biodefenders. Scientists can feel baffled by the tremendous speed at which gene editing, and synthetic biology more broadly, is speeding up, and how information is spreading online.
“We look at the overall risk front of the technology, how it keeps evolving, and how hard [it is] to stay on top of it, and how fast people are throwing out scenarios, and it’s hard to rationally address that risk,” Schoeniger says. In the meantime, he says, learning how to block CRISPR, in its classic, simplest form, seems like a good place to start. “It seems obvious we would like to modulate the technology, so let’s do that while trying to sort out the priorities,” says Schoeniger. “To a certain extent, it’s a mess; new technology is exploding so fast.”