The atom bomb took more than 10 years to go from scientific conception to reality. The next weapon of mass destruction could take much less.
On Thursday (Nov. 19), a special agent from the FBI met with researchers in Washington, DC to talk to a scientific panel about the risks of a powerful new genetic technology: “gene drive.” It allows scientists to, essentially, hijack the process of evolution, spreading a new gene through a population with incredible speed. And while it was developed with peaceful uses in mind, such as eradicating mosquitoes to end malaria, it could be used for ill too—it’s cheap and easy enough to master that bioterrorists could get their hands on it.
Gene drives have existed in nature for a long time. Normally, an organism has a 50% chance of inheriting any given gene from each of its parents. But certain genes can increase their own chances of being inherited. One way they do it is by having a mechanism that lets them make multiple copies of themselves in the parent’s genome.
For decades, nobody knew how to replicate this natural phenomenon. In 2003, Austin Burt, a professor of evolutionary genetics at Imperial College London, proposed a method for doing so, but his idea required a genetic tool that could precisely home in on one gene and replace it with another gene, along with the copying mechanism. At the time, no tool of required precision existed.
But then in 2012 US researchers developed just such a high-precision gene-editing tool—CRISPR-Cas9, which has opened up whole new realms of genetic engineering. By the end of 2014, researchers at Harvard University had used it to develop a gene drive in yeast that was inherited 99% of the time instead of 50%. In March, a research group at the University of California, San Diego showed that gene drives could work with 97% inheritance in fruit flies, a much more complex organism than yeast.
Genetic modification involves replacing one gene with another—to give mosquitoes a gene for resistance to the malarial parasite, for instance. Normally, a mosquito with that gene doesn’t always pass it on to the next generation. A gene drive, correctly programmed, could in theory both spread the parasite-resistant version of the gene to all the mosquito’s offspring and overwrite the non-malaria-resistant version, ensuring that the resistance never goes away.
But gene drives could also be used to build a terrifying bioweapon. In theory, a terrorist wouldn’t need to create vast amounts of a lethal virus to unleash on the world. Instead, he could create a handful of mosquitoes with a gene for making a toxin, and power it with a gene drive. Soon all the world’s mosquitoes would make the toxin, and every mosquito bite would be lethal.
Hence there is talk now of regulation. Thursday’s meeting was the last of six organized by the US’s National Academy of Sciences to discuss how gene drives might be regulated. The aim is to deliver a report next spring. Other security experts have briefed the UN’s bioweapons office about the potential risks of gene-drive-based weapons, and the US’s National Science Advisory Board for Biosecurity is likely to start looking into them too. This might lead to such things as a moratorium on federal funding for gene-drive research, much like the one the US government last year imposed on research into modifying viruses.
However, Amesh Adalja, a biosecurity expert at the University of Pittsburgh who was also present at Thursday’s meeting, told the health site Stat, “if a lone wolf or terrorist group is working on this, the regulation wouldn’t make any difference.”
But Austin Burt, who proposed the theoretical method for making gene drives in 2003, has said—and told Quartz—that the threat of gene drives isn’t as great, or as near-term, as it might seem.
First, gene drives can work only in sexually reproducing species. We’ve become very good at genetically engineering microbes, for instance, which can be used both to make useful compounds like synthetic drugs, or to create nasty diseases. But the vast majority of microbes (and plants, another biotech workhorse) reproduce asexually, so you couldn’t use gene drive to turn them into rapidly-spreading superbugs.
Second, the current gene drives have so far been shown to work only for one generation. Over several generations, natural mutations might destroy the copying mechanism, making the new genes peter out.
Third, we still understand only a few organisms at the level of detail needed to engineer a successful gene drive. It’s taken biologists decades of work to get to know fruit flies as well as they do. Bioterrorists won’t be able to simply adapt the idea to a mosquito or any other creature they please.
Still, scientists wanting to use gene drives for good aren’t taking chances.
To that end, Harvard researchers recently developed two possible safeguards. In a study published on Nov. 16, they show the use of a “split drive” in yeast, where some parts of the gene drive are inserted in the yeast’s DNA, and others are carried as separate strands in the yeast cell. That way, if by mistake the yeast gets out in the wild, not all the components will be inherited together, thus slowing the gene drive down.
They also developed a second gene drive—a ”molecular eraser” that undid the genetic changes in 99% of the offspring that had inherited the first gene drive. So one way to regulate gene drives might be to allow their use in the wild only if they have an “undo” button like this one.
Scientists have reason to be cautious. They don’t want a public backlash like the one that has surrounded genetically modified organisms. “If anyone messes up and a gene drive gets out into the wild, there will be a huge media circus,” Harvard researcher Kevin Esvelt, who developed the first synthetic gene drive in yeast, told Nature. “The message will be that scientists cannot be trusted to deal with this technology, and we will be set back by years.”