"Gene drive." Learn the term. Because it could one day transform the world.

Imagine if we programmed the mosquitoes that carry malaria to self-destruct. Not just a few of them. All of them. Every last bloodsucker on the planet.

That's no longer an outlandish fantasy. With a jaw-droppingly powerful new genetic editing tool known as a "gene drive," scientists could, in theory, transform entire species. I'll explain the mechanics below, but the basic idea is that scientists would alter the DNA of a few individual organisms and use gene drives to spread the modification throughout an entire population — by tweaking the rules of genetic inheritance.

Ponder the possibilities: We might eliminate malaria and save millions of lives. Modify ticks so that they no longer carry Lyme disease. Eliminate weeds that overrun our farms. Neuter the invasive lionfish devouring ecosystems in the Caribbean. We might even, one day, alter endangered species to make them more resilient and better able to survive modern-day stresses:

(National Academies of Sciences, "Gene Drives on the Horizon")

But also ponder the potential risks and dangers: We might wipe out an entire species only to learn later that it was vital in some unforeseen way. We might modify a pest only to find out that it emerges stronger than ever. Once a gene drive starts spreading throughout a species, it's hard to stop — so far, scientists haven't discovered an ideal kill switch. This is serious business.

That's why, this week, the National Academies of Sciences urged caution in a big report outlining guidelines for researching this new technology. Gene drives are incredibly promising, the panel argued, and merit further research. The potential benefits are too huge to ignore. But we're not yet ready to deploy this stuff. Our understanding of what might happen if a gene drive-modified organism were released into the wild is still far too murky.

The basics of how gene drives work

Scientists have long been able to modify the DNA of plants, insects, and animals. Think of corn modified to be resistant to rootworm, or Oxitec's mosquitoes programmed to die unless they receive antibiotics.

Yet these alterations are unlikely to dominate a wild population because of the usual rules of genetic inheritance, discovered 150 years ago by Gregor Mendel. If a fruit fly carrying a modified gene mates with a wild fruit fly, there's only a 50-50 chance that the offspring will inherit the modified gene. There's an even smaller chance that the offspring's offspring will carry it. And so on, down the line.

Gene drives change those rules of inheritance. Using advanced genetic editing tools known as CRISPR/Cas9, scientists have been able to modify fruit flies so that there's nearly a 100 percent chance the modified gene gets passed down. When a GM fruit fly mates, virtually all of its offspring carries the modified gene. When those offspring mate, all of the offspring's offspring have it.

Within a few generations, everyone has the modified gene (in blue):

(Oye et al 2014)

In the lab, scientists have demonstrated gene drive systems for yeast, fruit flies, and mosquitoes. In one experiment, researchers modified a few fruit flies to be yellow, released them into a population of unmodified flies, and pretty soon all the fruit flies were yellow. Scientists have dubbed it a "mutagenic chain reaction."

How scientists might use gene drives to wipe out mosquito populations

So how might we use this technology? Earlier this year, I talked to Zach Adelman, a molecular geneticist at Virginia Polytechnic Institute and State University in Blacksburg, who is researching gene drive systems for Aedes aegypti mosquitoes, which carry Zika and dengue fever.

Specifically, he's looking at whether it's possible to modify mosquitoes so that they only produce male offspring. If this gene modification propagated across multiple generations and the entire population, the mosquito would eventually go extinct. Good-bye Zika vector.

(Aedes aegypti/Wiki Commons)

Adelman cautioned that researchers still have a ways to go before they can actually pull this off: "You need to be careful about the timeline here." Scientists haven't even described a gene drive system for Aedes aegypti yet. After that, they'd need to show that this "males only" gene could actually propagate through an entire population in a lab setting.

Researchers might also need to develop a fail-safe — a way to halt the gene drive if something went wrong. "This is an active area of research," Adelman said. "There are a few proposals for how to do this, but I find them unsatisfactory." For example, scientists have described a "reversal drive" that would propagate the original genes back through the population, but it's uncertain how well this would work.

Finally, once researchers proved the concept in labs, they'd want to run field tests first to see if they could actually crash Aedes aegypti populations in the real world. Do you just need to release five mosquitoes? A hundred? Thousands? In one place? Many? Mosquito populations are "clumpy," Adelman said. If your modified males all hung out in one region while a surviving group of females hung out elsewhere, the population would survive.

That prospect of a field test is what concerns many people. What happens if the mosquitoes escape and they turn out to be absurdly effective? And what happens if, say, the wild Aedes aegypti population crashes and this turns out to be a dire outcome, ecologically?

That's where this new report comes in. Because while no one is conducting gene drive field tests right now, that day may be coming sooner than we think. It's not too soon to talk oversight.

Why scientists are urging caution with gene drives

The National Academies of Sciences' report on gene drives is long, detailed, and utterly fascinating. But the basic point is that gene drive technology is developing faster than our understanding of the potential ecological impacts. There are dozens of lab experiments using CRISPR/Cas9 to create gene drives, but far less research on what it would actually mean to deploy a "mutagenic chain reaction" in the wild.

Before any tightly controlled field tests begin, the report argues, we need to better understand how a gene drive would propagate through a wild population of the target species, how it would affect evolutionary fitness, and whether modified genes might "flow" to other unrelated species. That will involve both modeling and empirical research.

And if we're going to use a gene drive to modify or eradicate (say) an entire mosquito species, we want to answer some crucial questions. Like: What role does this species play in nature? Would other species fill a similar niche if it vanished? How might a dramatic change in this species affect other, nearby species?

It's quite possible that using gene drives to exterminate certain mosquito species would be benign. Aedes aegypti, for instance, isn't native to South America; it hitched a ride over from Africa during the slave trade. And we nearly wiped it out during the 1950s with an intensive pesticide campaign (it eventually rebounded after eradication efforts let up). So perhaps we wouldn't regret it one bit. But it's worth double-checking.

Back in 2010, Janet Fang wrote a great piece for Nature about whether we'd miss mosquitoes if we exterminated them all. In some ecosystems, this might not be a worry — animals that relied on mosquitoes for food would just gobble up other insects that filled the niche.

But there could be exceptions. In the Arctic, where mosquitoes are extraordinarily abundant, migratory birds might decline in population if this key food source vanished. The loss of mosquitoes could alter the paths of caribou herds, which migrate great distances to avoid being bitten. That might in turn affect lichen or wolf populations in unknown ways. The point is, when you're messing with 100 million years of evolution, you want to be careful.

The NAS report recommends the development of strict "ecological risk assessments" before field-testing this technology — and notes that current US safeguards under the National Environmental Protection Act (NEPA) may not be sufficient. Essentially, scientists would want to better understand various cause-and-effect pathways in the environment and quantify the risks of various ecological outcomes before this technology gets out of the lab.

The report also points out (wisely) that the use and deployment of gene drives isn't a purely scientific question. They involve ethical and political questions, like: How much risk do we want to take? What do we want to use this technology for? Note that rejected gene drives and sticking with the status quo isn't always the "safest" option. The ability to eliminate malaria would be an extraordinary benefit for mankind. Millions of lives saved. If we could do it, should we? Is it immoral not to?

That's not a call that scientists should make alone, and the report stresses "the importance of engaging affected communities and broader publics in decision making about activities involving gene drives." Put differently, humanity is on the verge of wielding unprecedented power over nature. We should put some thought into how best to wield it.

Further reading:

• Here's a collection of expert responses to the NAS gene drive report. Among other comments, Kevin Esvelt of MIT made this interesting point: "Adding a gene drive changes our default expectation from ‘this alteration is unlikely to spread in the environment’ to ‘this is likely to spread in the environment’."
• In the Guardian, Jim Thomas argues that the NAS report missed some key topics — like how and whether gene-drive technology could be weaponized. This 2014 piece in Science offers some thoughts on how gene drives might need to be regulated going forward.
• If you like TED talks, Jennifer Kahn has a 12-minute one on gene drives.

Explanatory journalism is a public good

At Vox, we believe that everyone deserves access to information that helps them understand and shape the world they live in. That's why we keep our work free. Support our mission and help keep Vox free for all by making a financial contribution to Vox today.

One-Time Monthly Annual

$95/year

$120/year

$250/year

$350/year

Other

$

Yes, I'll give $120/year

Yes, I'll give $120/year

We accept credit card, Apple Pay, and Google Pay. You can also contribute via