search slide
search slide
pages bottom

‘Gene drive’ breakthrough creates weaponized mosquito extinction strain

Here’s a question that occurs only to madmen and geneticists: How do you get a gene that kills an organism to spread through a whole population of that organism? You can either make your gene deadly, and thus impossible to pass on, or not, and thus useless as a vector of attack. The solution has long been to try “silent” genes that can spread with no negative effects, either introducing a deadly weakness to a man-made chemical we withhold for a while, or by waiting for deadly activation by such a chemical. But recently, with the advent of advanced new in vivo gene editing technology, it’s become possible to make genes that seem to defy evolution — and that means we could soon start releasing animals carrying doomsday genes that spread with astonishing speed, quickly killing entire populations.

Such an animal is currently sitting in a laboratory at Imperial College London, an apocalypse mosquito carrying a gene that could one day end its entire species. It represents a controversial proposal to end the scourge of malaria, which kills hundreds of thousands of people each and every year, by wiping out the mosquitoes that spread the disease. It also represents a fundamentally new ability for humanity: the power to easily and selectively snuff out a subcategory of life on Earth. The name for this power is called gene drive.

Gene drive is simply the use of some strategy to artificially increase a gene’s inheritance rate. Such strategies are found all over nature, but despite decades of theorizing, nobody had a really viable way for mankind to harness this functionality through biotechnology. That’s changed thanks to the incredible advances in direct gene editing we’ve seen over the past half-decade, in particular the CRISPR/Cas9 gene editing suite.

These “molecular scissors” are actually borrowed from viruses, allowing scientists to swap out a gene in a living organism for one of their choice, edit it right into the genome so it will be passed on as the cells reproduce. If you can get your gene spliced into the “germ cells,” the pre-sperm or -egg cells of these organisms, then you can even introduce a chance that it will be passed on to the next generation — classically, without gene drive, you can introduce a 50% chance.

The chance is 50% because germ cells, like virtually all other cell types in humans and mosquitoes, have two copies of our genome. When we splice in our attack gene, it will end up sitting across from a second, totally normal copy of the gene it just replaced. This means that when the two copies get pulled apart to form the half-genomes of two new, separate sperm cells, only one of those new sperm cells will have our spliced-in sequence. The other will carry the same gene it would have, regardless.

So, if our spliced-in gene lowers evolutionary fitness, then all that will happen is the other half of the offspring will thrive, and the infected individuals will be quickly bred out of the population. And even if it’s a seemingly harmless silent gene that does nothing at first, it will still spread too slowly to change the overall population much at all (see the image above).

Our mosquito doomsday device gets around these problems by applying two innovations.

First, it forces itself into 99% of a mosquito’s sperm cells, and thus into 99% of its offspring. It can do this by exploiting the natural process of genome proof-reading. Once our synthetic gene has been spliced in to replace a target gene, we can design the system to intentionally damage the other, natural copy of that target gene. Do enough damage, and the cell’s machinery shows up to repair it back to normal. But what’s normal? Well, the double-helix provides a template — and the repair enzymes end up using our spliced-in gene as the guide for what the natural version is supposed to look like.

Once the natural version of our gene has been “repaired” into the engineered version, both copies (“alleles”) of the gene have our man-made sequence. Now, when the germ cell divides into two sperm cells, both those sperm cells get our modified version of the gene. So now, no matter which of those sperm cells goes on to fertilize an egg, the resulting mosquito will inherit our inserted attack gene. And the germ cells of those offspring get the gene drive effect, carrying it on to the next generation, and the next, and the next.

But there’s still a problem: if infected mosquitoes give birth to 99% weakened mosquitoes, then those mosquitoes will simply get bred out by normal individuals from completely separate parents, and our attack will go nowhere. The key to this mosquito bomb is that while 99% of offspring get the engineered gene, that gene doesn’t cause any problems when there’s only one copy.

So, try to follow the counting here: Scientists splice one copy of their experimental gene directly into a germ cell, where it then destroys and replaces the other copy and makes itself into the cell’s sole version of that gene. Then, the germ cell splits into two sperm cells, each of which has one infected copy in its half-sized genome. This infected half-genome then fertilizes a non-infected egg through normal breeding with another mosquito, combining to make a new mosquito with one copy of our synthetic gene (from our infected male’s sperm cell) and one copy of the natural gene (from the uninfected female’s egg cell).

doomsday mozzyNow, 99% of our infected male’s offspring are infected with a single copy of the gene we inserted, and are thus carriers who display no adverse effects. Most regular genes spreading through the mosquito population by natural processes have to spread without the gene drive ability to power them into 99% of offspring very quickly, meaning that even very advantageous genes won’t be able to spread as fast as our silent, seemingly useless one. With no downside to bias evolution against it, our gene will spread through the population in just a scant few generations.

Eventually, it will reach such a level of saturation in the population that these fully functioning carrier mosquitoes will begin to mate with one another through sheer chance, each donating infected sperm or egg cells 99% of the time, and thus giving rise to virtually all double-infected offspring. It’s these offspring, with both copies of the gene infected from conception, that express our genetic attack: the females of such mating events are completely sterile, while the males are free to continue breeding and passing on the disease.


What this means is that by the time the gene drive starts creating a substantial number of sterile, double-infected mosquitoes, the overall population will already have been infected too heavily to breed it out. With gene drive to keep the gene spreading in spite of evolution, the process should continue until there are simply no viable female mosquitoes left to breed with. By the Imperial College team’s calculations, this gene drive approach could completely destroy a population of mosquitoes in as little as 11 generations, or about a year.

Now, it’s also possible that the gene drive approach will be a total failure — we can’t know until we release it. In the lab, in small isolated populations of mosquitoes, it mostly seems to cause exactly the population collapse we would expect. Some of these populations dipped and then quickly recovered, however, likely due to some compensatory mutation that offsets our gene’s sterility effect. More research is needed, but the effects are already impressive, and the potential is undeniable.

Of course, simply because we can release this strain doesn’t mean that we should.

Some ecologists have argued that mosquitoes are totally replaceable in most environments and that killing them all wouldn’t cause some catastrophic reaction in the food chain. Others simply refuse to accept this, pointing out that we have only very partial understanding of these food webs and can’t accurately predict the second- and third-order consequences of this sort of enormous intervention.

It’s a tough decision. Do you accept hundreds of thousands of deaths every year without using every technology that might be able to stop it? Or do you risk destroying potentially far more than is at risk right now, helplessly watching as your attempt to help people ends up victimizing them even further?

Right now, we still have some time to decide while the pure research continues. But soon enough, the decision point will arrive: should we release this sort of technology into the wild? And, the more serious question: When these sorts of abilities are just a few commercially available enzymes away, how could we ever stop someone with an increasingly common skill-set from deciding to make one and release it, entirely on their own?

Now read: What is gene therapy?

Leave a Reply

Captcha image