12 September 2011 by Henry Nicholls
We can now genetically modify animals to kill off their own kind, while leaving other species unharmed
Editorial: "Give geo- and genetic engineering a fair trial"
IN THE urban jungle of Juazeiro in Brazil, an army is being unleashed. It is an army like no other: the soldiers' mission is to copulate rather than fight. But they are harbingers of death, not love. Their children appear healthy at first but die just before they reach adulthood, struck down by the killer genes their fathers passed on to them.
These soldiers are the first of a new kind of creature - "autocidal" maniacs genetically modified to wipe out their own kind without harming other creatures. The first animals being targeted with these "living pesticides" are disease-carrying mosquitoes and crop-munching caterpillars, but the approach should work with just about any animal - from invasive fish and frogs to rats and rabbits. If it is successful, it could transform the way we think about genetically engineered animals.
In essence, much the same method has already been successfully employed for more than half a century. In the so-called "sterile male technique", large numbers of the target pest are bred, sterilised and the males let loose. When they mate with wild females, the resulting eggs are not viable, so releasing enough sterile males can eventually exterminate wild populations.
Infographic: Compare the sterile insect and autocidal techniques
This method is widely used and has notched up many successes. For instance, it helped to eliminate the screwworm fly from the US and other areas - the fly's larvae burrow into the flesh of livestock and even people. It also helped clear the tsetse fly from Zanzibar.
And it is not limited to insects: a similar approach is being used to try to control an invasive parasitic fish in the American Great Lakes. Male lampreys are being trapped, chemically sterilised and released.
The sterile male technique has the huge benefit of being incredibly focused, homing in only on the species you want to control. Pesticides, by contrast, harm a wide range of other species, including us.
So why isn't the method more widely used? The main problem is that it is very difficult to sterilise animals without harming them in other ways. The usual way of sterilising insects is to zap them with radiation, for instance, which leaves the males weakened. Establishing the optimal dose of radiation for a species is thus crucial - too little and fertile insects will be released, too much and the males will be too feeble to compete for females. Working out the optimal dose is best done in the field, but the task is laborious without an easy way to distinguish sterilised insects from wild ones.
Enter Oxitec, a biotechnology company based just outside Oxford in the UK. It has created a pink bollworm - a moth caterpillar - with a built-in fluorescent marker called DsRed. The bollworm is a major pest in cotton fields, and in 2002 an eradication campaign was launched in the US, part of which includes releasing sterilised bollworms. In 2006, Oxitec's fluorescent bollworm became the first genetically engineered animal to be deliberately released into the environment. Over three years of successful trials, more than 20 million moths have been released in the US so far.
These modified moths are just the start. When the founder of Oxitec, Luke Alphey, first learned about the sterile insect technique from a colleague in the 1990s, he realised that the molecular tools he was using in his everyday research might provide a better alternative. Within a matter of years, he had created fruit flies with genes that kill their offspring (Science, vol 287, p 2474).
In theory, unlike zapping animals with radiation or chemosterilisation, the genetic approach should work well with just about any species. Besides the pink bollworm, Oxitec is targeting the mosquito Aedes aegypti, the single most important carrier of dengue, a viral disease that affects 50 to 100 million people in tropical regions every year, including a few in Florida and Queensland, Australia. Usually the symptoms are mild, but around 1 in 20 people become seriously ill. There is no vaccine and no treatment, so the only way to combat the disease is to kill the mosquitoes that carry it - and they are becoming resistant to pesticides.
Alphey and his colleagues have created a strain of A. aegypti with two copies of a gene that disrupts the development of offspring. The gene is switched off in the presence of the antibiotic tetracycline, allowing large numbers of perfectly fit mosquitoes to be bred for release. "With our system, the mosquitoes are fundamentally sterile and we're keeping them alive by giving them an artificial antidote," says Alphey. The insects also have the DsRed marker gene, to enable them to be easily monitored.
When these mosquitoes mate with wild females, the eggs hatch and the larvae develop normally until they reach the pupae stage, when the killer genes kick in. Delaying death like this is actually a cunning trick: the doomed larvae compete with wild larvae for resources, further reducing their numbers.
In 2009, Oxitec began a series of field trials of this strain of A. aegypti in the Cayman Islands in the Caribbean, making it the first engineered mosquito to be let loose in the wild. The trial showed that the engineered mosquitoes survive and disperse from the site of release, and compete well for mates.
With the Oxitec mosquitoes accounting for just under 20 per cent of all males in the population, around 10 per cent of the eggs produced contained the engineered genes. "We got about half as many transgenic offspring as you would have expected had everything been equal," he says. "But this is way more than you need for success." The mosquitoes also performed well in a small trial in Malaysia, says Alphey.
Now a bigger trial is getting under way in Juazeiro, Brazil, which Alphey hopes will be scaled up into a full-scale control programme. In the meantime, Oxitec has been busy developing other, more sophisticated strains, including one particularly devious one.
With both the classical sterile-insect method and the genetic variations on it, it is normally vital to release just males. If you release males and females at the same time, they will mate with each other, reducing the impact upon the wild population. Unfortunately, separating the sexes takes a lot of time and effort in some species. "This is a huge problem if there is a need to release millions of males per week," says Mauro Marelli, an epidemiologist at the University of São Paulo, Brazil, who is working with Oxitec on the trials in Juazeiro.
With the tsetse fly, for instance, separation by hand is common. Separation is much easier with A. aegypti, as the female pupae are quite a bit bigger than the male pupae. Nevertheless, the process still requires dedicated facilities at the release site, which makes it expensive. Transporting adults is not feasible. "Adult mosquitoes are all spindly and if you pack them into any kind of space, you end up with legs tangled up with wings and a lot of physical damage," says Alphey.
Instead, his team has created a strain in which the females cannot fly. The work was based on the discovery that female mosquitoes have a unique flight muscle protein that males lack, perhaps because females have to fly after a blood meal and so must fly with a much heavier load. Flightless females cannot find people to feed on and cannot mate either, so there is no need to separate the sexes. Envelopes containing millions of eggs could simply be mailed to wherever they are needed. "Just add water and you get instant mosquitoes," says Alphey.
The males that hatch from the eggs will appear normal and can pass the flightless gene to their daughters. Their sons will also inherit a single copy, so they too will produce some flightless daughters. "The construct will persist in the population for several generations but not for long due to its high fitness cost," says Alphey.
Oxitex also has malaria-carrying mosquitoes in its sights, but this is a greater challenge than dengue. For a start, there is often more than one malaria-carrying mosquito species responsible for transmission in a particular area, so effective control would mean targeting each these species separately. In addition, malarial mosquitoes often bite other animals besides humans, so their distribution is less predictable than A. aegypti's.
For malaria then, this kind of technology is unlikely to be as effective as for dengue but it could still be helpful. "It'll turn out to be extremely valuable in some places, a piece of the jigsaw puzzle in other regions and perhaps not that relevant in other areas," says Alphey.
While Oxitec is leading the way, many other groups around the world are working on similar approaches - and not just for killing insects. "On technical grounds, there's no reason why the logic of sterile insects could not be transferred to vertebrates," says Ronald Thresher, an ecologist working for Australia's national scientific agency, CSIRO.
He thinks the autocidal approach could not only be used to control invasive species such as cane toads, but that it is the only method that could work in many cases. "It's the only hope we have for the long-term control and eradication of these pests," he says. "Other efforts help, but in the end they are Band-Aids in the absence of a real solution."
Thresher has come up with a way to create fish that produce only male offspring. Releasing enough of these "daughterless" fish into the wild, with each passing on the daughterless habit, would turn a thriving invasive population into a bunch of reluctant bachelors destined for extinction.
His method relies on the fact that an enzyme called aromatase is crucial for generating female hormones in fish. Switch off the aromatase gene and you've created a fish that can only produce sons. He has shown the approach works in lab tests on zebrafish, skewing the sex ratio in favour of males for at least three generations. The plan is to tackle carp, an invasive fish blamed for the decline in native fish species and the erosion of riverbanks across the vast Murray-Darling river basin in south-east Australia.
Models suggest that releasing enough daughterless carp to make up 5 per cent of the total population would effectively eradicate carp in the Murray-Darling basin by 2030.
Thresher's models also suggest pests such as cane toads and rats could be tackled this way. However, breeding large animals is labour intensive. "So the expense of such a programme quickly becomes an issue," says Thresher. Public acceptance could also be a huge issue, Alphey points out. "A large number of adult male rats - sterile or not - is probably not the way you want to go."
Nevertheless, if autocidal technology lives up to its promise, it could be about as environmentally friendly as pest control can get. It could largely or entirely replace pesticides, and it affects only the target species. Last but not least, it is hard to see what could go wrong.
Many engineered plants, for instance, are being given advantageous traits such as disease resistance, so these genes could well spread among wild relatives. Autocidal traits, by contrast, are a great disadvantage and should disappear from the wild within a few generations after releases stop. "We are putting genes with huge, huge fitness penalties like death into something that's undesirable in the first place," says Alphey.
In theory, wild insects might be able to evolve resistance, for instance, by somehow learning to recognise and avoid insects with lethal genes. But this is much less likely to develop than pesticide resistance, and could be overcome by altering the release strain.
Needless to say, those opposed to genetic engineering are not convinced. "Genetic modification leads to both intended and unintended effects," says Ricarda Steinbrecher of EcoNexus, which describes itself as "a not-for-profit, public interest research organization". "There are potential knock-on effects on many other organisms," she claims.
Most biologists, though, agree the risks are minimal. "It is true that some of the regulations are being put together as the programmes are moving along, but the risks are really very, very small," says Mark Benedict, an entomologist at the University of Perugia in Italy. For him, the big question is whether it is a cost-effective approach.
The risks also have to be weighed against the potential benefits. Dengue, for instance, is spreading rapidly in the tropics. There are promising vaccines in trials and several potential antiviral candidates, but engineered mosquitoes are potentially a very powerful way of preventing the disease, says Jeremy Farrar, director of the Oxford University Clinical Research Unit in Ho Chi Minh City in Vietnam. "We need to really push the developments, be hard-nosed about assessment of what works, and ensure that what does work gets implemented."
Much could ride on the success or failure of autocidal technology. So far, most GM organisms have offered few visible benefits to consumers. Opposition to their use remains widespread, particularly in Europe. But if they are seen to save lives by helping control pests and diseases, opinions could change.
That could open the door to other, even more ambitious genetic approaches to pest control. One that has particular promise exploits chunks of "selfish" DNA that can spread themselves through the population and kill only when two copies are inherited. In theory a one-time release of just a few insects, rather than the continual release of millions, could wipe out a wild population (New Scientist, 22 March 2003).
Success could also change people's attitude to genetic engineering more widely. "Our aim is to reduce the burden of dengue in vulnerable populations," says Alphey, "but if this helps promote a more nuanced discourse about genetic technology in general, that would certainly be a welcome side effect."
Henry Nicholls is a freelance writer based in London and author of The Way of the Panda (Profile, 2010)