Mario Tama / Kevin Frayer / Getty / James Gathany / AP / Zak Bickel / The Atlantic

How to Beat Dengue and Zika: Add a Microbe to Mosquitoes

After 30 years of development, virus-beating insects are finally being deployed in megacities around the world.  

On January 4th, 2011, in the first hours of a crisp Australian morning, Scott O’Neill walks up to a yellow bungalow in a suburb of Cairns. He sports glasses, a goatee, jeans, and an off-white shirt, with “Eliminate Dengue” written over the breast pocket. That’s both the name of the organization that O’Neill founded and its goal: eliminate dengue fever, from Cairns, from Australia, and perhaps eventually from the world. The tools with which he will accomplish this feat are sitting in the small plastic cup that he now holds in his hand.

He carries it towards the house, past a fence, down a flower-lined patio, and up to a large palm tree. His pace is deliberate and a little self-conscious. This is a big moment. Around 20 people are watching, filming, joking. O’Neill stops and looks up. “Are you ready?” he says. The crowd cheers. They have been waiting for this for a long time. O’Neill pulls the lid off the cup, and a few dozen mosquitoes fly out into the morning air. “Go, babies, go!” says an onlooker.

This article is adapted from Yong’s forthcoming book.

These mosquitoes are Aedes aegypti, a black-and-white species that transmits the virus that causes dengue fever. Through its bites, it infects as many as 400 million people every year. O’Neill has never had dengue himself, but he has seen others go through it. He knows about the fevers, headaches, rashes, and the severe joint and muscle pains. He knows that there is no vaccine or effec­tive treatment.

The only real way of controlling dengue is through prevention. We can kill Aedes mosquitoes with insecticides. We can stop them from biting, using repellents or nets. We can remove the open, stagnant water in which the insects breed. But despite these strategies, dengue fever is still common, and increasingly so. O’Neill’s plan, unorthodox though it sounds, is to beat the disease by releasing even more of the Aedes mosquitoes that carry it. But his insects are different from their wild counterparts. O’Neill has loaded them with a marvelous bacterium called Wolbachia.

It is not unusual for science writers who regularly write about microbiology to pick a favorite bacterium, much as people would choose a favorite film or band. Wolbachia is mine. It is breathtaking in its behavior and majestic in its spread.

It was discovered unceremoniously. In 1924, pathologist Marshall Hertig and entomologist Simeon Burt Wolbach were looking inside common brown mosquitoes, Culex pipens, which they had collected near Boston and Minneapolis, when they found a new microbe. It looked a bit like the Rickettsia bacteria that Wolbach had previously identified as the cause of Rocky Mountain spotted fever and typhus. But this new microbe didn’t seem responsible for any disease – and so was largely ignored. It took twelve years for Hertig to formally name it Wolbachia pipientis, in honour of his friend who found it and the mosquito that carried it.

For decades, the bacterium was ignored—it seemed rare and unimportant. Then, in the 1980s and 1990s, when scientists started identifying microbes by sequencing their genes, they started find­ing Wolbachia everywhere.

Richard Stouthamer discovered a group of asexual, all-female wasps, which only repro­duced by cloning themselves. This trait was the work of a bacterium, Wolbachia: when Stouthamer treated the wasps with antibiotics, the males suddenly reappeared and both sexes started mating again. Thierry Rigaud found bacteria in woodlice that transformed males into females by interfering with the production of male hormones; it was Wolbachia, too. In Fiji and Samoa, Greg Hurst found that a bacte­rium was killing the male embryos of the magnificent blue-moon but­terfly, so that the females outnumbered the males by a hundred to one. Again: Wolbachia. Maybe not exactly the same strain, but all were dif­ferent versions of the microbe from Hertig and Wolbach’s mosquito.

There’s a reason why all of these strategies are bad news for males. Wolbachia can only pass to the next generation of hosts in eggs; sperm are too small to contain it. Females are its ticket to the future; males are an evolutionary dead end. So it has evolved many ways of screwing over male hosts to expand its pool of female ones. It kills them, as in Hurst’s butterflies. It feminizes them, as in Rigaud’s woodlice. It elim­inates the need for them entirely by allowing females to reproduce asexually, as in Stouthamer’s wasps. None of these manipulations is unique to Wolbachia, but it is the only bacterium to use them all.

Where Wolbachia does allow males to survive, it still manipulates them. It often changes their sperm so that they cannot success­fully fertilize eggs unless the eggs are infected with the same strain of Wolbachia. From the females’ perspective, this incompatibility means that infected females (which can mate with whomever they like) gain a competitive advantage over uninfected females (which can only mate with uninfected males). With every passing generation, the infected females become more common, as do the Wolbachia they carry. This is called cytoplasmic incompatibility, and it’s Wolbachia’s most common and most successful strategy – the strains that use it spread so quickly through a population that they typically infect 100 percent of their potential hosts.

Aside from these misandrist tricks, Wolbachia also excels at invading ovaries and entering egg cells, so it quickly becomes a hand-me-down that insects pass on to their offspring. For these reasons, it has become exceptionally common. One recent study estimated that it infects at least four in every 10 species of arthropods—the animal group that includes insects, spiders, scorpions, mites, woodlice, and more. That is a preposterous proportion! The majority of the 7.8 million or so liv­ing animal species are arthropods. If Wolbachia infects 40 percent of them, it is arguably the most successful bacterium in the world, at least on land. And, rather tragically, Wolbach never knew. He died in 1954, unaware that his name had been grafted onto one of the greatest pandemics in the history of life.

In many cases, Wolbachia is a reproductive parasite. But it has a beneficent side, too. It provides some unknown benefit to certain nem­atode worms, which cannot survive without it. Bed bugs use it as a dietary supplement, which provides the B-vitamins that are missing from their blood meals. The caterpillar of the spotted tentiform leaf miner—a kind of moth—uses Wolbachia to stop leaves from yellowing, creating green islands in which the very hungry insects can feed. And most importantly for us humans, Wolbachia seems to protect many flies and mosquitoes from viruses and other pathogens.

That’s the simple basis of O’Neill’s dengue-beating plan: Wolbachia stops Aedes mosquitoes from car­rying dengue viruses, turning them from vectors into dead ends. If he can get the bacterium into enough wild mosquitoes, he would cut the chains of dengue transmission and eliminate the disease.

Of course, it would be impossible to collect every wild mosquito and shoot them up with a symbiont, but O’Neill doesn’t have to. Wolbachia is such an exceptional manipulator that once released into a population, it should spread quickly. All O’Neill has to do is to release enough Wolbachia­-carrying mosquitoes into the wild, and wait. The ones he set loose in Cairns were the first. This was the culmination of decades of obsessive hard work and hair-pulling frus­tration. “It seems like my whole life,” says O’Neill.

Scott O’Neill holds a cup of Wolbachia-infected mosquitoes. Credit: Eliminate Dengue

His quest to turn Wolbachia into a dengue fighter began in the 1980s, meandered through several wasted years, and hit many a dead end. It only started bearing fruit in 1997, when he read about an unu­sually virulent strain of Wolbachia that infects fruit flies. This strain, known as ‘popcorn,’ would reproduce like mad in the muscles, eyes, and brains of adults, filling a fly’s neurons so thoroughly that they become “akin to a bag full of popcorn,” hence the name. These infections are so severe that they can halve a fly’s lifespan.

“That was a lightbulb moment for me,” says O’Neill. He knew that dengue viruses take time to reproduce in mosquitoes, and even more time to reach the salivary glands where they can jump into a new host. This means that only old mosquitoes can transmit dengue. If O’Neill could halve the insects’ lives, they would die before they got a chance to spread the virus. All he needed to do was to get popcorn into Aedes.

Wolbachia infects many mosquitoes—remember that it was origi­nally discovered in a Culex before anyone realised how omnipres­ent it is. But as luck would have it, it doesn’t touch either of the two groups that cause the most human suffering: Anopheles, which carries malaria, or Aedes, which spreads Chikungunya, yellow fever, and den­gue. O’Neill was going to have to play matchmaker and create a new symbiosis from scratch. He couldn’t just inject adults with Wolbachia, though; he needed to inject an egg, so that every part of the resulting insect would carry the microbe.

He and his team would look down a microscope and, ever so delicately, try to lightly puncture a mosquito egg with a needle bearing Wolbachia. They did this hundreds of thou­sands of times, over many years. It never worked. “I burnt the careers of all these students and I was so frustrated that I was ready to walk away,” says O’Neill. “But I just had this sadistic streak in me. This par­ticularly bright student came into the lab in 2004, and I couldn’t help myself. I put the old project in front of him and he bit really hard. He was Conor McMeniman. He was one of the best students I ever had. He made it work.”

It took thousands more attempts, but McMeniman finally managed to stably infect an egg in 2006, creating a line of Aedes that naturally carried Wolbachia. Most alliances between animals and microbes are millions of years old. Here’s one that is, at the time of writing, 10 years old.

But after all that work, the team discovered a fatal flaw in their plans: the popcorn strain was too virulent. Besides killing females pre­maturely, it also reduced the number of eggs they laid, and the viabil­ity of those eggs, thus sabotaging its own chances of moving into the next generation of mosquitoes. Theoretical work by Michael Turelli at the University of California, Davis, who first showed that Wolbachia would spread quickly in natural populations, revealed that the popcorn strain would not. That was terrible news.

O’Neill soon learned that none of that mattered. In 2008, two groups of researchers independently discovered that Wolbachia made fruit flies resistant to the group of viruses responsible for dengue, yel­low fever, West Nile fever and other diseases. When O’Neill saw that, he immediately asked his team to feed their Wolbachia-infected mos­quitoes with blood that had been spiked with dengue virus. The virus utterly failed to take hold. Even when the team injected it straight into the insects’ guts, Wolbachia stopped it from replicating.

That changed everything. The team didn’t need Wolbachia to shorten a mosquito’s lifespan. Its mere presence would be enough to prevent the spread of dengue! Better still, the team didn’t need popcorn any more. Other less virulent strains were similarly protective, and would spread far more easily. “After years and years of banging our head against the wall, we suddenly realized that we didn’t need to,” says O’Neill.

The team switched to a different strain called wMel, which had a track record of spreading through wild insect populations, but was an altogether gentler companion than popcorn, with none of the same life-shortening, brain-destroying, egg-killing effects. But would it spread?

An Eliminate Dengue volunteer leaves a bucket of Wolbachia-carrying mosquitoes in an Australian backyard.

To find out, Ary Hoffmann and Scott Ritchie built two insect aviaries: giant, walk-in cages, which they filled with mosquitoes. For every one uninfected insect, they added two wMel carriers. They also included a makeshift porch for the mosquitoes to hide under and a pile of sweaty gym towels to attract them. And for fifteen minutes a day, they added some succu­lent team members to feed the Wolbachia-infected mosquitoes. Every few days, the team collected eggs from the cages and checked them for Wolbachia. They found that, within three months, every mosquito larva inside was infected with wMel. Everything suggested that their big idea would work. All the signs were saying: Go.

So, they did. Since 2006, well before the team had a mosquito with Wolbachia, they had been talking to the residents of two Cairns sub­urbs—Yorkeys Knob and Gordonvale—about their plans. Hi, they said, we have a plan to get rid of dengue fever. Yes, we know that you’ve always been told to kill mosquitoes because they make you sick, but now, we’d appreciate it if you let us release more mosquitoes. No, they’re not genetically modified, but we have loaded them up with a microbe with a penchant for spreading rapidly. Also, Aedes mosqui­toes don’t migrate very far, so for this plan to work, we’re going to have to do lots of releases, including on your property. Yes, they’ll probably bite you. No, no one has ever done this before. Are you in?

Amazingly enough, they were. For two years, the Eliminate Dengue team ran focus groups, talks in town halls and local pubs, and a shopfront drop-in clinic where people could ask questions. They knocked on a lot of doors.

O’Neill knew what could happen if scientists ignored local communities. In 1969, World Health Organization scientists travelled to India to try a variety of new techniques for controlling mosquitoes, including genetic modification, irradiation, and Wolbachia to try and control mosquito populations. The project was secretive and people grew suspicious. Newspapers started accus­ing the scientists, some of whom were American, of using India as a test-bed for experiments that were too dangerous for US soil, and even developing biological weapons. The team responded by not responding at all. “It was a PR nightmare,” says O’Neill. “They were thrown out of the country, and the controversy made the genetic modification of mosquitoes taboo for 20 years.” He wanted to avoid making the same mistake.

“The project requires a lot of trust, and we got it, but it didn’t happen overnight,” he says. “We were very authentic in how we listened to people. When they had concerns, we addressed them. We even did experiments.” For example, they showed that Wolbachia couldn’t infect fish, spiders and other predators that bit the mosquitoes, or humans whom the mosquitoes bit. Slowly, even skeptics became supporters. “This local volunteer group, who mobilize people to help the community if floods and cyclones happen, asked if they could go door-to-door on our behalf to get people to release mos­quitoes from their houses,” says O’Neill. “That was a real turning point for me.” By 2011, when the mosquitoes were ready, the project had the support of 87 percent of the residents.

It began in earnest on that January morning, with the cup that O’Neill ceremoniously opened. “We were all a bit giddy,” O’Neill recalls. “We had been working on this thing for frigging decades. A whole bunch of us were there for that moment, people who had been on the journey for a long time.” the team marched through the streets, pausing at every fourth house to release a few dozen mosqui­toes. Within two months, they had liberated some 300,000 of them, pausing only to duck an incoming cyclone.

Every two weeks, the team would then collect mosquitoes from the suburbs using a grid of traps, and test the insects for Wolbachia. “It actually worked better than expected,” says O’Neill. By May, Wolbachia was sitting happily in 80 percent of the Gordonvale mosquitoes and 90 percent of those in Yorkeys Knob. In just four months, the dengue-proof insects had almost totally replaced the native ones. For the first time in history, scientists had transformed a population of wild insects to stop them from spreading human diseases. And they did it through symbiosis.

But O’Neill’s organisation isn’t called “Transform Mosquitoes.” It’s called “Eliminate Dengue.” Have they done that? There certainly haven’t been any new cases in the two suburbs since 2011—an encouraging sign, if not a definitive one. Neither area was a dengue hotspot to begin with. Nor is Australia, for that matter. O’Neill will be able to declare victory only when his mosquitoes repress dengue in the coun­tries where it’s most prevalent, which is why he is now expanding his work to Brazil, Colombia, Indonesia, and Vietnam. When he started Eliminate Dengue in 2004, it was just him and his lab members. Now, it’s an international team of scientists and health workers. “In two to three years, we should have good evidence showing its impact,” says O’Neill. “In 10 to 15 years, we should be able to make a significant dent in dengue.”

The team have carried out several small releases in all of their target countries, and shown that the Wolbachia-carrying mosquitoes can establish themselves in a variety of climates. Each new place presents its own challenges. For example, if a city is gratuitous in its use of insecticides, the resident mosquitoes will probably be partly resistant. Releasing naïve Australian-born mosquitoes into such an environment would be pointless: they would succumb to poison long before they passed on their symbionts. So, the Wolbachia-infused mosquitoes need to be at least as resistant as the local ones. Cross-breeding can help.

At the Indonesian chapter of Eliminate Dengue, scientists breed the Wolbachia-carriers with local mosquitoes for several generations, so that the insects they release are as close to the indigenous ones as possible. That should help them to mate more successfully, too. “Every location is unique,” says O’Neill, “but we’re seeing that Wolbachia works well in every setting. Everything suggests that it should be possible to roll it out globally.” It’s also encouraging that wherever the team have deployed their mosquitoes, dengue transmission has plummeted. “Dengue’s a focal disease, so when people get sick, so do their neighbours,” says O’Neill. “And we’re not seeing any of that clustering.”

They are now gearing up to go big. Back in Australia, they have started to disperse their mosqui­toes through the northern city of Townsville. With some 200,000 res­idents to address, the team can’t go knocking on every door. Instead, they rely on media coverage, big public events, and citizen science initiatives, where local people – even schoolchildren – volunteer their time. It’s also too cumbersome to release adult mosquitoes. Instead, the team hands containers with eggs, water, and food to homeown­ers, who let the mosquitoes grow up in their gardens. “Ultimately, we want to go to tropical megacities,” says O’Neill.

He means places like Yogyakarta in Indonesia, a metropolis that’s home to some 4 million people. The Eliminate Dengue team have already shown that Wolbachia could spread throughout the mosquitoes of two small nearby towns. Now, they’re gearing up for a bigger trial in the city itself. Their plan is to release their Wolbachia-carrying mosquitoes over some randomly selected areas, and see if dengue transmission falls compared to nearby regions.

This approach isn’t limited to dengue. Wolbachia also stops mosquitoes from carrying the viruses behind Chikungunya and yellow fever, and the Plasmodium parasites that cause malaria; a team of Chinese and American scientists has now successfully melded the microbe with the Anopheles mosquito that spreads malaria. Other research­ers are trying to use Wolbachia to control insect pests like tsetse flies, which spread sleeping sickness, and bed bugs, which spread sleepless nights.

And in recent months, Eliminate Dengue researchers in Brazil and Colombia have shown that Wolbachia could help to curtail the spread of Zika—the now-infamous viral disease that has spread explosively throughout the Americas and causes microcephaly, a birth defect that stunts brain development. When carrying Wolbachia, mosquitoes do not seem to harbor the Zika virus in their saliva, and so are very unlikely to transmit it. For that reason, the World Health Organization recently recommended testing Wolbachia-carrying mosquitoes as a way of curtailing the ongoing Zika crisis.

O’Neill’s team have answered the call. Over the next few years, they will release their insects in three to-be-confirmed locations in Brazil, Colombia, and the Asia-Pacific. Two of those will be home to around 2.5 million people. “We’re talking close to entire cities,” says O’Neill.


Skeptics would argue that evolution produces a countermeasure to every measure, a parry to every thrust. Dengue viruses should even­tually become resistant to the encroaching wave of Wolbachia, and start infecting mosquitoes again. (As British scientist Leslie Orgel once famously said: “Evolution is cleverer than you are.”)

But Elizabeth McGraw, a long-standing member of the Eliminate Dengue team, is optimistic. Her team has shown that Wolbachia protects against viral infections in several ways. It boosts the mosquito’s immune system. It also competes for nutrients like fatty acids and cholesterol, which dengue virus needs in order to reproduce. “The more mechanisms you have, the less likely you’ll get resistance,” she says. “For an evolu­tionary biologist, that’s really heartening.”

O’Neill and McGraw also argue that the spectre of resist­ance haunts every possible control measure, such as insecticides and vaccines. Unlike these other solutions, Wolbachia is alive, and could counter-adapt to any viral adaptations. It is also safe and cost-effective. While insecticides are toxic and must be continuously resprayed, Wolbachia-carrying mosquitoes have no side effects and can sustain themselves when released. “Once it’s going, it’s ongoing,” says O’Neill. “We’re trying to bring the cost in to two to three dollars per person.”

Eliminate Dengue is not the only organisation planning to release Wolbachia-infected mosquitoes. A Kentucky-based startup called MosquitoMate has carried out a few small trials in several US states, while researchers from Michigan State University are doing the same in two Chinese islands. Both groups have shackled Wolbachia to a different mosquito species, the Asian tiger mosquito Aedes albopictus, although MosquitoMate is also planning trials with the dengue-spreading Aedes aegypti. And both groups are profit-making ventures, while Eliminate Dengue is not. “We have a lousy business model,” says O’Neill. “Once our mosquitoes are out there, there’s nothing for us to sell.”

O’Neill marvels at how far the study of Wolbachia has come. “We were a fairly innocent lab that studied symbiosis,” he says. “It was an area of basic science, but something wonderful and applied will come out of it.” “This is just part of the whole new way of thinking, about the microbial ecology of organisms and about how that relates to disease.”