Bacteria vs virus: the battle to defeat dengue

Dengue remains the world’s fastest-spreading mosquito-borne disease. With climate change and urbanisation bringing more people into contact with the Aedes aegypti mosquitoes, the need for a solution has never been greater. Now, the team behind a novel approach to fighting dengue fever – first tested in Australia and demonstrated at scale in Indonesia – is gearing up to launch its most ambitious phase yet. Clare Watson reports.

In an unassuming red-brick building on the southside of Colombia’s second-largest city, Medellín, millions of mosquitoes are bred each week to combat dengue fever. One’s natural instinct is to recoil at the metallic smell of blood wafting out of a humid room where netted cages hold thousands of mosquitoes. But these blood-sucking insects are part of a daring strategy to rid cities like Medellín of dengue fever, one suburb at a time.

From the backs of motorbikes and out car windows, adult mosquitoes are spread across the city and neighbouring urban areas of the Aburrá Valley that is home to three million people. These Aedes aegypti mosquitoes have been bred to carry Wolbachia, a species of bacteria naturally found in up to 60% of insect species, but not usually A. aegypti. When it is, Wolbachia stops A. aegypti from transmitting viruses to humans.

Only a few years ago, dengue fever was rampant in Medellín. But since the non-profit World Mosquito Program (WMP) started releasing its modified mosquitoes in 2017, dengue case numbers have plummeted to their lowest levels in 20 years. It’s the largest continuous release of Wolbachia-infected mosquitoes anywhere in the world, and dengue incidence is down by 94%.

According to Cameron Simmons, an infectious disease scientist at Monash University and WMP’s director of global implementation, the challenge now is “to industrialise the supply of Wolbachia” to reach more cities. Now, the program’s planned expansion is moving into its most ambitious phase yet.

Backyard beginnings

It’s a technique that was first trialled in Townsville almost a decade ago. An affable entomologist by the name of Scott O’Neill had been researching how the Wolbachia bacterium, taken from flies, shut off A. aegypti’s ability to transmit dengue and other viruses. Scientists still aren’t exactly sure of the mechanism, but they think Wolbachia piques the mosquito’s immune system, which stops the virus replicating, and also outcompetes the virus for critical resources.

Being a bacterium, it spreads through mosquito populations of its own accord. Another early experiment to simulate a real-world release with caged mosquitoes, by O’Neill and his team at The University of Queensland, indicated how with a few weekly top-ups of freshly infected mosquitoes, Wolbachia could sweep through contained populations within a few generations.

“I just thought this was another enthusiastic scientist, with a solution for dengue that was going to crash on the rocks of disappointment.”

But then came the same stumbling block faced by other mosquito-control methods: how to roll out the technology citywide. To overcome this, O’Neill had to trial releasing modified mosquitoes in a community willing to give it a go. Townsville in northern Queensland might seem an unlikely choice, with the disastrous consequences of introducing cane toads still fresh in people’s minds. But the town was game; as scientists explained, Wolbachia is a Trojan horse of sorts, with little effect on A. aegypti mosquitoes other than curtailing their short lives and quashing their ability to spread viruses. 

Over 28 months, starting in late 2014, some four million mosquitoes were released from plastic tubs and takeaway food containers in backyards across Townsville. Residents became “Wolbachia Warriors” and primary school kids joined in too. Sampling showed Wolbachia quickly spread to local A. aegypti populations as the lab-bred mosquitoes mated with wild ones.

This is the main advantage of using Wolbachia: unlike repeatedly spraying insecticides to control mosquitoes, the bacterium and the mosquitoes it infects do much of the hard work. The bacterium is passed from a female mosquito to her offspring so none of them become viral vectors, and if an infected male mates with a female mosquito without Wolbachia, her eggs won’t hatch. Once more than about 30% of mosquitoes in an area are infected , Wolbachia spreads through the population in a self-sustaining way, no more releases required (though close monitoring continues).

Within months of releasing the first batches of mosquitoes in Townsville, suburbs had upwards of 80% of mosquitoes carrying Wolbachia. Once the whole city was covered, 66 sq. km in total, locally acquired dengue fever was effectively stamped out and imported cases were trending in the same direction. However, the study had no control areas without mosquito releases with which to compare. Still, O’Neill, now at Monash University in Melbourne, and his colleagues were optimistic they could gather more data in time from places with a far greater dengue disease burden than Australia – data that they hoped would convince others their approach really worked.

Netting new evidence

Fast forward 10 years and the World Mosquito Program (WMP) has now deployed its technology in select cities in a dozen countries. In each one, initial hesitation about whether or not the approach would work in local communities has been overcome with successful pilot projects. “Success is long journey,” says Simmons – and, he adds, it often hinges on feasibility.

The next leap is to establish a mass-breeding facility in Brazil, expected to be operational in 2024. WMP has already released mosquitoes in five Brazilian cities, but the new facility will be capable of producing up to 100 million mosquito eggs per week – enough to cloak many more of the country’s sprawling urban areas in Wolbachia’s protection. Up to 70 million people in Brazil could be protected from dengue in the next 10 years, the program estimates.

Dengue fever remains the world’s fastest­spreading mosquito-borne disease, and its grip on Brazil is growing particularly fast. Climate change coupled with rapid urbanisation is bringing millions more people into contact with A. aegypti. Cities across Asia have also recorded huge surges in dengue cases this past year. Singapore, for example, had its second-worst year on record in 2022, according to its National Environment Agency; more than 32,000 dengue cases were recorded, representing a six-fold jump on 2021.

Simmons has seen first-hand the havoc dengue fever wreaks on hospital systems when working on ward rounds in the early 2000s in Ho Chi Minh City, Vietnam. “When there are raging dengue outbreaks, hospitals suffer and patients line the hallways,” he recalls. He also witnessed trial after trial end in disappointment; none of the drugs or antivirals ever tested for dengue fever delivered any real benefit.

So, when Simmons met WMP founder Scott O’Neill in 2010, whose idea it was to breed A. aegypti with Wolbachia, Simmons wasn’t optimistic. “I just thought this was another enthusiastic scientist, this time from academia, with a solution for dengue that – like most other things – was going to crash on the rocks of disappointment,” he says.

Simmons has since changed his tune, as have other epidemiologists, upon seeing the staggering reductions in dengue case numbers in areas where Wolbachia-infected mosquitoes have been released. After Townsville, studies in Indonesia, Vietnam, Colombia and Brazil have reported falling dengue cases numbers wherever Wolbachia mosquitoes have flown.

Most notably, a randomised, controlled trial – representing the gold-standard of clinical evidence – was conducted in Yogyakarta, Indonesia. Conceiving the trial was easy, says Simmons, but designing it took several years and three more to run. A large chunk of the city was divvied up into areas with and without Wolbachia, so the team could directly compare dengue rates. The results, published in the New England Journal of Medicine in 2021, showed that releasing Wolbachia-infected mosquitoes reduced dengue incidence by 77% over 27 months compared to untreated neighbourhoods. Dengue-related hospitalisations also dropped by 86%.

“We won’t go into a community unless they show support for what we do. We don’t wish to push the technology onto people.”

The government of Yogyakarta province has since elected to roll out Wolbachia-carrying mosquitoes across the whole city and many of the urban areas surrounding it, Simmons says. “There’s now nearly two million people in Yogyakarta province living with the benefit of Wolbachia in their local mosquito population.”

That’s just the beginning of WMP’s plans in the Asia-Pacific region. The program is expanding into Bali, planning to release its first batch of modified mosquitoes on the island later this year, says Simmons. Pilot projects are also underway in Laos and Sri Lanka, while New Caledonia has decided to expand its pilot after early successes. But Brazil poses the biggest test yet, the first time WMP is attempting to disperse Wolbachia-infected mosquitoes nationwide.

While every country and city is different, Simmons says there are commonalities in how these societies work at the neighbourhood level. The program enlists the help of local health agents who do everything from vaccination to maternity care. Their involvement is “hugely important”, Simmons says, in helping reach affected communities, especially those wracked by conflict. Drones are also being tested to disperse adult mosquitoes. Pellets of dried eggs to be plopped in water are likewise sent out in at-home kits.

But it can take months of consultation to ensure communities are on board with releasing Wolbachia mosquitoes in their local area. Part of the challenge is explaining to people who have been warned for decades of the grave danger mosquitoes pose that these insects are now their ally in the fight against dengue fever. Building trust is also paramount.

“Community engagement is a very central element of what we do,” O’Neill says. “We won’t go into a community unless the community shows support for what we do. We don’t wish to push the technology onto people.”

However, dengue-riddled communities also know how little progress has been made by spraying insecticides. “Communities are quick to accept any intervention that can diminish this problem,” says Maria Patricia Arbelaez, an epidemiologist at Colombia’s University of Antioquia tracking the program’s progress in Medellín.

Tomorrow the world

Not every city has seen such a steep drop in dengue case numbers, though. According to the WMP, studies in Rio de Janeiro, Brazil and Vinh Luong, Vietnam, have recorded smaller though still substantial reductions in dengue incidence, of 44% and 48%, respectively. The difference, Simmons says, is patchy coverage in Brazilian favelas where security concerns disrupt the best-laid plans, or only a small umbrella of protection in a bustling Vietnamese city where people come and go.

Dengue outbreaks also differ from place to place: every rainy season in Asia brings a spike in dengue cases, whereas the disease moves through Latin America in waves that can be half a decade apart. This can muddy epidemiological findings that become clearer over time. “You need to be very patient,” says Simmons.

Another randomised, controlled trial led by a different US team is already underway in Belo Horizonte, Brazil. And in a matter of months, WMP is expecting to hear the outcome of more than two years of consultation with the World Health Organisation, which has been considering, based on available evidence, whether to endorse the method as a dengue control tool.

WMP founder O’Neill says the results from Medellín and Yogyakarta demonstrate that the method can work at scale, in big cities, and protect millions of people from dengue fever. Projections also estimate that the healthcare savings from keeping people out of hospital would offset the cost of deploying Wolbachia mosquitoes within five to 10 years. “It’s very rare to have a cost-saving intervention,” says O’Neill.

“Then there’s the real-world evidence,” says Simmons. “When you look at the city of Yogyakarta, their dengue case numbers are lower than they have been for 30 years and north Queensland is essentially dengue-free. So if you ask people in public health in Yogyakarta or north Queensland ‘does Wolbachia work?’, the answer is very clear.”

Gathering data about the technology’s effectiveness against other diseases transmitted by A. aegypti might prove trickier – though that is the hope. Dengue is more predictable than Zika virus, yellow fever and chikungunya, which crop up sporadically. Trials need luck on their side to coincide with a disease outbreak or else it might seem like the technology does little to reduce disease burden. Encouragingly, a 2021 study found that chikungunya cases were halved in Niterói, Brazil, three years after releasing Wolbachia mosquitoes that also slashed dengue cases by 69%; however, Zika incidence only dropped by a third.

Another potential spanner in the works is that viruses might one day become resistant to Wolbachia, evolving an escape route to bypass the bacterium, as it does insecticides. Simmons says this is unlikely because of the many ways Wolbachia outcompetes a virus like dengue: hoovering up critical resources such as cholesterol, blocking its entry into cells and suppressing it by boosting the mosquito’s immune system .

“We’re not so naïve to think the virus will never escape,” says Simmons. WMP’s teams are monitoring for viral resistance to Wolbachia in mosquitoes scooped up from around Townsville, a decade after the technology was first deployed, and elsewhere. “But we hypothesise that it’s going to take a very, very long time before that happens, so our expectation is that we’re going to buy decades of protection.”