Malaria kills more than half a million people every year, most of them children under age 5. Yet strategies to control the disease are riddled with challenges: Mosquitoes develop resistance to insecticides and vaccines provide incomplete protection. Now, researchers say they’ve found a new weapon: a naturally occurring bacterium that, when fed to mosquitoes, halts the development of the malaria parasite in the insects’ guts.

The findings offer a complement to existing malaria prevention methods, says Carolina Barillas-Mury, who leads the malaria and vector research lab at the U.S. National Institutes of Health. The approach “has great potential to be implemented” in areas where malaria is endemic, adds Barillas-Mury, who was not involved with the work.

Researchers have tried to use microbes to control mosquito-borne diseases before. The virus-fighting bacterium Wolbachia pipientis has shown particular promise against dengue fever in recent clinical trials and is already used in some areas of the world. But most methods for blocking malaria-causing Plasmodium parasites, which are transmitted by different mosquito species from dengue, have relied on genetically modified bacteria. That’s a major obstacle to regulatory and public acceptance, Barillas-Mury notes, given the unknowns of releasing edited organisms into the wild.

The bacterium in the new study, published today in Scienceinhibits the malaria parasite without any genetic tinkering by humans. Janneth Rodrigues, a scientific lead in global health medicines R&D at GlaxoSmithKline, and colleagues stumbled across the microbe at a GSK research center in Spain, after noticing the mosquitoes they were using for malaria research were getting harder to infect with Plasmodium.

It turned out there was an unexpected microbe, a bacterial strain called Delftia tsuruhatensis TC1, inside these mosquitoes’ guts, Rodrigues says. It’s not clear how the bug got there, though the species has been seen before in the guts of some insects, as well as in water, soil, and, rarely, hospital-acquired infections. But when the team fed the strain to other mosquitoes, it seemed to block the parasite from infecting them.

To learn more, Rodrigues’s group teamed up with Marcelo Jacobs-Lorena, a malaria researcher at the Johns Hopkins Bloomberg School of Public Health. His lab discovered  D. tsuruhatensis was disrupting Plasmodium’s growth in the mosquito gut—where the parasite develops before moving to the insect’s salivary glands. Compared with mosquitoes lacking the bacterium, insects with D. tsuruhatensis had about 75% fewer Plasmodium oocysts—egglike structures the parasite forms in the insect gut.

Experiments in rodents revealed that this disrupted Plasmodium growth led to reduced transmission: Only one-third of mice bitten by bacterium-carrying mosquitoes became infected, compared with 100% of mice bitten by regular mosquitoes.

The research identifies D. tsuruhatensis as an attractive tool for malaria prevention, Jacobs-Lorena says. Mosquitoes only need to eat a little of the microbe to become colonized, and once inside the insect, the bacterium stays there, continually blocking parasite development, he explains. What’s more, D. tsuruhatensis doesn’t appear to affect the mosquito’s survival or that of its offspring—meaning the insects are unlikely to evolve resistance to it.

The Johns Hopkins researchers also homed in on how the bacterium hurts Plasmodium: by secreting a molecule called harmane—a compound found in plants used for traditional medicine in some cultures. Feeding this compound on its own to the mosquito or just getting the insects to walk on it hobbled the parasite’s development—though the effect wore off after a day or so. The findings suggest harmane might be used to treat bed nets or other surfaces mosquitoes come into contact with, Rodrigues says.

In a last round of experiments, the scientists teamed up with researchers in Burkina Faso to test the approach on mosquitoes in a roughly 10-meter-by-10-meter-by-5-meter netted enclosure designed to simulate the real world with plants and breeding areas. Leaving out balls of cotton wool soaked in sugar and D. tsuruhatensis was enough to colonize some three-quarters of the mosquito population with the bacterium overnight. When those mosquitoes fed on blood from people with malaria, the bacterium blocked the parasite’s development, just like it had in the lab.

“It’s a really impressive piece of work,” says Grant Hughes, a vector biologist at the Liverpool School of Tropical Medicine who was not involved in the research. He praises the study’s use of multiple experimental approaches to understand the bacterium’s effects on malaria transmission and says the identification of harmane as the Plasmodium-attacking compound is particularly useful for future work.

Barillas-Mury highlights another important finding from the study: Unlike some gut bacteriaD. tsuruhatensis isn’t passed between mosquitoes. Although this could be a limitation down the line, it’s a plus from a safety perspective, and may make the path to regulatory approval easier. “When you are ready to deploy in massive amounts, you would like [the bacterium] to spread,” she says. “But before you do that, you really want one that doesn’t spread too much.” That way, she says, researchers can evaluate its efficacy and safety before it takes over a whole population.

Rodrigues, who along with another GSK researcher is named as co-inventor on a patent related to the discovery, says an eventual product could take the form of a bacterium-containing powder or even harmane itself mixed with sugary baits to attract the insects. For now, though, the researchers are running additional experiments to test the method’s safety—making sure, for example, that that bacterium doesn’t harm other insects such as bees.

Hughes says he hopes the field will be spurred to look for other gut bacteria that might have similar effects. “There are so many different bacteria and microbes that are associated with these mosquitoes,” he says. “The implications [for using] these microbes to control disease are very promising.”