ABOVE: Mosquitoes demonstrating a preference for the smell of infected mice (right) Courtesy of Gong Cheng et al.

Some of the most notorious human viruses, including dengue and Zika, belong to a group of viruses called flaviviruses. They need mosquitoes to ferry them from host to host, and research now suggests they play an active role in ensuring that transfer occurs. The viruses are able to manipulate their hosts’ skin microbes so that they produce an increased amount of a chemical that attracts mosquitoes to the host, researchers from Tsinghua University in China report today (June 30) in Cell.

Lead author and Tsinghua University microbiologist Gong Cheng tells The Scientist in a written statement that he and his colleagues wanted to understand how mosquito-vectored viruses spread given that infected hosts are often vastly outnumbered by noninfected ones. He notes that “mosquitoes need to actively seek and feed on a viremic host to acquire infectious viral particles; however, the absolute number of infected individuals are very low,” oftentimes only a single person in a thousand. The team was therefore interested in finding out how “mosquitoes effectively orient to viremic hosts with a high frequency.”

The researchers had a hunch that the insects might be attracted to scents emitted by infected hosts based on previous studies on other insects. So, for a period of six days, they placed 60 hungry Aedes aegypti mosquitoes in a chamber system where they could smell either mice infected with dengue and Zika viruses or healthy mice and move towards their preferred scent. By the last day, 70 percent of the mosquitoes opted to be in the chamber containing the infected mice’s odor. The team then repeated the experiment on two other strains of A. aegypti—and again, the mosquitoes moved towards the smell of infected mice. 

See “Malarial Mosquitoes Drawn to Human Smell

Helen Lazear, a virologist at the University of North Carolina at Chapel Hill who was not involved in the study, says that they made an “interesting observation” about factors that might attract mosquitoes to infected individuals, but that she would have liked to see how the mosquitoes react to a greater diversity of pathogens, especially ones not vectored by mosquitoes. “There should have been a comparison with other infections in the mice to show that this effect is specific to viruses like dengue [and] Zika that are transmitted by these mosquitoes,” says Lazear. “It would be interesting to know if you cause that immune response with some other virus, would you see the same effect, or is this something that is really specific to these viruses?”

To ferret out what might be attracting the mosquitoes to the infected mice, the team analyzed the volatile compounds emitted by bodily fluids from infected and healthy mice and pinpointed the ones common to the infected mice. Then, the researchers applied each of the compounds to the mosquitoes’ antennae and monitored their olfactory response with an electroantennogram. Of the 20 chemicals tested, acetophenone—a sweet-smelling chemical produced by bacteria—emerged as the most potent stimulator. And in chamber tests, it was the only one that attracted more mosquitoes than a control scent—including when the chemical was topically applied to human hands. “We were very excited to identify acetophenone from host skin microbiota as the targeted volatile compound to manipulate the feeding motivation of mosquitoes,” Cheng writes. 

two hands in chambers with mosquitoes on the mesh at the end
Hands laced with acetophenone (right) attract more mosquitoes than those coated in a control solvent (left).
Courtesy of Gong Cheng et al.


Previous literature had documented that a derivative of acetophenone secreted by Pseudomonas aeruginosa, a human pathogen, attracts several fly species to food contaminated by the bacteria. Vector biologist Matthew Aliota of the University of Minnesota explained that similar ideas have been raised about the human skin microbiome. And while studies have suggested nonviral human pathogens, particularly malaria parasites, alter people’s smell in a way that attracts mosquitoes, this is the first experiment that shows that a virus-infected animal is more attractive to a host-seeking mosquito.

See “Researchers Discover What Attracts Mosquitoes to Humans

To investigate the source of the acetophenone, the researchers rid the experimental mice of internal bacteria with oral antibiotics, but this did not deflect the mosquitoes from the mice. What did lower the insects’ interest was a topical antibiotic on the mice’s skin that similarly wiped away their external microbiome. That strongly pointed to skin-dwelling bacteria as the source of acetophenone. And indeed, when Cheng and his team examined cultures of mouse skin microbial communities, they found that acetophenone-producing bacteria like Bacillus increased following infection. 

The team then traced the overgrowth of acetophenone-producing bacteria to the activity of a gene that expresses RELMα—a protein involved in protecting the mice against pathogenic skin microbes. Dengue and Zika infections reduced the expression of this gene and consequently, RELMα, though this effect was reversed by a dietary supplement of Vitamin A derivatives, hinting at potential therapeutic routes for reducing bite frequency.

“Their study shows convincingly, at least in a mouse model, that there are specific host cues that are altered in animals that are infected with flaviviruses like dengue and Zika that may increase an individual’s attractiveness to a host-seeking mosquito,” says Aliota, adding that their study is a “somewhat provocative one” and “a good starting point to try to understand the underlying basis for these coevolutionary relationships that are occurring between mosquito virus and human host.”