Dengue Fever and Zika Virus Make Humans More Attractive to Mosquitoes

Scientists conducted human and mouse studies to find the skin bacteria responsible for the draw

Mosquito on Human Skin
The yellow fever mosquito, Aedes aegypti, spreads dengue fever, Zika, chikungunya and other viruses that infect humans. Smith Collection / Gado/Getty Images

Diseases transmitted by mosquitoes cause hundreds of thousands of deaths annually. With malaria causing more than 600,000 fatalities a year, it is the most famous. But around 4 billion individuals live in regions with a high risk of dengue fever infections, which cause about 40,000 deaths every year. And individuals in at least 86 countries have been infected with the Zika virus. Cases are rarely fatal, but they have been associated with serious birth defects. Scientists studying such diseases are now investigating whether viruses could tweak human physiology to their benefit, and if so, how they do it.

That quest led Gong Cheng, a microbiologist at the Tsinghua University-Peking University Joint Center for Life Sciences, to test whether humans infected with dengue fever and Zika virus are more attractive to mosquitos. Their new study published today in Cell reveals that mosquitoes become more attracted to hosts that are infected with both flaviviruses, diseases in the same family as West Nile and yellow fever. Their results show that a chemical produced by bacteria in the skin is responsible for this increased allure to the insects. Cheng writes in an email that his team’s findings could “inform real-world public health strategies for controlling mosquito-borne flaviviral viral diseases such as dengue and Zika.”

In the first part of their multi-step study, Cheng’s team tested whether two species of mosquitoes, Aedes aegypti and Aedes albopictus, were more attracted to Zika or dengue infected mice than to uninfected mice. They put mosquitoes in a plastic box connected with tubes to two other boxes, one on each side. These side chambers had air piped in from nearby containers of mice, which were either uninfected or infected with Zika or dengue.

The scientists released 60 mosquitoes into the central chamber and watched their movements carefully over the course of a week. At first, similar numbers of mosquitoes entered each of the side boxes. But by day four, the scientists had noticed a clear pattern: around 70 percent of the mosquitoes entered the box connected to the infected mice, while only 30 percent entered the box connected to the uninfected mice. But when they repeated the experiment after adding a deodorization device that blocked the smelly chemicals from entering the boxes, the mosquitoes no longer showed any preference. The scientists concluded that virus infection changes a mouse’s odor, making it more attractive to mosquitoes.

To test whether humans become more attractive to mosquitoes when they get infected, Cheng’s team recruited both dengue patients and uninfected participants. The researchers swabbed their armpits to collect body odor chemicals and then had them hold a piece of paper with odor compounds in one hand and untreated paper in the other hand. Using similar methods as before, mosquitoes were allowed to choose between the two hands. As with the mice, the mosquitoes showed the strongest attraction to odors from humans that were infected with dengue.

To determine what specific chemical compound changes with flavivirus infection, the scientists isolated chemicals that were released into the air by infected and uninfected mice. Twenty chemical compounds differed between the mice infected with Zika or dengue compared to the uninfected mice. The researchers then tested whether each of these compounds could trigger a nerve impulse from the mosquitoes’ antennae to their brains, indicating that mosquitoes can sense the chemical.

A compound called acetophenone became the prime suspect since it caused the strongest electrical response in the mosquito antenna test. Acetophenone was emitted into the air more by mice that were infected with Zika or dengue compared to uninfected mice. Human dengue patients also showed higher levels of acetophenone emissions than uninfected participants.

Acetophenone was also the only chemical that mosquitoes showed a clear attraction to when it was applied to a mouse’s skin. When acetophenone was applied to one human hand and both hands were placed in chambers connected to a mosquito cage, the mosquitoes flocked towards the chemical laden hand.

While previous studies had found evidence of increased mosquito attraction in humans infected with mosquito-borne diseases, this study is one of the first to clearly identify a specific chemical as the reason for this attraction. “That they found this acetophenone, and that they found this effect so clearly, both in mice and in humans, I think that's really the highlight of the study,” says Niels Verhulst, a vector entomologist at the University of Zürich’s Institute of Parasitology who was not involved in this study.

The researchers still wondered why acetophenone emissions increased after viral infection. Acetophenone is a common bacterial waste product, so the researchers figured it might be produced by bacteria living on skin. So they killed off the bacteria from the skin of some mice, and found that the mosquitoes chose to fly towards the uninfected mice just as often as the infected ones. Mice that lacked skin bacteria produced hardly any acetophenone.

The scientists dug in more. They found that infected mice had more Bacillus bacteria than uninfected mice. When Bacillus bacteria were added to a mouse’s skin, it became more attractive to mosquitoes. Bacillus bacteria also produced the highest acetophenone emissions among all the skin bacteria that were tested.

Taken together, the study’s findings suggest that dengue or Zika infection causes an increase in Bacillus bacteria on the skin, leading to higher acetophenone emissions, and greater mosquito attraction. According to Julien Martinez, an evolutionary biologist at the MRC-University of Glasgow Centre for Virus Research who was not involved in this study, having each step of this pathway so clearly understood is quite rare. “We know that parasites often evolve ways to manipulate the host to increase their transmission. And we know many examples of this in the animal kingdom,” he says, “but most of the time, we have no clue about the mechanisms.”

While figuring out this mosquito-attraction mechanism is an important advancement, Verhulst expects that it is not the only one at play. Based on previous research, he suspects that bacteria other than Bacillus might also be involved in attracting mosquitoes. “If you think of skin bacteria that attract mosquitoes, there are like four or five studies that have shown that Staphylococcus play a role,” he says. Cheng similarly notes that compounds besides acetophenone may also be involved in the increased mosquito attractiveness of infected hosts.

Cheng’s research team is currently considering ways to apply their research findings to reduce the spread of dengue and related viruses. They are looking into treating dengue patients with drugs that reduce their acetophenone emissions. A common acne medicine called isotretinoin is known to increase production of an antimicrobial protein that is especially effective at killing Bacillus bacteria. When infected mice were given isotretinoin, their Bacillus load decreased and their mosquito-attractiveness dropped. But considering that the majority of this research was done on mice, more studies are needed on humans. Cheng also notes that isotretinoin has potentially severe neurological side-effects, so his team is investigating safer alternative treatments.

Other potential applications of these findings could include designing more effective mosquito traps and faster tests for dengue or Zika infection. Health professionals may even be able to determine infection status by measuring a person’s acetophenone emissions from their skin. “This can be quick, much quicker than taking a blood sample and doing a test,” Verhulst says.

The research team is also planning to take a closer look at the mosquitoes themselves. Cheng is now hoping to find the genes that enable mosquitoes to sense acetophenone. If they can turn these off, they may make the vectors less attracted to infected humans and less likely to spread the virus.

Of course, Martinez also points out that accurately predicting how mosquitoes behave could depend on how the virus affects them. “If you want to understand the epidemiology of the virus,” he says, “you need to look at all the different steps of its lifecycle. So when it's in the human, and when it's in the mosquito.”

Get the latest Science stories in your inbox.