Leslie Vosshall

Researchers work towards throwing mosquitoes off the scent.

Mosquitoes track their victims by following carbon dioxide trails from their prey's breath. Like most insects, mosquitoes detect this gas with great sensitivity — something humans cannot do. To Leslie Vosshall at the Rockefeller University in New York this capability “was a fascinating unsolved problem”.

Scientists had long been aware that most insects can detect CO₂, but they knew little about the molecules involved in this process. Identification of mosquitoes' CO₂ receptors could help scientists develop compounds to throw the insects off their human trail, thereby preventing the spread of mosquito-borne diseases such as malaria and West Nile virus.

Vosshall became interested in insect CO₂ detection in 2004 when she was finishing her postdoc on the fruitfly Drosophila in Richard Axel's laboratory at Columbia University, New York. When this insect nears something that gives off CO₂, it quickly turns away. Work by Axel's group and John Carlson at Yale University in New Haven, Connecticut, determined that one neuronal type in Drosphila antennae is responsible for this avoidance behaviour. In addition, Axel's group found that the protein Gr21a is expressed specifically on the CO₂-detecting neurons.

“These were the building blocks for my work,” recalls Vosshall. Once she had established her own laboratory, Vosshall and postdoc Walton Jones developed possible hypotheses for finding the elusive CO₂ receptors in Drosophila, and their team started considering approaches to test it. They knew that in Drosophila two distinct proteins always team up to detect a specific odour. From this they proposed that Drosophila would use a pair of receptors to detect CO₂. The only clue they had to these receptors' identity was the protein Gr21a, which is expressed on CO₂-detecting neurons. So Jones searched for another Drosophila protein that would co-localize with Gr21a, which led him to Gr63a.

The technology for making targeted mutants in not as mature for flies as for mice. We were lucky that we got one mutant after one year.

The next hypothesis was that the genes encoding Gr21a and Gr63a would be conserved in other insects, demonstrating that they perform a critical function. A screen for homologues in the mosquito Anopheles gambiae, one of the species that carry malaria, found two matches (see page 86).

The next step was the most challenging. To show that Gr21a and Gr63a are necessary for detecting CO₂, Vosshall and Jones needed to knock one or both genes out and show that the flies could not detect the gas. “The technology for making targeted mutants in not as mature for flies as for mice. There is an element of luck involved,” says Vosshall. “We were lucky that we got one mutant after one year.” Consistent with their third hypothesis, their mutant fly that lacked Gr63a did not avoid CO₂.

“The study was done very sequentially,” says Vosshall. “It was one of the simplest studies done in my lab.” But the work was not without surprises. The team found that both Gr21a and Gr63a are membrane proteins. In vertebrates, gas receptors typically reside inside cells. “It is possible that in vertebrates the receptors for carbon dioxide are also membrane proteins,” says Vosshall. But for the moment she will be turning her attention not to vertebrates but back to mosquitoes.

Vosshall and her colleagues next hope to show that the homologous receptors identified in mosquitoes also function as CO₂ detectors. To do this, they can either carry out genetic manipulations directly in mosquitoes, which is tricky to do, or use Drosophila or frog oocytes as 'synthetic mosquitoes.' They will then be able to start testing different compounds to block the receptors and keep mosquitoes away from their food source.