Because the gene is found in a wide variety of insect species, the results point to a new strategy for developing insect repellents. Repellents that block the gene, and thus prevent disease-carrying insects from finding human hosts, might eventually help fight malaria and other infectious diseases.
"This finding has a direct applied potential," says assistant professor Leslie B. Vosshall, Ph.D., head of the Laboratory of Neurogenetics and Behavior. "Insects are the primary vectors for malaria, dengue fever, yellow fever, and West Nile encephalitis, and they locate human hosts largely through their exquisitely sensitive olfactory systems."
"We need better insect repellents to use as weapons against the spread of infectious disease," Vosshall adds. "Most insect repellents are based on trial and error, or folk remedies. Now we have a scientific, rational basis for designing insect repellents."
An anomalous smell receptor
The sense of smell is very direct. In order for humans or fruit flies to smell bananas, for example, molecules from the fruit must waft through the air to specialized nerve cells that detect them. In humans these cells are at the top of the nasal passages. In the fruit fly they are located on the antennae and the maxillary palp, an appendage near the fly mouth. The odorant molecules bind to receptors on the neurons, fitting like chemical keys into the lock-like receptors, and set off a series of signals that the brain perceives as smell.
A different gene codes for each kind of receptor, and the Or83b receptor is unusual in two ways. First, nearly all the olfactory neurons in the fly have the Or83b receptor, whereas other kinds of receptors are divided up among small groups of olfactory neurons.
In addition, Or83b is found in widely divergent species of insects, including locusts, mosquitoes, moths, honeybees and medflies. "In all of these insects the gene is found in nearly all the olfactory neurons. The vast majority of cells responsible for smell have it," says Vosshall.
This is surprising because insects have evolved smell receptors tailored to their individual ecological niches: fruit flies detect fruit, for example, and mosquitoes sense humans and other warm-blooded animals. As a result, although different insect species have different genes coding for smell receptors, all of them have a gene that looks like the fruit fly Or83b gene.
Solving an old mystery
The ubiquitous Or83b gene has perplexed scientists since 1994, when Vosshall and her colleague Hubert Amrein, now at Duke University Medical Center, first identified it in fruit flies while they were postdoctoral fellows in Richard Axel's laboratory at Columbia University. But newly available genetic techniques allowed Vosshall to investigate its function by creating a strain of fruit flies lacking the gene.
"Only in fruit flies can you take the gene away and test the effect," she says, because -- unlike other insects -- fruit flies are common laboratory organisms with well-understood genetics.
The researchers wanted to know whether Or83b worked independently, like other odorant receptors, to sense a specific chemical, or played another more general role in the sense of smell. If Or83b receptors detected only one or a few odors, then flies without these receptors would still be able to smell other things. But if Or83b's function was a general one, underlying the fly's ability to smell, knocking it out would blind the insects to all odors.
Mutant flies lack a sense of smell
The researchers used a technique called gene targeting to create a strain of fruit flies lacking the Or83b gene. The flies appeared to be normal, but they could not smell at all, either as larvae or as adults.
The scientists tested the flies' sense of smell by three methods. First they placed larvae one at a time in the middle of a Petri dish, with a drop of fruity-smelling ethyl acetate near the edge. Normal fly larvae sense odors with cells in an organ on the top of the head, and they crawl toward the smell. The mutant larvae wandered aimlessly in the dish because they did not detect the smell. But a dose of Or83b receptor, delivered to the odor-sensing cells through a sort of fruit fly gene therapy, "cured" them -- the larvae crawled toward odors. That confirmed that the Or83b gene is essential to smell.
Vosshall and her colleagues also looked under the microscope to see, with special staining techniques, whether the olfactory neurons of the mutant flies had certain receptors and where these were located. In flies lacking Or83b receptors, the other smell receptors ended up in the wrong part of the nerve cells. Instead of being located at the ends of the cell dendrites, where they would be exposed to odor molecules in the air, they were clustered in the cell body.
Finally, they tested whether the nerve cells with incorrectly located smell receptors were activated in the presence of odors such as rosy-smelling 2-phenylethanol and methyl salicylate, which smells like wintergreen. They attached a tiny recording electrode to the antenna of a fly, then puffed the odorant molecules into the air nearby. In normal flies, a dip registered on the electroantennogram (similar to the chart that records heartbeats on an electrocardiogram), indicating that nerve cells had sent signals. The mutant flies, however, showed no response.
All these results show that Or83b is essential for fruit flies to smell. "Not every gene that is broadly expressed has an important function," says Vosshall. "But for a fly to smell anything the Or83b receptor has to be present."
"We still haven't figured out how or why it works," she adds. It could be that Or83b combines with other receptors to form different-shaped "locks" in which odorant molecules fit. Another possibility is that Or83b acts as a chaperone to direct other receptors to their proper placement in the cell. Or Or83b may be essential to the series of molecular signals that trigger perception of a scent.
"The reason this is really important is that we've reached an impasse in the fight against bad insects," says Vosshall. "We have DEET, but it has to be reapplied, you use large amounts, and every part of the skin has to be covered. It can't be used for infants or toddlers because of its toxicity. So it's not very useful against disease-vector insects."
"Now there's hope that, in Africa, in regions where malaria is endemic, bed nets could be impregnated with a repellent that works by blunting or eliminating the mosquito's olfactory response. For backyard barbeques, you might have candles that burn and release a repellent. Having a lot of different tools -- repellents to ward off biting insects like mosquitoes as well as drugs to fight the diseases they spread -- will help us prevent diseases like malaria."
Co-authors on this study are Mattias C. Larsson, now at the Swedish University of Agricultural Sciences, Ana I. Domingos, Walton D. Jones, M. Eugenia Chiappe, and Hubert Amrein, now at Duke University Medical Center.
This research was supported by grants from the National Science Foundation, the National Institutes of Health, and the McKnight, Beckman and Merck Foundations.