All That Buzz
The newest hope for disease control? Hack mosquitoes’ sense of smell.
Story by Alexander Gelfand • Photography by Harry Giglio
"This,” Conor McMeniman, PhD, says with a flourish, “is where the magic happens.”
A wry grin plays across McMeniman’s face. He and his research assistant, graduate student Winter Okoth, are standing inside a walk-in mosquito incubator on the fourth floor of the Bloomberg School—one of several in the Insect Core Facility of the Johns Hopkins Malaria Research Institute (JHMRI), which McMeniman joined last fall as an assistant professor of Molecular Microbiology and Immunology. Directed by Nobel laureate Peter Agre, MD, JHMRI is dedicated to pursuing basic scientific and field research that will help cure and prevent malaria.
Though McMeniman is clearly delighted, the room—close, warm and humid—is an entomophobe’s nightmare: The walls are lined with shelves containing tens of thousands of mosquitoes at various stages of maturation, from eggs and larvae floating in water-filled trays to adults housed in cages whose floors are splattered with the dull red remains of recent blood meals. All belong to one of the most dangerous animal species on the planet: Anopheles gambiae, the primary vector of malaria in sub-Saharan Africa. For countless millennia, An. gambiae has played host to Plasmodium falciparum, the deadliest strain of malaria parasite; as such, the winged arthropod plays a leading role in the spread of a disease that sickens more than 200 million people around the world each year and kills nearly 500,000—most of them Africans under the age of 5.
McMeniman—blonde, blue-eyed and well over 6 feet tall—grabs a cage full of adult mosquitoes and hoists it aloft, blowing gently through one of its mesh walls. Roughly half of the 400 or so mosquitoes inside immediately go into a frenzy, flying about in every direction. “That’s called activation,” McMeniman says, referring to the way in which the carbon dioxide in his breath prompts the blood-sucking insects to look for someone to bite. (Only female mosquitoes consume blood, which provides a source of protein essential to egg production, so only they exhibit such host-seeking behaviors.)
He then holds the palm of one hand close to the same mesh wall. The female mosquitoes, already primed for action by the carbon dioxide, instantly swoop toward it, homing in on his body heat. The same thing happens when he puffs into the cage and presses a hot plastic water bottle against it; the mosquitoes are so excited that they begin probing through the mesh with their needle-like proboscises, stabbing at the hard plastic as if it were a warm-blooded snack.
Some of the trays in the incubator are labeled with cryptic words like “ORCO” and “GFP.” These represent lines of genetically modified mosquitoes that McMeniman and Christopher Potter, PhD, an assistant professor of Neuroscience in the School of Medicine, are using to better understand how—and what—mosquitoes smell. Potter is interested in the basic neuroscience of mosquito olfaction. McMeniman, working with Okoth, postdoctoral researchers Genevieve Tauxe, PhD, Shruti Shankar, PhD, graduate student Gibran Nasir and others, is after something different: He wants to understand how the mosquito’s sense of smell leads it to us, so that we can manipulate it to our advantage.
The consequences for global health could be immense. Even the best vaccines against malaria are only roughly 30 percent effective, and P. falciparum, the parasite that kills so many every year, is disturbingly adept at developing resistance to the drugs that are used to treat the disease. (Resistance to the latest class of antimalarial drugs, known as artemisinins, has already appeared in Southeast Asia, the same region where resistance to their predecessor, chloroquine, first surfaced in the late 1950s.) Consequently, vector control, or managing the insects that spread the illness, is crucial.
Unfortunately, while conventional vector-control measures like indoor spraying and insecticide-treated bed nets—both of which were designed specifically with indoor, night-biting mosquitoes like An. gambiae in mind—have been successful in some locales, they have not succeeded in shutting down malaria transmission everywhere. People continue to become infected for a variety of reasons, from faulty bed nets to mosquitoes that are capable of evolving resistance to insecticides and of altering their feeding habits—foraging outdoors instead of indoors, for example, or during the day instead of at night. As a result, new methods of control are urgently needed.
This is why McMeniman’s JHMRI colleagues Marcelo Jacobs-Lorena, PhD, and George Dimopoulos, PhD, are trying to use genetically modified bacteria to kill the malaria parasite before it can exit its mosquito host, and also boosting the insect’s own immune system in order to thwart infection and block transmission.
It is also why McMeniman and his fellow mosquito olfaction researchers are trying to determine how smell drives mosquitoes toward us. Such knowledge could lead to the development of better repellents that would prove noxious to mosquitoes but not to us, of chemical compounds that might confuse mosquitoes’ sensory systems so they could not target us effectively, or even, someday, to genetically modified mosquitoes that aren’t interested in us at all.
“The more you understand about an organism, the better chance you have of controlling it,” says James Logan, PhD, an entomologist at the London School of Hygiene and Tropical Medicine who, like McMeniman, uses the methods of modern neuroscience and molecular biology to study and manipulate the mosquito’s sense of smell.
McMeniman’s work could also prove a boon to entomologists like JHMRI’s Douglas Norris, PhD, who conducts research at the Institute’s field sites in Zambia, a country where malaria is endemic. As Norris explains, researchers trap mosquitoes for various public health purposes. If you want to determine the transmission rate of a mosquito-borne illness in a particular region, for instance, you must first figure out how often people are being bitten. Similarly, if you want to put effective control measures into place, you must determine which particular species or subspecies of mosquito is involved. (It would not make much sense to invest heavily in bed nets and indoor spraying, for example, if the mosquitoes involved prefer to bite outdoors during the day.) Both require capturing representative samples of insects, like the ones whose tiny corpses are sealed within dozens of Ziploc baggies in Norris’ lab in Baltimore, awaiting identification through genetic analysis. But current trapping methods leave much to be desired.
The gold standard, known as a “human landing catch,” uses people as bait. It therefore involves a certain amount of risk—participants are, after all, subjecting themselves to attack by malaria-carrying mosquitoes. It is also labor intensive: Field workers must be rotated through nightlong trapping sessions. Mechanical traps, which use carbon dioxide and other attractants to lure mosquitoes, are cheaper and easier to manage, but they work better indoors than out and don’t much resemble an actual human host. If McMeniman’s research leads to compounds that more accurately mimic the scent of a real person, researchers could collect as many mosquitoes as they like without ever having to use flesh-and-blood bait. Better mechanical traps might even be capable of capturing—and killing—mosquitoes in large enough numbers to affect transmission.
Growing up in Brisbane, Australia, a city whose subtropical climate guarantees regular mosquito infestations, McMeniman pretty much hated mosquitoes. It was only after he began studying virology and parasitology at the University of Queensland that he became intrigued by their role as an important vector for various pathogens and decided to help develop new ways of tackling the mosquito-borne diseases that afflict hundreds of millions of people around the world every year.
"A MOSQUITO CAN ONLY DETECT YOUR BODY HEAT ONCE IT IS A FEW METERS FROM YOUR SKIN, BUT IT CAN PICK UP YOUR SCENT FROM 30 METERS AWAY."
Success came early. As a doctoral student at Queensland, McMeniman found a way of transferring a strain of the bacterium known as Wolbachia from Drosophila melanogaster, aka the vinegar fly, and more commonly referred to as the fruit fly—an intensively studied, easy-to-rear insect that serves as a model animal for mosquitoes, much as rats and mice do for humans—into Aedes aegypti, a day-biting mosquito that has the distinction of being the primary vector for dengue virus, yellow fever, chikungunya and Zika. The idea was to shorten the mosquito’s lifespan, thereby interfering with its ability to spread dengue in particular. (Dengue virus requires a long incubation period in its insect host, so if the insect dies too early, it can’t pass its viral cargo along.) The results exceeded all expectations: Wolbachia has since been shown to completely inoculate Ae. aegypti not only against dengue but also against chikungunya and yellow fever; as-yet-unpublished research indicates that it is also effective against Zika. As a result, Wolbachia-infected mosquitoes are now being field-tested in locations ranging from Vietnam to Brazil.
McMeniman then landed a postdoctoral position at Rockefeller University in the lab of Leslie B. Vosshall, PhD, a molecular geneticist who did pioneering work on olfaction in flies before branching out into humans and mosquitoes.
The idea that mosquitoes can smell anything at all, much less use that sense of smell to target us with malign accuracy at night under poor light conditions, might seem counterintuitive. They do not, after all, have noses. They do, however, have their own highly sensitive olfactory organs; namely, their antennae and a couple of nearby antennae-like structures called the maxillary palps. All are loaded with olfactory sensory neurons like the ones that line our own nasal cavities. (Such neurons are also found on the mosquito’s proboscis and even on its feet, but these are believed to mediate taste more than smell.) Each of those neurons expresses one of many different finely tuned olfactory receptors: proteins that bind to specific odorants, causing signals to be sent into an area of the insect’s brain called the antennal lobe. These receptors are grouped into several different classes, and the signals generated by each class of receptor are in turn routed to a particular cluster of neurons, or glomerulus, located in the antennal lobe.
In the absence of a vaccine, the only way to manage the Zika outbreak sweeping across Latin America and beyond is to control the virus’s primary vector, Aedes aegypti. But that won’t be easy. According to McMeniman, Ae. aegypti is an incredibly resourceful mosquito. Active during the day, it feeds both indoors and out, and it can bite right through thin clothing. What’s more, it breeds in standing water in and around human habitation—and it doesn’t require much: A soda-bottle cap will do. Removing such breeding sites is therefore crucial but can be tough in places where plumbing is scarce and people must store water for drinking and cleaning. Insecticides are helpful, but over the long term, resistance can be an issue. The Wolbachia-infected mosquitoes that McMeniman helped develop could prove to be even more useful and are currently being tested in Rio de Janeiro.
Meanwhile, people living or traveling in Zika zones can take personal protective measures like ensuring that doors and windows have screens; applying an EPA-registered mosquito repellent to exposed areas, especially the lower extremities where Ae. aegypti likes to bite; and covering up. McMeniman advises wearing long sleeves and proper shoes (“Sandals won’t cut it,” he says), and tucking pants into socks. Best to avoid spandex, he adds: “They’ll bite straight through that.” » AG
Mosquitoes also use other kinds of sensory information, like heat and visual cues, to locate us. But olfactory signals seem to be overwhelmingly important—so much so, says Logan, that under certain circumstances they can even override the others. These signals also operate over considerable distances: A mosquito can only detect your body heat once it is a few meters from your skin, but it can pick up your scent from 30 meters away.
In recent years, researchers have determined which receptors are responsible for picking up many of the chemicals that draw mosquitoes to us, including carbon dioxide. They have also identified some of the genes that allow mosquitoes to manufacture these receptors, and have learned how to manipulate those genes.
At Rockefeller, for example, McMeniman engineered a strain of Ae. aegypti that could not sense carbon dioxide. The goal was to see if these CO2-insensitive mutants would still be attracted to human hosts. If not, compounds that interfered with mosquitoes’ CO2-sensing abilities might help ward the insects off. Indeed, other researchers had already identified a number of such compounds in anticipation of their usefulness.
Alas, McMeniman’s experiments revealed that even CO2-blind mosquitoes could still track us down. They were, in fact, only 15 percent less likely to find their human targets, presumably because they were still able to make use of other sensory signals, such as body heat, visual cues and the many and diverse compounds that comprise human body odor. As McMeniman and others have demonstrated, the telltale signs that mosquitoes use to zero in on us often reinforce one another, resulting in a highly redundant targeting system that is not easily disabled.
Those findings have all sorts of implications, many of which now drive McMeniman’s research program at the JHMRI. For one thing, it seems likely that attacking a single olfactory pathway—like the one governing carbon dioxide reception—will not be the most effective way of either repelling or attracting mosquitoes. Instead, the best baits might consist of blends of many different chemicals that mimic the specific components of human scent that are most attractive to mosquitoes, combined with other cues like heat.
The best repellents, meanwhile, might be ones that are either innately repulsive to mosquitoes or ones that so effectively confuse their olfactory circuits that the insects can no longer pinpoint us, given the overall importance of smell and the distance over which it travels. Vinegar flies, for example, are highly averse to the smell of geosmin, a compound produced by bacteria that are toxic to them. Mosquitoes might also have a built-in aversion to specific odors. And there is evidence that DEET, the most common mosquito repellent on the market, works by stimulating some olfactory sensory neurons but inhibiting others. This, in turn, suggests that it might function not only by actively repelling mosquitoes but also by scrambling the olfactory “code” that allows them to perceive us—something that other chemical compounds might do even more effectively.
More broadly, McMeniman’s previous research suggests there is much more work to be done on the basic neuroscience of mosquito olfaction—on the chemical cues, the neural circuits and the underlying genes that drive mosquitoes to feed on us—as well as on the range and relative importance of the various odors to which they respond.
This is no simple task: Human body odor is a complex mixture of chemicals, and research has shown that mosquitoes are not only attracted to the odorants that we exhale and secrete but also to the ones that are produced by the bacteria that live on our skin. Scientists have even demonstrated that mosquitoes find the volatile compounds produced by certain strains of bacteria more appealing than others, suggesting that our delectability depends to some extent on our individual microbiota. In addition, the malaria parasite itself appears to influence both mosquito olfaction and human body odor. For example, Logan and his colleagues have shown that mosquitoes infected with P. falciparum are more strongly attracted to human body odor than uninfected ones—while others have demonstrated that malaria-infected humans are similarly more attractive to mosquitoes than their uninfected counterparts. Sorting through all of this is going to take a very sophisticated, multipronged approach.
Luckily, this is just what McMeniman has planned—with a little help from his friends in the Potter lab.
Using techniques he first devised for vinegar flies, Potter has genetically engineered An. gambiae to express green fluorescent protein, or GFP, in an important class of olfactory sensory neurons. He has done this using a gene called orco that those neurons need in order to function; hence the labels (“GFP,” “ORCO”) on the trays back in the mosquito incubator.
The results have been illuminating. Seated before a computer screen, Olena Riabinina, PhD, a postdoctoral researcher in Potter’s lab, is able to call up an image of a mosquito’s antennal lobe that resembles a lumpy Christmas tree. Most of it is a crenellated blue mass. In the upper left- and right-hand corners, however, two small, blobby ornaments glow a vivid green, and a series of similarly fluorescent strands stream away from its lower circumference.
The strands are axons that carry messages from the olfactory sensory neurons in the insect’s antennae; the blobby ornaments are glomeruli that receive those messages for processing. Together they provide a partial blueprint of An. gambiae’s olfactory system—a wiring diagram, as it were, that illustrates how signals travel from at least some of the mosquito’s olfactory receptors into its brain. McMeniman is already working to expand this neural schematic by labeling yet another class of neurons that is responsible for detecting human odor. Meanwhile, Potter recently managed to insert a calcium-sensitive protein called GCaMP into An. gambiae’s olfactory sensory neurons. Whenever a neuron receives an incoming signal, calcium ions rush in, causing GCaMP to fluoresce. As a result, he and McMeniman are now in a position to waft odors over their modified mosquitoes to see which neurons light up in response.
McMeniman plans to use a sophisticated form of chemical analysis known as gas chromatography to figure out all of the different components that make up human body odor—including that of malaria-infected volunteers from places like Zambia and the Democratic Republic of Congo, where the JHMRI is also beginning to establish field operations.
He then intends to puff those components onto the sensory organs of mosquitoes that have been kitted out with calcium sensors and to image their antennal lobes in real time, observing which parts of their brains fizzle and pop in response to specific chemicals. And he will do this not only with uninfected mosquitoes but also with ones that carry Plasmodium.
Together with a more comprehensive wiring diagram of An. gambiae’s olfactory system, these experiments will give McMeniman what he calls a “proper receptor-to-glomerulus” map, along with a list of the scents that evoke the strongest neural responses in his bloodthirsty test subjects. McMeniman will then see how these odors actually affect mosquito behavior by trying them out on infected mosquitoes confined to a wind tunnel in the insectary, and then on insects out in the field in Africa—something he hopes to do within the next two years. And while he is currently focused on combating malaria, he would in the future like to apply his methods to Ae. aegypti and its rogues’ gallery of viruses, including Zika.
It’s an ambitious plan—not least because McMeniman’s winged nemeses have evolved over millions of years to be stubbornly persistent human-seeking missiles. “They’re wired to find us,” he says of An. gambiae and its kin. But if he and his colleagues can hack their olfactory systems, they might just find a way to put these tiny, disease-ridden bloodhounds off our scent once and for all.