Why do mosquitoes prefer us over other animals? In episode 94, we talk with Zhilei Zhao and Lindy McBride from Princeton about their research into how mosquitoes that can carry dangerous diseases – such as Zika, dengue, West Nile virus and malaria – are able to track us down so quickly while ignoring other warm-blooded animals; an ability they’ve developed in just the past few thousand years. Their preprint manuscript “Chemical signatures of human odour generate a unique neural code in the brain of Aedes aegypti mosquitoes,” was posted to BioRXiv with multiple other co-authors on November 2, 2020.
Websites and other resources
News and Media
🔊 Access bonus content here.
We’re not a registered tax-exempt organization, so unfortunately gifts aren’t tax deductible.
Hosts / Producers
Ryan Watkins & Doug Leigh
How to Cite
What’s The Angle? by Shane Ivers
Zhilei Zhao: What’s very interesting about the system is: in Africa, the mosquitoes in Africa, they don’t specifically like humans.
Lindy McBride: Yeah, they are generalists.
Ryan Watkins: This is Parsing Science, the unpublished stories behind the world’s most compelling science, as told by the researchers themselves. I’m Ryan Watkins.
Doug Leigh: And I’m Doug Leigh. Today, in Episode 94 of Parsing Science, we’ll talk with Zhilei Zhao and Lindy McBride from Princeton about their research into how mosquitoes that can carry dangerous diseases – such as Zika, dengue, West Nile virus and malaria – are able to track us down so quickly, while ignoring other warm blooded animals … an ability they’ve evolved in just the past few thousand years. Here’s Zhilei Zhao and Lindy McBride.
Zhao: Hello, my name is Zhilei Zhao. I’m a graduate student in Lindy’s lab. Actually, the first graduate student in Lindy’s lab. I’m from, like, a small village in China. I went to Peking University in Beijing [to] study Life Sciences. So it’s very broad. Like, we study a lot of things. I did my thesis on the evolutionary genomics in Drosophila fruit flies. I took a gap year in the same lab to finish the study. And then I applied to the greatest school in the United States. And in Lindy’s lab, I’ve been studying … focusing on the neurobiology side, to study the question: at the neural level, how could the mosquito tell the difference between humans and animals?
McBride: And my name is Lindy McBride. I grew up in upstate New York, a suburb of Rochester, New York. I was always interested in animals and plants and being outside. I ended up choosing a really kind of wilderness-y place to go to college, Williams College in western Massachusetts, in the Berkshire mountains. Then I took a few years off. I wanted to travel – I knew I wanted to go to graduate school, but I decided to travel a little first – I did field research in Tanzania. I became an Outward Bound instructor in Minnesota for a year. Did some more field research in Peru. And ended up at University of California Davis for graduate school. And from there – I was a grad student, actually for seven years, so it’s quite a long PhD – then I went and did a postdoc at Rockefeller University in New York. Before starting my position here at Princeton University, where I’m jointly appointed in the Department of Ecology and Evolutionary Biology and the Neuroscience Institute.
Watkins: A mosquito is any member of a group of about 3500 species of small insects belonging to the order diptera, or flies. And, as we learned in Episode 78 with Richard Bomphrey – on how mosquitoes can detect surfaces using the airflow caused by the movement of their own wings – only female mosquitoes bite. Since Zhilei and Lindy study the yellow fever mosquito – Aedes aegypti – we asked them to tell us more about this particular species.
Aedes aegypti‘s origins
Zhao: The species we study is called Aedes aegypti. It’s a mosquito found in the tropical and subtropical regions all over the world. It’s a very dangerous disease vector because it spreads virus like the yellow fever, dengue in recently the Zika virus. And one major reason they become so efficient in transmitting disease is they specifically target humans. If you present a human and animal, most of the time the mosquitoes which was a human is a targets for blood feeding. Therefore [they] transmit the disease to humans. And what’s very interesting about the system is: in Africa, the mosquitoes in Africa, they don’t specifically like humans.
McBride: Yeah, they are generalists.
Zhao: Yeah, that’s right. Sometimes they don’t have a strong preference of humans over animals. Sometimes they even prefer the animals. And we know from the genomic study, the mosquito in Africa is the ancestral form. And the mosquitoes that prefer humans actually evolved from those generalist mosquitoes. So the idea is, during evolution, human preferring form of the Aedes aegypti evolved from the ancestral generalist form in Africa. It’s very recent evolution. So our research question is to study how this process happened during evolution.
McBride: Yeah, the species is probably about 200,000 years old. It diverged from ancestors in the Indian Ocean islands, or possibly Madagascar, and arrived on the African continent. And this is somewhat speculative, but this is what people think. And then, you know, became what we identify today as Aedes aegypti in Africa on the order of 200,000 years ago. But the human preferring form of Aedes aegypti then emerged about 10,000 years ago. And that form is considered a subspecies. Yeah, a postdoc in the lab actually just did, I think, some of the best genomic dating to date. No pun intended. And he estimated that the shift to humans probably occurred around the end of what’s called the “African humid period.” And during the African human period – which was, say 10 to 15,000 years ago – the Sahara was actually quite green. It was more of a Savanna, much wetter than it is today. And it looks like as the humid period ended, and the Sahara dried, that’s when these human preferring mosquitoes evolved. And that’s relevant because we think they originally evolved to specialize in biting humans – not because our blood is so sweet or anything like that – but because breeding in humans stored water is possibly the only way to survive and really arid or seasonal climates.
[ Back to topics ]
Genetically engineering a model mosquito
Leigh: To examine the neural mechanisms at work when they detect the scent of humans, Zhilei and Lindy required some specialized mosquitoes. And CRISPR – a technique by which genomes of living organisms can be altered by the transfer of genes from another species or breed – was the ideal tool for creating genetically-modified transgenic mosquitoes, with the unique characteristic that regions have their brains would glow with fluorescent protein when activated by odors. Up next, Zhilei explains what but then to approach their study in this way.
Zhao: I think the key aspect for us is we want[ed] to get a comprehensive picture of how the neural response looks. Because before people [had] done like the electrophysiology: they plug an electrode into the antenna and record the activity. But you cannot get all the neurons from those recordings. In the beginning, for us, we use CRISPR-Cas9 – genome editing tools – to create a transgenic mosquito. And those mosquitoes are broadly useful, not only for our own study, but also for the community. We developed a transgenic mosquito that labels all the neurons in the brain. For us, we just focus on the olfactory side. But other people – since we post[ed] the paper to BioRXiv – several groups have contact[ed] us for the transgenic mosquito, because they can use this line for their own questions. For example, some people might study the taste, some people might study the thermo-sensation … so those too are broadly useful for them. For us, we focus on the brain region called antenna lobe. It’s where the axon of the sensory neurons – an antenna – project to the brain. So it’s the primary station of the olfactory information in the mosquito brain. And what we did is we target gene called orco. It’s a co-receptor for the olfactory receptors. So it’s broadly expressed in the olfactory sensory neurons. And we target this gene; we insert a transgenic construct that has a protein called GCaMP. It’s a fluorescent protein, so it kind of lights up in the two-photon microscope image. It’s a very useful protein, because when the neuron is active, the fluorescence will increase because the protein is sensitive to the calcium concentration. And when neurons are active, the calcium concentration increase, so the fluorescence increase[s]. And once we have this mosquito, we can just remove the cuticles, and we can look at the neural activity by measuring the fluorescence under the two photon microscope. [How] it shows up … when we do the two-photon calcium imaging is we have … you can imagine we have a kind of an image of the antenna lobe. So basically, it’s a lot of circles in one image, but it’s the base nine: the fluorescence is kind of low, so it’s kind of dim. But when we stimulate the mosquito with one human odor, for example, we might see one of the circle[s] increase in fluorescence. And we know this region is active to these human odorants.
McBride: I was just gonna ask Zhilei, do you actually watch the screen when you do the recording? To see what’s happening?
Zhao: Yeah, I do. Yeah, I still remember the first time we had the transgenic mosquito and we want[ed] to try it on a human odorant. I still remember the excitement when I saw one of the brain region kind of lights up when we stimulate the mosquito with the odorant.
[ Back to topics ]
Hypothesizing how mosquitoes preferentially target humans
Watkins: “Glomeruli” are spherical structures located in the olfactory bulb, and are the initial sites for the brain’s processing of odor information coming in from the nose, or – in the case of mosquitoes – their antenna lobe. In addition to possessing universal glomeruli that are sensitive to both human and animal odors, Aedes aegypti have glomeruli that differentiate the smell of humans from animals. So they’re an ideal location to decipher how mosquitoes decode complex odor blends, as we’ll here after the short break.
Watkins: Here again [are] Zhilei Zhao and Lindy McBride.
Zhao: The behavior of the phenomena we want to explain is: why the mosquitoes can tell the difference between human odor and animal odor. And we know for those mosquitoes, they like the human odor. So based on this phenomena, we have three hypotheses about how this can be achieved at the neural level. One way is very simple, we can imagine maybe the human odor just evokes much stronger, like overall much stronger response in the mosquito brain, compared to the animal odor. That’s the first hypothesis. The second hypothesis is maybe these specific brain region that responds only to the human odor, not the animal odor. So we can imagine maybe this one or two commanders that are human unique, so that I can explain why the mosquitoes like the human odor. The third hypothesis is because we know the human [odor] and motor are very complex, we can imagine maybe there is a very complex pattern of neural activity in the olfactory center, almost the brain. And we can only achieve the discrimination by building some complex model. Those are the three hypotheses we have.
[ Back to topics ]
Extracting and delivering odors
Leigh: In order to compare mosquitoes’ reaction to the scent of humans versus that of other warm-blooded animals, Zhilei and Lindy collected samples from eight humans, two rats, two guinea pigs, two quail, a sheep, and four dogs, as well as two nectar-related stimuli, the mosquitoes find attractive: milkweed flowers and honey. But collecting these diverse samples was only half of the equation: they also had to deliver them both to the same mosquito in exact odor concentrations, and deliver replicate puffs of the same sample to different mosquitoes, all while maintaining the original blend ratios. And as no existing odor delivery technology existed to accomplish this kind of precise quantitative control, they had to develop one themselves.
McBride: The actual collection of odors, those protocols have been well established by other groups. They’re kind of normal ways to collect odor. But you know, most people who study olfaction use synthetic odorants that you can buy from Sigma [Aldrich] or something. And that’s because, you know, using simple stimuli can provide a lot of insight. And also because it’s just much easier. Using complex, like, real odors … they’re hard to control, they’re hard to, you know … it’s hard to bring the source to the microscope to produce the odor. And in previous works … people had done this sort of thing with fruit, where you can just put fruit in a little bottle and puff air through the bottle. Or anything really, as long as it’s small enough, you can get it into this bottle. But we obviously couldn’t bring, you know, animals and stick them in this bottle, let alone humans. So typically, for an animal – for a small animal – we put them in, ideally, a glass container. And you blow air – clean air – through the container, and you collect the volatile chemicals on a filter as the air leaves the chamber. For some larger animals – like we wanted to do sheep and dog – you know, we can’t easily get a glass chamber to put those animals in, and also approval to do so. So, for those animals, we can use hair and put the hair in a glass bottle and extract the odor from hair. It’s not probably, almost certainly, not exactly the same as the whole body odor. But we think it’s certainly a start, good enough for our purposes. And then for humans, you know, you can have a human put their arm in the chamber. A lot of people do that or their hand. But we had some collaborators in Sweden who developed a body bag assay where you have a human get into a big Teflon – or somehow odorless bag – and lie on a cot and watch a movie for two hours while air is slowly being blown through the bag. Of course, you want to avoid using any sorts of products or lotions or soaps for several days in advance: get really stinky and natural and then get in the bag. Another alternative would be to have the human lying in a body bag next to the microscope for every experiment, but the logistics of that would be really difficult. So we needed to collect the odor in advance which we did. But then you have all this odor … and the question is how to get it to the mosquito and the mosquitoes flailing under this fancy microscope and the neuroscience center. And there’s just all sorts of technical details of why that’s difficult. You can try to elute the odor in a solvent, but then all of the different hundreds of components will start evaporating out of [the] solution at different rates, and the blend that’s in the air will actually be quite different than the blend that’s in [the] solution. So we didn’t know what to do about this. This is actually one delay we had. Zhilei was very fast and came up with all these creative ways to make transgenic mosquitoes that we could use for imaging. And then we spent at least six months or a year trying to figure out what we’re going to do with these mosquitoes because we didn’t have a way to deliver human odor that was reliable. And around the same time we decided that we needed our own GCMS machine in the lab – a gas chromatography, mass spectraphotometer – for analyzing odor. Because I was kind of in the middle of negotiating with all these companies and learning about that technology – it’s not something that we had done before – we came across this approach for injecting odor into a GCMS. And it’s called thermal desorption. You basically take your filter that has that compounds on it, and you stick it into the thermal desorption unit, and it heats it up really fast – ballistically – to like 300 degrees Celsius in a couple seconds. And that releases all the odorants, supposedly simultaneously, off this little filter. And I thought, “Well, if we can use that to get the odor into the GCMS, why can’t we use that to get the odor to the mosquito?” So we talked to the company about it. They were of course willing to be paid to come out and help us set it up. And Zhilei worked with one of the people from this [company], Markes International, for several days. And, is that the time you are referring to Zhilei? When the excitement of seeing … I mean it’s the first time, like, we got human odor onto the mosquito and could see a response. Mind blowing for us.
[ Back to topics ]
Watkins: For their study, Zhilei and Lindy wanted to observe the neurological activity of the Aedes aegypti while they’re alive and actively detecting odors. To do so they use two-photon excitation microscopy, a fluorescence imaging technique that allows for the imaging of living tissue up to about one millimeter in thickness. Doug and I wanted to learn more about what it’s like to work with such a microscope.
Zhao: The microscope is huge, because we have different components, like the laser or the light path of the objective. But the mosquito is very tiny. Like, the brain of the mosquito is 500 microns by 500 microns by 200 microns. So it’s half a millimeter in diameter, basically. So it’s very small. So what we [did] is we designed a holder that kind of has a very thin metal plate. And we drill a very tiny hole on the plate. Then we could do the head of the mosquito to the tiny hole. We use UV glue because it’s very convenient to use. Because, usually, it’s a liquid, but when we shine the UV [light] it can also solidify. Then what we do is we use the forceps to remove the cuticle [that] was covering the brain, to expose the brain. Then we place this holder – this mosquito – under the objective of the two-photon microscope so we can see how the fluorescence things change.
[ Back to topics ]
How humans and animals odors differ
Leigh: Human body odor is a unique identity feature of each individual, as well as an established composite of numerous volatile organic compounds, or VOCs … which we first learned about in Episode 41 with Jonathan Williams, who analyzed their emission during movies as a marker for stress, and therefore of how scary they might be. One such VOC, decanal, occurs naturally – and is used in fragrances and flavoring – and differs in abundance not only among humans, but also between humans and animals. So we asked how the odor profiles of humans and animals differ.
McBride: Do you wanna answer that, Zhilei?
Zhao: Yeah, sure. Actually, when we look at the odor profiles of humans or animals, there’s no human-unique compound … like in the strictest sense – like only in humans, not in animals – there’s no such compound. But we have a compound that’s much more abundant in humans, compared to animals. One compound is called decanal. And this compound is very interesting because it’s produced from fatty acid on the skin. And we know this fatty acid – sapienic acid – acid is human-unique. And you can tell from the name, “sapienic,” it’s named after the Homo sapiens. So people think this fatty acid is human-unique. And this fatty acid will be oxidized into the decanal. And that can explain why humans have much more decanal than animals. And [there’s] research saying, like, the sapienic acid plays some role in protecting our skin. So you can imagine – because we can all lost our hair during evolution – maybe our skin cannot separate sapienic acid to protect our skin. And sapienic acid gets broke down into the decanal. And mosquitoes took advantage of this, and used this unique kind of abundant compound in human odor to achieve the discrimination. That’s kind of the scenario in our mind[s].
McBride: Yeah, that’s what we think. I guess we don’t know that for sure yet, but they’re very sensitive to decanal. So it makes sense that they would use that signal for discrimination.
[ Back to topics ]
Differences within humans’ odor profiles
Watkins: In figure 4A of their manuscript – which you can view at parsingscience.org/e94 – Zhilei and Lindy present the odor profiles of the people, animals, and plants from which they conducted their samples. In it, the odor profile of one of the human subjects stood out as having relatively fewer decanals, and having a profile more similar to that of a dog. So Doug and I wondered if this might explain why some people claim they’re almost never bitten by mosquitoes, while others of us seem to get bitten all the time.
Zhao: The first person in the figure, we know this person is actually … is kind of [an] outlier for other humans, or the average human. And we know from their behaviors, this person is not very attractive to mosquitoes. So the idea is maybe if you have a, kind of, abnormal ratio of those compounds in your odor profiles, maybe you are a not attractive to the mosquitoes. And we did another analysis, where we correlate the attractiveness of the human subjects with the abundance of long-chain aldehydes. And we saw a very nice correlation between the attractiveness and the distance of this human subject to a typical .. like an average human. So the idea is: if you are very different from an average human, you are less attractive to mosquito.
McBride: Yeah, as Zhilei said, there are real differences between humans and how attractive their odor is to mosquitoes. But there’s some data to show that people’s perceptions of how attractive they are, are driven mostly by how strongly they react to bites. And as a case in point, there’s a postdoc in my lab whose husband absolutely does not respond to mosquito bites. Like I’ve never seen someone like this before. He can’t feel them biting, there’s absolutely no red dot, there’s no itch. So he could be bitten tons and he would never know it. Whereas the person who has a welt that hangs around for three days, they’ll remember every bite. What also makes a lot of sense, in light of that interpretation, is that for someone to say, “you know, I’m so attracted to mosquitoes” … well, there’s different mosquitoes living in different places that are flying at different times of year. So it would make more sense that if your immune system is highly reactive, then you respond to the bites of all sorts of different mosquitoes – or even no-see-ums – strongly, [it] doesn’t make as much sense to me that all of these different biting insects would be using the same cues to find you.
[ Back to topics ]
Deterring mosquitoes by altering our scent or their olfaction
Leigh: Given that they strongly prefer the smell of humans over non-human animals, we asked if there might be a way to dissuade mosquitoes – or at least make us less visible to them – by altering our scent, either by way of some kind of topical solution, or by something we could ingest, to change the way we smell to them.
McBride: You know, I used to be very skeptical of the idea that we could manipulate mosquitos’ sense of smell or manipulate human odor to, you know, prevent disease transmission. But, you know, after getting these results, I’m a little bit more open to that idea. Certainly, I think that the data we collect could help us design super-attractive lures to bring mosquitoes into lethal old traps, for example. Or repellents, you know, we could look for compounds that inhibit the activity of these decanal-sensitive neurons. But in terms of altering human odor … I don’t know, it’ll be really interesting to see. I mean, you wouldn’t … you could imagine just decreasing, for example, sebum levels across the board, but then you might also have side effects there, you know, related to the function of sebum.
Zhao: Yeah, there’s some recent research … like, we know that currently the most efficient repellent is, like, the DEET. And there’s our recent research saying, like, the reason why DEET works so well is it kind of reduce[s] the volatility of the other compounds on the human skin. So I can now mask the human odor from the mosquitoes: it’s what one recent study suggested.
[ Back to topics ]
Future technological advancements
Watkins: One of our favorite things about the paper is that Zhilei and Lindy used CRISPR, a technology that scarcely existed just 10 years ago, and which just last year, was a major contributor to the Nobel Prize in Chemistry awarded to biochemists Emmanuelle Charpentier and Jennifer Doudna. This led us to wonder what kinds of technologies Zhilei and Lindy believe might come to fruition in the near future and at – say, 10 years from now – we might similarly count on as a regular part of science.
McBride: CRISPR can’t be overstated … its impact on science, of course, in humans. Certainly for us, like, we couldn’t have dreamed of doing these things. Well, I actually did dream of doing some of these things before CRISPR. But I was gonna do it in a different way that would have been much less successful. But – I don’t know what you think Zhilei – but for me, we have an issue right now with a resolution of our microscope. And I just feel like that’s just going to disappear within a couple years. We do volumetric imaging where we get a whole 3D volume of the antennal lobe’s response to an odor three times every second. And when we do that volumetric imaging, the resolution is quite low, which can make it difficult for us to identify the individual glomeruli. Those plots in the paper, those – we call them “cloudograms” – where you have these like, little clouds of red corresponding to activated glomeruli … it would be nice to be able to do that in a slightly more concrete way. That’s one thing: the optics. That problem will be solved.
Zhao: For me a lot of things on the transgenetic side. Like into Drosophila – in fruit flies – we have so many tools that can basically give you access to a single neuron, a specific type of neuron. But in mosquitoes, we don’t have that yet. So, for example, if we say we want to control a specific neuron in the mosquito brain, we cannot do that yet. So one very promising direction, I feel like, is to develop those cell-type specific drivers for mosquitoes. So we can study specific neurons in the central brain of mosquitoes … more fine control of what neuron we are targeting.
McBride: Yeah, absolutely. Right now we can image from all neurons or from all olfactory sensory neurons. But we can’t image from one particular class, you know, among a 100,000 different neurons. Yeah, you know, let’s say the brain has hundreds of thousands of neurons in these insects. We can’t image from just one of those 100,000. In Drosophila, we can do that: we can access just one neuron. We can either activate it optogenetically, or we can record from it using GCaMP. But with mosquitoes, we’re limited right now to large sets of neurons.
[ Back to topics ]
How neuroscience can inform other disciplines
Leigh: As neuroscientists who study the genetic and neural basis for behavioral evolution, Zhilei and Lindy operate in a field that’s both influenced by, and itself influences, a wide variety of disciplines from computer science to psychology. So Ryan and I closed out our conversation by asking what they believe researchers in other fields can glean from the methods of theirs.
McBride: This wouldn’t apply, you know, to some scientific fields like math or physics. But I think for psychology … I mean, one thing that we value a lot in the lab is natural diversity, natural variation. We could study this human-preferring mosquito without thinking about its evolutionary context. But we always have that in mind. And we take advantage of these comparisons that are presented to us in nature, like between these two subspecies of Aedes aegypti. So personally, that resonates with me: taking advantage of the diversity that exists in nature to understand science. But that’s kind of relevant other [natural] scientific fields.
Zhao: For me, I’ll probably emphasize on the technical side. Our research is not possible without the innovation in the methods or in the techniques. So I feel like students or researchers from other fields – for example, like computer science, engineering, or even math – they’ll play very large roles in biological research, because they can bring in new methods, and also new concepts into the research … biological research.
[ Back to topics ]
Links to manuscript, bonus audio and other materials
Leigh: That was Zhilei Zhao and Linda McBride discussing their preprint manuscript “Chemical signatures of human odor, generate a unique neural code in the brain of Aedes aegypti mosquitoes,” posted to bioRxiv on November 2nd 2020, along with multiple co-authors. You’ll find a link to their paper at parsingscience.org/e94, along with transcripts, bonus audio clips, and other materials that we discussed during the episode.
Watkins: If you missed any recent episodes of Parsing Science – like our conversation with Luke Cuddy about what we can learn from a video game about our own theory of knowledge, or with Ángela Zorro Medina about the unintended consequences of legal reforms – then head over to parsingscience.org where you can have a listen to any of our previous episodes.
[ Back to topics ]
Preview of episode 95
Leigh: Next time, in Episode 95 of Parsing Science, we’ll talk with Shiri Melumad from the University of Pennsylvania’s Wharton School of Business about her research, which shows that when – much like the children’s game of telephone – news is repeatedly retold, it undergoes a stylistic transformation through which the original facts are increasingly replaced by opinions and interpretations, with a slant towards negativity
Shiri Melumad: In contrast to conspiracy theories or fake news, this distortion seems to be coming from a benevolent place, right? So again, the reason it’s happening is because as a reteller, I’m concerned about my audience sort of getting the key takeaways, and I want to help guide them on it.
Leigh: We hope that you’ll join us again.
[ Back to topics ]