A persistent behavioral state enables sustained predation of humans by mosquitoes

Background

Following an odor trail back to the source is an incredibly difficult task. Yet, as humans, we aren’t particularly aware of hard this task is. Who has the time to painstakingly follow that appetizing smell wafting into the room when you know you can run to the kitchen to check it out? Or if you are outside, just looking for nearby shops or restaurants might quickly help narrow down your search. These, however, are luxuries which we can afford thanks to our big brains.

Now consider the female mosquito that keeps following you around until it gets its fill of blood, while giving you diseases as a bonus. It can’t remember where humans cluster. Nor can it see well enough to distinguish humans from the background, especially at a distance. So how do these insects reliably follow you around when they can’t use the tricks we rely on?

Mosquitoes follow you around by tracking the carbon dioxide (CO₂) you exhale: a hard task as CO₂ comes in puffs, and there is no steady trail they can follow. Yet, they follow this patchy plume in three dimensions, all while flying around. Once they are close enough to smell your sweat (or skin odor), they shift gears and use that to narrow in. That combined with your body heat tells them that they have landed on a living blood bag, which leads them to pierce their syringe-like stylets into your body and have their fill.

Although scientists have known about mosquitoes tracking CO₂, skin odor and heat for a while, it is how these sensory cues are consistently pooled together to give rise to the tracking abilities of the mosquito. In this preprint, Sorrells et al. use Aedes aegypti as their model system (the mosquito that spreads a lot of deadly diseases), to investigate the effect of the first whiff of CO₂ on mosquito behavior. They show that just a 5-second exposure to CO₂ is sufficient to induce a state of heightened activity that lasts for up to 10 minutes, which facilitates a search for the source of CO₂ and triggers feeding once the insect finds the source (Figure 1).

Key findings

1. The authors genetically express light-activated ion channels in CO₂ sensing neurons (in the maxillary palp); this allows these CO₂ sensing neurons to be activated by shining red light. Using this approach (often called optogenetics), the authors can deliver a “fictive” CO₂ stimulus to the mosquito and measure its behavioral response. The optogenetic approach decouples the effect of airflow from CO₂ and allows precise temporal control of the odor delivered. Using this approach, the authors see a clear increase in overall activity of the mosquito after a fictive CO₂ stimulus (Video 1). In addition, if a feeder containing warm blood is placed on the cage containing mosquitoes, a fictive CO₂ stimulus is sufficient to get the mosquito to feed blood. Note that without the fictive CO₂ stimulus, the mosquito does not feed from the warm blood feeder – i.e., CO₂ is necessary to initiate feeding (also seen from previous studies). This shows that the response to the optogenetic stimulus sufficiently mimics the known response of mosquitos to CO₂.
2. Next, the authors dig into the behavioral effects of a 5-second pulse of fictive CO₂ To do this, they leverage machine learning tools that track the positions of different parts of the mosquito’s body and classifies the behavior it is performing. Using this method, they broadly characterize the activity of the mosquito into four types – flight, walking, probing (sticking its proboscis/stylet into the mesh cage) and none (sitting near motionless in the cage). When given a pulse of fictive CO₂, they find that mosquitoes begin flying, walking and probing, and sustain this activity for minutes (half-time of ~4 minutes; Video 1). In comparison, giving the mosquito a short heat stimulus only increases probing behavior, which quickly disappears (half-time of 0.4 minutes). This suggests that the CO₂ pulse causes a long-term change in the behavior of mosquitoes. Subsequent experiments show that this long-term sustained activity change only occurs in unfed female mosquitoes.
3. Because the long-term change induced by CO₂ is essentially a memory of the past whiff, the authors next ask how long this memory (or behavioral change) lasts. To do this, they use the result that CO₂ presentation is required to initiate feeding, even in the presence of warm blood. By staggering the CO₂ stimulus from the heat stimulus that warms a blood mimic, they determine that mosquitoes begin feeding up to 14 minutes after the CO₂

Why I think the work is important

The trick to following a noisy stimulus is to average it over a long period, so it makes sense to keep a memory of the last odor (CO₂) encounter. In fact, such long-term changes to odor stimuli have been seen in other insects, but it has been difficult to quantify this memory window.

The authors beautifully use the female mosquito’s feeding cascade to tease out the memory window. They find that a short 5-second stimulus causes a behavioral state change that lasts for 10-15 minutes. This is surprising, because the change in behavioral state is at least two orders more than the duration of the stimulus! Given that such long-term changes have been seen in other insects, this provides us with an upper bound on how long such state changes can last. The next step is to start investigating the mechanisms of such a state change. All these are reasons why I find the work exciting (and important).

Future directions / Questions to authors

1. It is surprising to see that fictive CO₂ did not often initiate flight. Why do you think that is the case? Is it because of the experimental setup (not enough room to fly)? Or maybe because there is no airflow coupled with CO₂ to trigger flight?
2. How do you think flight (airflow) will affect the duration of this long-term state? Do you expect it to change?
3. Given the temporal precision of optogenetic activation, why did you use a 5-second activation window instead of a much shorter one? A 10 or 100 millisecond pulse would have more accurately matched the timescale of a whiff of CO₂ in a turbulent plume. Did you try experiments with shorter pulses? If so, does the 10+ minutes of long-term change occur with short presentations of fictive odor too?
4. The mosquito system seems to provide a wonderful opportunity to perform comparative studies.

• • Wildtype male vs. ones lacking the fruitless gene could provide a way of finding out where this behavioral state change occurs.
  • Another comparative aspect would be to compare tracking flower odors with CO₂, i.e., do male mosquitoes have similar state changes when they encounter flower odor?

If possible, could you elaborate on your thoughts on these directions and maybe share your plans if you do intend to pursue them?

Tags: behavioral state, mosquito, multisensory integration, odor tracking, optogenetics, short-term memory

doi: https://doi.org/10.1242/prelights.31278

Read preprint (1 votes)