The scent of a kitten: Fear may smell like cats, if you're a mouse

May 22 2013 Published by under Behavioral Neuro, Uncategorized

What does terror smell like? Well, if you're a mouse, terror smells like something that's going to eat you. Maybe a cat, a fox, or a large bird. As prey animals, mice need all the help they can get in avoiding potential predators. And they get a lot of help from smell. Most predators produce pheromones that mice can sense, and these are often the only warning the mouse has that it's about to become dinner.

But not all dangers produce pheromones, and mice still have to let each other know that something is coming. So, as these authors show, mice may produce pheromones of their own that can communicate alarm to other mice.

If you need to keep the mice away, be prepared to wear...ear d'terror.

Brechbuhl et al. "Mouse alarm pheromone shares structural similarity
with predator scents" PNAS, 2013.


(Or get a cat. That works too. Behold, Scicat).

The authors of this paper had previously described a new ganglion, a group of neurons, which they believed to be involved in pheromone sensing, particularly of alarms. They had called this the Grueneberg ganglion, a small group of cells located at the VERY tip of the mouse nose. The neurons in this area specifically responded to alarm pheromones in previous studies.

But those were alarm pheromones from predator species. What about alarm pheromones from mice themselves? Do mice produce alarm pheromones to warn other mice of danger, and if they do, what are they?

To figure this out, the authors exposed a group of mice to carbon dioxide, which causes a major stress and alarm reaction. They then microextracted the chemicals the animals produced, and identified them. They found a total of 44 chemicals, 32 of which were emitted by control, non-alarmed mice, so those could be discounted. That left 8. 3 of these were unidentified, but the other 5 were gathered together an tested, to see what effects they might have.

First, they exposed the Grueneberg ganglion, where they believed the alarm pheromones to have the biggest effect, to the different chemicals they had isolated. In theory, if the any of the particular chemicals were an alarm pheromone, they would produce chemical potentials in the Grueneberg ganglion, which could be detected my imaging for calcium responses.

While most of the chemicals didn't do anything to the Grueneberg ganglion, one did, SBT (2-sec-butyl-4,5-dihydrothiazole). This one produced a nice big calcium response in the Grueneberg ganglion.

(Figure 2B)

You can see the calcium response on the left. But this is just in mouse slices of ganglion. If you want to prove something is a predator indicator, we need to see some fear.

SBT as a chemical is similar in structure to regular predator chemicals, but does it act the same way? The authors compared SBT with two known predator phermones, TMT and 2-PT, and looked for behavioral responses. They saw a significant increase in the stress hormone corticosterone in response to SBT, and also saw good marks of fear behaviors in the mice exposed to SBT, TMT, and 2-PT, including freezing (a well known fear response in mice), and risk assessment (when mice stretch out to their full length to investigate, leaving the hind paws in relative safety so they can spring back if there's danger).

But what this all working through the Grueneberg ganglion? After all, there's another very well known pheromone detection system: the vomeronasal organ. The authors of this paper believe that the Grueneberg ganglion plays an even bigger role than the VNO when it comes to fear responses.

You can see above the percent of responding cells in each organ (at left), including the Grueneberg ganglion, the Main Olfactory Epithelium (a major scent area), and the VNO. And it looks like SBT, TMT, and 2-PT, all the potential predatory indicators, produced the biggest effect in the Grueneberg ganglion.

So it appears the Grueneberg ganglion responds in a big way to these chemicals. But is it required for the fear response? The authors took a group of mice and looked for fear responses again in response to the pheromones. But this time, half of the mice had had the axons coming from the Grueneberg ganglion CUT. The Grueneberg ganglion was still there, but it couldn't send messages on anymore.

What you can see above are the fear responses in the control mice (grey), and the mice with the cut axons (black). Without the Grueneberg ganglion signals, the mice showed no significant freezing responses to SBT, TMT, or 2-PT, suggesting that the Grueneberg ganglion is required for this fear response. But it's not ENTIRELY required, as the mice with cut axons still showed increased risk assessment (bottom section).

So it appears that the Grueneberg ganglion plays a big role in fear in mice, and that SBT is an important pheromone for mouse to mouse communication. SBT is released in the urine and could be used to indicate to other mice that something wicked this way comes (and the urine is bound to happen, as anyone who has ever handled mice will tell you, one of the first things they will do is pee on you).

The authors note that SBT is structurally similar to TMT and 2-PT, both of which are natural predator odors, and so SBT might be activating similar receptors, giving the mouse to mouse version of the smell of terror. What are those receptors? In theory, it would be the TAAR4 receptor, which is known to mediate responses to predator odor in the VNO. But we don't know yet whether the mouse pheromone works the same way.

But there are still further questions. When they were causing mice to produce alarm pheromones, CO2 was the alarm condition they used (they also tried restraint and cold temperature). It produces a major alarm response, sure, but is it the SAME alarm response to the presence of a predator? It would be interesting to see how the chemical components compared to, say when the animals were exposed to real cat or fox odor. I understand the limitations of using the odors (the responses aren't as reliable, and getting fox odor out of anything is...almost impossible), but they did use some of them for other experiments, and it would definitely be the best way to determine the true similarity (though the chemical similarity and the behavior they got is a good indicator).

This is especially important because SBT is more than just a potential alarm cue. When presented to male and female mice, it promotes aggression in males, while it females it synchronizes estrous. Both of these things (but especially the aggression) can produce the increases in corticosterone they saw in the animals after SBT exposure. So it would be interesting to see both more specifically how SBT is acting, and also how SBT responses compare to true predator odor responses.

Additionally, they still had three unlisted chemicals. I really wonder what those are, though I imagine they are probably being examined separately. It would be interesting if they also played a role.

But it's an interesting thing to think about. Could pheromones like this be used as, say, a good way to get rid of rodents (though, since most of them REALLY smell, even to humans, it's probably not the best idea). And are there other roles for SBT that might be more than just eau d'terror? Future studies will have to find out.

Brechbuhl, J., Moine, F., Klaey, M., Nenniger-Tosato, M., Hurni, N., Sporkert, F., Giroud, C., & Broillet, M. (2013). Mouse alarm pheromone shares structural similarity with predator scents Proceedings of the National Academy of Sciences, 110 (12), 4762-4767 DOI: 10.1073/pnas.1214249110

3 responses so far

Leave a Reply to The Scent of a Feline Cancel reply