The smell of fear

NEARLY 70 years ago, Karl von Frisch described the alarm response in a species of small freshwater fish called the European minnow (Phoxinus phoxinus). Frisch, who was one of the founders ethology - the scientific study of animal behaviour - demonstrated that when a minnow was eaten by a predator, a chemical released from its damaged skin elicited defensive behaviour in other minnows that were close by. In response to the chemical, they would at first dart about randomly, form a tight school and then retreat from the source of the chemical.

Frisch called this substance schreckstoff, meaning "scary stuff", and we now know that similar chemicals are used throughout the animal and plant kingdoms. When faced with a threatening situation, many organisms, including, insects, fish and mammals, release volatile small chemicals alarm pheromones which signal the danger to other members of the same species. However, neither the pheromones, nor the sensory systems and mechanisms that detect them, have been identified. But now a team of researchers from the University of Lausanne in Switzerland show that mice detect alarm pheromones by means of a recently identified sensory system in the nose. They report their findings in the current issue of the journal Science.

Julien Brechbuhl and his colleagues examined a structure called the Grueneberg ganglion (GG), which in mammals is located at both sides in the tip of the nose, close to the openings of the nostrils. The GG is an arrow-shaped structure which in mice is approximately 1mm in length. It consists of 300-500 neurons arranged in several densely packed grape-like clusters found in discrete pockets of the septum which separates the nostrils. The axons of these cells bundle together very close to the cell bodies and project backwards along the roof of the nasal cavity to the olfactory bulb, which relays signals related to the sense of smell to the brain.

When the GG was first discovered in 1973, its anatomy was not known in such detail and so it was thought to be a non-sensory structure. The connections to the olfactory bulb were first described just 3 years ago. Thus, it is only very recently that the olfactory system has come to be viewed as containing 3 distinct "channels", each with a unique structure and, probably, function. The main channel is involved in detecting aromatic molecules; the second channel is called the vomeronasal system, and is an "accessory" olfactory system which is now known to be involved in the detection of pheromones; the GG constitutes a third component of the olfactory pathway, one that was thought to be involved in mother-pup recognition and suckling behaviour, because it is present at the time of birth.

In the new study, Brechbuhl and his colleagues used a combination of techniques to better characterize the structure and function of the GG in mice. First, scanning electron microscopy showed that each GG neuron possess between 4-8 hair-like projections called ciliary processes. However, when they used another method of electron microscopy, they got conflicting results: each cell appeared to have between 30-40 of these processes. Further investigation revealed that the GG also contains a population of glial cells which trap most of the ciliary processes within the cell cluster.

Grueneberg_ganglion_cells.bmp
Scanning electron microscope images of the mouse Grueneberg ganglion. Left: a cluster of neurons (GC) in a meshwork of fibroblasts (Fb) Right: and a higher magnification of a single GG neuron (green), with its axon (red) and thin ciliary process (blue). Scale bars: 20 microns (L) and 5 microns (R).  From Brechbuhl et al (2008).

 

Antibody staining showed that the GG cells express genes related to those that encode olfactory receptors and, although the ciliary processes do not project into the nasal cavity, the layer of epithelial cells surrounding the GG cell clusters was found to be permeable to substances that are soluble in water. The cilia are therefore likely sites of transduction, the process by which a sensory stimulus is converted into a pattern of nervous impulses, and the finding that the epithelium was permeable to water-soluble substances suggests a mechanism whereby such substances can gain access to the GG neurons.

To test for such a mechanism, the researchers took slices of tissue from the noses of their mice, and injected the GG cells with a calcium dye. They then exposed the tissue slices to various chemical stimuli. The dye injected into the cells is one that emits a fluorescence in response to elevated levels of calcium inside the cell, something that is characteristic of neuronal activity. Thus, if the GG neurons bind any of the chemicals, they become active, and the fluorescence can be detected.

In these experiments, no calcium increases were observed in response to either to either mouse milk or secretions from the mammary glands of lactating mice. Mixtures of odours, mouse urine and known mouse pheromones did not elicit increases in fluorescence either. But when the cells were exposed to alarm pheromones collected during the killing of mice by carbon dioxide asphyxiation, they exhibited a robust and reversible increase in fluorescence.

Finally, the researchers sought to investigate the role of the GG in behaviour. Because of its location, the GG is easliy accessible, so they were able to cut the axons of GG neurons in live mice (a process called axotomy), thus preventing any signals from reaching the brain. 30 days later, the responses of the animals to alarm pheromones were tested. Typically, mice respond to these chemicals by freezing in their tracks. But when the axotomized mice were exposed to the alarm pheromones, they did not freeze, but instead continued to explore their environment.

This study clearly shows that in mice the GG is involved in detecting alarm pheromones, rather than in mother-pup interactions, as was previously thought. The GG has a specialized and very basic structure which enables it to perform this primordial function - it consists of a small cluster of cells that is separated from the outside world by nothing more than a water-permeable sheet of epithelial cells. The location of the GG, far away from the main olfactory system, allows for the rapid detection of alarm pheromones. Such a mechanism is of course crucial to the survival of the organism, and the GG is found in every mammalian species examined so far, including humans.

Related:


Brechbuhl, J., et al (2008). Grueneberg ganglion cells mediate alarm pheromone detection in mice. Science 321: 1092-1095. DOI: 10.1126/science.1160770

More like this

I have heard that when cats and dogs at the pound are to be killed ('euthanized'), they seem to sense immediate mortal danger. If the dying give off some pheromone, then this 'animal sense' makes good sense.

I was going to say that these researchers must have some great confocal apparatus, but then I looked at the original paper. Those pictures you've posted here are from a scanning electron microscope, not confocal imaging.

So THAT explains the name of the animated ogre -- "Frisch called this substance schreckstoff, meaning "scary stuff"

Now, what do we know, as a unified theory, of pheromones in human beings?

Is there a basis for that Broadway show title: "The roar of the greasepaint, the smell of the crowd"?

Can humans smell fear?

And can you help me out in my ongoing argument with a physicist, who denies my saying that rats have been proven to be able to smell carbon dioxide?

Do humans have this GG structure?

By Luna_the_cat (not verified) on 25 Aug 2008 #permalink

Jonathan Vos Post:
There is some evidence that humans *can* smell fear.
http://www.ncbi.nlm.nih.gov/pubmed/12011790?dopt=Citation

"The scent of fear." Ackerl K, Atzmueller M, Grammer K.
Neuro Endocrinol Lett. 2002 Apr;23(2):79-84.

I'm pretty sure there is something more recent than that, too.

Speaking just for myself, *I* can smell it; I notice a distinct change in my own odor when I am either afraid or very stressed, and can tell from my husband's clothes if he has had "that kind of day" as well. FWIW, I have noticed in the past that animals which have been habitually maltreated or which have become chronically fearful for one reason or another "stink" in a way which happy animals do not, and this holds for rodents, rabbits, dogs, cats, and horses.

By Luna_the_cat (not verified) on 25 Aug 2008 #permalink

Do humans have this GG structure?

Did you read the whole post? Look at the last sentence.

At the time I finished my Ph.D. (1965), it was thought that fright substance was confined to minnows. One of my contemporaries got his Ph.D. by showing that it occurred in other fish groups. I can recall a couple of similar studies in that era. Wikipedia, and most of the references which come up with Google, treat it as a more or less restricted, minnow, ostariophysan, fish, aquatic organisms, whatever, phenomenon. Your post here makes the case that it is a broad and general phenomenon. Very interesting.

By Jim Thomerson (not verified) on 26 Aug 2008 #permalink