Skip to main content
Text Size: sML

"I'll Have What She's Having!"

New NIDCD research offers intriguing clues into the role of smell in food preference

Two mice on top of cheese
Two mice exploring their lunch option

If two mice run into each other after one of them has just eaten, an interesting thing happens: the mouse with an empty stomach will subsequently choose the food that his buddy has snacked on over any other food source. For decades, researchers have wondered about the origins of this behavior, which they call socially transmitted food preference (STFP). They believed that it had to be nature's way of stacking the deck against eating foods that could be harmful or poisonous. It allowed rodents to lessen the risk of eating unsafe foods by basing their food preferences on the successful eating experiences of their peers.

In earlier studies, researchers had established that it was the odor of carbon disulfide (a byproduct of food metabolism) on the breath of the snacker that acted as a chemical social cue. When combined with the scent of bits of food still clinging to her fur and whiskers, it was enough to prompt a nearby hungry mouse to think "Say, I'll have what she's having!" But in spite of many years of experimentation, no one had been able to explain on a molecular level how carbon disulfide combined with the food odor to create a signal in the brain that says "okay to eat."

Now, a recent paper published online in the journal Current Biology shows that the behavior is the result of a dedicated subsystem of specialized olfactory receptors in the nose and neural circuits in the brain whose job it is to associate an incoming odor with food that's safe to eat. The NIDCD-funded study was conducted by Steven Munger, Ph.D., at the University Of Maryland School of Medicine in Baltimore, Md., and an international team of researchers.

"Scientists are beginning to see the olfactory system not as a solitary unit, but as a collection of subsystems, some of them well known and some of them only recently described," says Munger. Each of these subsystems is distinguished by the location of its sensory neurons in the nasal cavity, the kinds of receptors and signaling mechanisms it uses, the region of the brain's olfactory bulb it signals, and likely the ultimate destination of those signals in the brain's olfactory cortex. This helps explain why the role of smell is so broad among mammals. In addition to helping an animal identify food and gauge its quality, a single sniff can alert it to another animal's reproductive status or warn it about a nearby predator, a potential poison, and even the presence of disease. Much of this olfactory communication between mammals takes place through the use of social cue chemicals, such as pheromones, which cause a behavioral or hormonal change in the sniffing animal.

Munger and his colleagues thought the most likely place to look for the STFP connection was a small family of olfactory receptors called the GC-D+ neurons. These receptors send signals to special clusters of other neurons called the necklace glomeruli, so named because they encircle the back end of the olfactory bulb like a string of beads. Glomeruli, in general, are important way stations in the path from the nose to the brain. It's here that information from the olfactory receptors is initially translated into the signals that travel onward to the olfactory cortex.

According to Munger, the necklace glomeruli are wired differently than those in the rest of the olfactory system. Unlike most glomeruli, which are usually only stimulated by signals from one type of olfactory receptor, the necklace glomeruli receive input from a number of different receptors as well as other glomeruli in the olfactory bulb. Since they seemed to be set up to integrate multiple sensory inputs, he and his colleagues thought that perhaps this was a place where social cue chemicals and general odors, such as those from food, could be processed together. They and others had already found substances that activated the GC-D+ neurons, including chemicals in urine (a rich source of social cues for mice), and carbon dioxide, a cousin of carbon disulfide.

To test their theory, Munger and his colleagues exposed the GC-D+ neurons to carbon disulfide to show that they did, indeed, respond to the chemical. Next, they used knockout mice specifically engineered to turn off the three proteins that GC-D+ neurons require to sense olfactory cues and noted that the olfactory response to carbon disulfide was drastically reduced. In addition, the knockout mice displayed no preference for the food odor, in this case cinnamon, which they were exposed to on a nearby mouse at the same time as the carbon disulfide. Normal mice, on the other hand, chose cinnamon-scented food over any other flavor if they had earlier sniffed cinnamon on a buddy.

"This is a way of detecting stimuli with a dedicated circuit," says Munger. "That's what these GC-D+ neurons and necklace glomuruli are doing when they link up to the olfactory brain." This direct link is likely critical to surviving in the wild where rodents have to make quick decisions about whether or not a food is safe to eat.

In most primates, including humans, a key gene that controls carbon disulfide detection isn't functional. Since humans have words and visual cues for food, while mice only have their noses in a nocturnal world, we might have outgrown the need for such a dedicated circuit. But it still offers clues about how we learn to associate behavior with odor.

"We may still have some variation of it," says Munger. "Odors alone can elicit all kinds of feelings about food. You smell the baking cookies and your stomach growls—that's coupling food detection with memory systems and internal sensations. Those types of things are still there."