Circuits of excitation and inhibition help the brain interpret odors.

Curry-infused Thai food, a gas leak, a dirty cat box. Through sensory cells in the nose, the brain makes quick sense of the good and the bad in our olfactory universe.

The initial phase of the sense of smell is well understood. So-called olfactory receptors—dogs have around 1,000 types of receptors, humans have hundreds, fruit flies around 50—sit on neurons in a vertebrate’s nose or on a fly’s antenna. Each neuron carries one type of receptor, and all neurons with the same type converge at an arborlike structure called a glomerulus and pass their information to other neurons.

All glomeruli reside in the same brain region: the antennal lobe in fruit flies and the olfactory bulb in vertebrates. Different odors activate different combinations of receptors, producing a kind of olfactory code, with each olfactory receptor acting as a chemical “channel” that conveys a stream of information like a radio channel.

Less clear is exactly what happens next. How do neural networks make sense of these activation patterns, direct the information to different parts of the brain, and alter an organism’s behavior in response? “The key problem in the field has become figuring out how information is processed across these channels,” says HHMI early career scientist Rachel Wilson at Harvard University, who, along with Liqun Luo of Stanford University, has begun to figure out how the brain processes smell information and routes it to different brain regions.

Luo, an HHMI investigator, is an expert at mapping how cells course through the nervous system as an animal develops to adulthood. He also has investigated where olfactory information goes after it leaves a glomerulus. In research reported online in Nature on December 22, 2010, Luo and his team mapped the connections from glomeruli in the olfactory bulb to different areas of the mouse brain.

The group found that the amygdala, a brain region responsible for emotion, receives inputs from glomeruli predominantly in one part of the olfactory bulb. Previously, scientists found that knocking out this part of the olfactory bulb makes a mouse no longer react with fear to the smell of a predator, leading Luo to speculate that the amygdala uses odor information to control the innate fear response. By contrast, the piriform cortex, a major area for smell processing, received a broad, unorganized set of inputs from glomeruli, suggesting that it calculates an overall view of the olfactory environment.

In a related study, HHMI investigator Richard Axel of Columbia University and colleagues injected dye into specific glomeruli to trace neurons to the piriform cortex and the amygdala. Squaring with the findings from Luo’s group, the researchers found that individual glomeruli send neurons to all parts of the piriform cortex, whereas they send neurons to more defined regions of the amygdala. Those findings appeared online on March 30, 2011, in Nature.

Wilson is exploring how signals are adjusted and combined at the glomeruli. Each glomerulus is a way station where long-distance neurons meet, but the glomeruli are also connected to each other by local neurons, called interneurons. Typically, these interneurons dampen the signals sent by other neurons, so scientists have reasoned that they must be involved in synthesizing and analyzing all the information coming through the different olfactory channels. But exactly what they are doing has been mysterious. “What’s going on with the local interneurons?” wondered Wilson. “Why do you need so many?”

The two investigators tackled these questions together by matching Luo’s expertise in deciphering wiring maps with Wilson’s skill at measuring neuron activity by recording electrical impulses. In research reported in Nature Neuroscience in April 2010, they mapped interneuron patterns and activity in the fruit fly brain.

Illustration: VSA Partners

PAGE 1 2		Continue