Scientists can Now See Sense of Smell
DURHAM, N.C. -- Using a high-resolution video technique on
laboratory rats, neurobiologists at Duke University Medical
Center have captured the first detailed images of the living
brain in the act of recognizing specific odor molecules. The
scientists say their achievement will open the way to
deciphering the brain's internal "language" of smell.
More broadly, say the researchers, the imaging technique can
give them new insights into the machinery of learning, as they
explore how training alters the odor-recognition process.
The researchers -- graduate student Benjamin Rubin and
Howard Hughes Medical Institute Investigator Lawrence Katz --
reported their achievement in the July issue of Neuron. To record
odor-recognition, the researchers refined an imaging technique
used to visualize brain activity in other parts of the brain.
They increased its resolution tenfold to detect changes in the
tiny hair-thin olfactory structures, called glomeruli, in rats'
They began their studies by thinning out a section of a
rat's skull until they could see the olfactory bulbs --
stem-like projections in the forebrain that receive signals
from the chemical receptor cells that line the nasal
Rubin's and Katz's objective was to visually detect when a
specific odorant triggered the activity of a specific
glomerulus -- tiny basketlike structures covering the surfaces
of the olfactory bulb. These glomeruli, each about the diameter
of a human hair, are the basic units of olfactory reception.
Previous research has shown that each of the 2,000 glomeruli in
the olfactory bulbs receives impulses from nasal receptors
tuned to specific odorants, relaying those signals to higher
processing centers in the brain.
The scientists' detection method depended on the fact that
active cells consume more oxygen, converting oxygen-carrying
oxyhemoglobin to deoxyhemoglobin. Since deoxyhemoglobin absorbs
red light more than the oxygenated form, the scientists could
distinguish activated glomeruli by imaging the olfactory bulbs
under a red light shone through the rat's skull. Activated
glomeruli showed up as distinctive dark spots on the
To map how glomeruli responded to odors, the scientists
recorded images of the olfactory bulbs as they exposed the rats
to chemical odorants that smelled like bananas, caraway and
spearmint, as well as the mixed-chemical smell of peanut
"We found that we could visualize individual glomeruli,
achieving the best resolution anyone has obtained so far," said
Katz. "And most exciting, we found that we could see
distinctive patterns of activated areas from different
What's more, said Katz, the researchers found that
activation patterns in olfactory bulbs on one side of an
animal's brain matched those on the other. Also, the patterns
of activation were highly similar from animal to animal.
"Thus, this technique can serve as our guide, our Rosetta
stone, for deciphering the olfactory response and even helping
us to understand higher olfactory processing areas of the
brain," said Katz.
For instance, Katz said, he and Rubin also reported
experiments in which they varied the concentration and
molecular structure of odorants and observed how the activation
pattern of glomeruli changed. They found that the activation
maps changed significantly as they increased the concentration
of the odorant chemical amyl acetate several thousandfold.
Also, the activation maps changed as they exposed the rats to a
series of slightly different aromatic chemicals called
aldehydes. These aldehydes differed only in the number of
carbons in their chainlike structure.
"Previously, you simply couldn't compare such responses to
multiple concentrations and multiple chemicals in the same
animal, because the only technique available was to use tracers
that could only detect response to a single odorant," said
Katz. "Also, that technique required killing the animal and
doing detailed analyses that took weeks."
Rubin's and Katz's experiments already have settled a
critical question about the nature of olfactory processing in
"There has been considerable debate in the field about
whether closely related odorants could be distinguished by
spatial mapping, or whether there was some timing of neuronal
firing involved. However, we have found that we could
distinguish odorants just on the basis of the pattern of
activated glomeruli," Katz said.
Katz believes the olfactory visualization system offers high
promise in studying the machinery of the learning process.
"In rodents, the olfactory system is the sensory system of
choice, and we believe we can see the early stages of learning
at the olfactory bulb level. Since our system is very rapid and
non-invasive, we think it offers an extraordinary pathway to
Such learning experiments could involve imaging changes in
animals' response to odorants as they are trained to associate
an odorant with a reward, said Katz. Also, he said, by studying
genetically altered mice with altered olfactory systems,
scientists could gain important clues to the molecular basis of
olfaction in particular, and sensory processing in general.