Neural Patterns Show Odor Detection Is Like Picking Up Voices At A Noisy Cocktail Party
A team of scientists led by Professor Venkatesh Murthy at Harvard Medical School’s Department of Molecular and Cellular Biology, found that odor molecules activate specific patterns of neurons, and that the ability of a mouse to distinguish between a learned, target odor and random, background odors, depends on the strength of neural pattern overlap between the two odors. Dr. Dan Rokni was the primary author.
“There is a continuous stream of information constantly arriving at our senses, coming from many different sources,” Professor Murthy said. “The classic example would be a cocktail party — though it may be noisy, and there may be many people talking, we are able to focus our attention on one person, while ignoring the background noise.”
The researchers first trained the animal to detect particular odors. Following, they presented the animals with mixtures of smells, sometimes and sometimes not including the trained, “target” scent. When the mouse licked a waterspout after being presented with a mixture containing the target scent, they were rewarded with a drop of water. When the mouse performed incorrectly by either licking the waterspout when the mixture did not contain a target, or not licking the waterspout when the target was present, it was “punished” with a 5 second pause. When the mouse correctly rejected the mixture without a target, there was neither reward nor punishment.
Overall, the mice seem to perform quite well, picking out the “target” with 85% accuracy. In fact, when the mixture has only one other scent, the accuracy was very high at 94%. This diminished as more scents were added until it bottomed out at 80% for a mixture of 14 scents. But observing animal behavior alone does not reveal its fundamental neural basis. It was still unclear how the mice were able to pick out target odors at all, and moreover, why was the average success only 85%.
To understand the molecular basis of these observations, Professor Murthy’s team engineered into olfactory bulb of the mice a fluorescent protein responsive to calcium ion flux. The olfactory bulb is the part of the brain that receives neural inputs from the sensory neurons in the nose. Calcium influx is an important player in neural synaptic transmission. Neurons carry signals from one to another through waves of calcium flux into the cells. Therefore, for each odorant molecule, the researchers could visualize the pattern of neurons that lit up during the test and recall phase. Only a patch of the olfactory bulb was imaged.
Each of the 14 tested odors elicited a different spatial pattern of firing neurons. In the cases where the mice did poor in detecting a target odor against background, it could be explained by the fact that the spatial patterns of the target overlap strongly with the background mixture.
The researchers took the analysis one step further by exploring whether the mice were getting confused because the mixture was very similar to the target, or whether the mixture was masking the target. The incentive structure of the experiment was deemed such that the mouse behaves the same because one could not tell whether the mouse was licking because it thought both target and mixture smelled the same, or whether it was licking because it could tell there was so many odors that the target must be there even though it could not detect it.
Based on the input odor composition and strength, they computed “similarity” and “masking” scores for the mixture and target. They then computed whether performance was more due to similarity or to masking. The conclusion was that masking was contributing to degradation of performance. Basically the mouse was commiting errors because when it could no longer detect the target due to the other distracting odors.
Professor Murthy says in conclusion, “This study is interesting because it first shows that smells are not always perceived as one whole object — they can be broken down into their pieces,” he added. “This is perhaps not a surprise — there are in fact coffee or wine specialists that can detect faint whiffs of particular elements within the complex mixture of flavors in each coffee or wine. But by doing these studies in mice, we can now get a better understanding of how the brain does this.”
This research was published on Aug 3, 2014 in Nature Neuroscience.