Odor recognition is a patterned, time-dependent process, research shows
PASADENA, Calif.-When Hamlet told the courtiers they would eventually "nose out" the hidden corpse of Polonius, he was perhaps a better neurobiologist than he realized. According to research by neuroscientists at the California Institute of Technology, the brain creates and uses subtle temporal codes to identify odors.
This research shows that the signals carried by certain neuron populations change over the duration of a sniff such that one first gets a general notion of the type of odor. Then, the wiring between these neurons performs work that leads to a more subtle discrimination, and thus, a precise recognition of the smell.
In the February 2 issue of the journal Science, Caltech biology and computation and neural systems professor Gilles Laurent and his colleague, postdoctoral scholar Rainer W. Friedrich, now at the Max Planck Institute in Heidelberg, Germany, report that the neurons of the olfactory bulb respond to an odor through a complicated process that evolves over a brief period of time. These neurons, called mitral cells because they resemble miters, thepointed hats worn by bishops, are found by the thousands in the olfactory bulb of humans.
"We're interested in how ensembles of neurons encode sensory information," explains Laurent, lead author of the study. "So we're less interested in where the relevant neurons lie, as revealed by brain mapping studies, than in the patterns of firing these neurons produce and in figuring out from these patterns how recognition, or decoding, works."
The researchers chose to use zebrafish in the study because these animals have comparatively few mitral cells and because much is already known about the types of odors that are behaviorally relevant to them. The Science study likely applies to other animals, including humans, because the olfactory systems of most living creatures appear to follow the same basic principles.
After placing electrodes in the brain of individual fish, the researchers subjected them sequentially to 16 amino-acid odors. Amino acids, the components of proteins, are found in the foods these fish normally go after in their natural environments.
By analyzing the signals produced by a population of mitral cells in response to each one of these odors, the researchers found that the information they could extract about the stimulus became more precise as time went by. The finding was surprising because the signals extracted from the neurons located upstream of the mitral cells, the receptors, showed no such temporal evolution.
"It looks as if the brain actively transforms static patterns into dynamic ones and in so doing, manages to amplify the subtle differences that are hard to perceive between static patterns," Laurent says.
"Music may provide a useful analogy. Imagine that the olfactory system is a chain of choruses-a receptor chorus, feeding onto a mitral-cell chorus and so on-and that each odor causes the receptor chorus to produce a chord.
"Two similar odors evoke two very similar chords from this chorus, making discrimination difficult to a listener," Laurent says. "What the mitral-cell chorus does is to transform each chord it hears into a musical phrase, in such a way that the difference between these phrases becomes greater over time. In this way, odors that, in this analogy, sounded alike, can progressively become more unique and more easily identified."
Applied to our own experience, this result could be described as follows: When we detect a citrus smell in a garden, for example, the odor is first conveyed by the receptors and the mitral cells. The initial firing of the cells allows for little more than the generic detection of the citrus nature of the smell.
Within a few tenths of a second, however, this initial activity causes new mitral cells to be recruited, leading the pattern of activity to change rapidly and become more unique. This quickly allows us to determine whether the citrus smell is actually a lemon or an orange.
However, the individual tuning of the mitral cells first stimulated by the citrus odor do not themselves become more specific. Instead, the manner in which the firing patterns unfold through the lateral circuitry of the olfactory bulb is ultimately responsible for the fine discrimination of the odor.
"Hence, as the system evolves, it loses information about the class of odors, but becomes able to convey information about precise identity," says Laurent. This study furthers progress toward understanding the logic of the olfactory coding.