Salience encoding in the mouse olfactory system

Abstract

Internal representations of the external world guide organisms in executing appropriate behavior. The olfactory system in mice is a prime model of this process as mice rely heavily on their sense of smell to survive and determine whether to approach and investigate or avoid and flee. Otherwise known as valence, this is the intrinsic quality of the attractiveness or averseness of an odor and is arguably the first aspect of description organisms use to describe and categorize odors. Even before this categorization, these odorants must be defined as salient. The salience of a stimulus is defined as the intensity of a stimulus and serves to direct the attention of an organism to focus their perceptual and cognitive resources on the pertinent sensory stimulus. Furthermore, many odors trigger instinctive aversive or attractive responses, thus carrying an innate, unlearned value.While several higher-order olfactory regions have been implicated in processing innate valence of odorants, none have been shown to harbor an internal representation of innate odor valence. The central amygdala (CeA) has largely been studied in the context of emotion and fear processing, with studies demonstrating encoding of both innate and learned valence to various stimuli. One subpopulation of cells in the CeA, activated by general anesthetics (CeAGA), has been shown to be responsive to innately salient pain stimuli. To probe whether cells within the CeA encode innate valence of odorant stimuli, I recorded odor-evoked neural responses from populations of optogenetically identified CeAGA cells in awake, head-fixed mice and characterized their odor response properties. I found that CeAGA neurons respond to a variety of odorants and encode innate odor salience. Because CeAGA receives substantial input from a region called the amygdala-piriform transition zone (AmPir), which has been implicated in predator odor processing, I recorded the odor-evoked neural responses from populations of cells within this region. I found that AmPir shows odor response properties similar to piriform and does not harbor a representation of innate odor salience. Finally, the strongly valent predator odor 2MT induces robust defensive behaviors in mice. To probe the role of CeA cells in these defensive behaviors, I optogenetically manipulated their activity in the presence of innately salient predator odor and pain stimuli. I found no robust changes in defensive behaviors to predator odor and pain stimuli. Together, these studies reveal the first evidence of innate odor categorization in the mouse through innate odor salience.