A growing body of scientific literature indicates that sensory processing in the nervous system incorporates not just "bottom-up" analysis of sensory information that originates from the sensory organs but also "top-down" expectations about the organization of the sensory world that derive from previous sensory experience. In the olfactory system the first convergence of bottom-up and top-down projections occurs in the glomeruli of the olfactory bulb, where optical imaging techniques now permit the visualization of presynaptic calcium signaling and neurotransmitter release from olfactory receptor neurons (the very first neurons in the olfactory system) in awake mice that are smelling odors and learning about their environment. The olfactory system is thus a uniquely powerful model system to study the synthesis of top-down and bottom-up information. Remarkably, in preliminary experiments we have found that this primary sensory input is strongly modulated by the mouse's expectations about olfactory stimuli, as established by prior sensory experience during the imaging session. For instance, if an odor is always presented after a "warning tone" cue, the unexpected presentation of that odor without the tone evokes much less presynaptic calcium influx and less neurotransmitter release from the olfactory nerve than when the odor follows the cue. This effect appears to occur via a GABAB receptor-mediated presynaptic inhibition of neurotransmitter release from these synapses. Because the output of the receptor neurons themselves is modulated by expectations, these data suggest that there is actually no purely bottom-up information in the olfactory system at all. This has profound implications for our understanding of neural representations of olfactory stimuli and will inform our understanding of other, less experimentally tractable sensory systems. This proposal confirms and extends these findings by testing the nature of the expectations (e.g. is the expectation odor-specific?), their time course (e.g. are expectation effects anticipatory?), and their neural mechanisms (e.g. are descending projections to the olfactory bulb necessary during the establishment of the expectation or only for detecting unexpected outcomes?). Further experimentation is designed to reveal the perceptual consequences of these neural changes (e.g. do expected odors smell stronger than unexpected odors?) and whether the difference in neurotransmitter release from receptor neurons is necessary for these perceptual effects to occur. Importantly, this work will be designed and analyzed in the context of information theory, a formal theory that allows the quantification of how expected or surprising a given stimulus is and thus provides testable quantitative predictions for assessing the neurophysiological and psychophysical consequences of expectation. These results should provide an essential step in understanding how cognition can play a role in perception as early as the first neurons in a sensory system.

Public Health Relevance

Sensory disorders affect 11.6 million Americans (2005 Census), but the development of effective therapeutics and prosthetic sensors to help patients with these disorders requires a better understanding of the basic neural mechanisms of sensation and perception. A variety of recent research indicates that the brain learns about its sensory environment and uses this information to develop expectations that shape its interpretation of subsequent sensory input. This project explores the role of these expectations in shaping the responses of the very first neurons in the olfactory system, which will provide important new insight into basic sensory function and inform the development of clinical treatments.

National Institute of Health (NIH)
National Institute on Deafness and Other Communication Disorders (NIDCD)
Research Project (R01)
Project #
Application #
Study Section
Somatosensory and Chemosensory Systems Study Section (SCS)
Program Officer
Sullivan, Susan L
Project Start
Project End
Budget Start
Budget End
Support Year
Fiscal Year
Total Cost
Indirect Cost
Rutgers University
Schools of Arts and Sciences
New Brunswick
United States
Zip Code
Kass, Marley D; Rosenthal, Michelle C; Pottackal, Joseph et al. (2013) Fear learning enhances neural responses to threat-predictive sensory stimuli. Science 342:1389-92