Active Sampling State Dynamically Enhances Olfactory Bulb Odor Representation

• DOI: 10.1016/j.neuron.2018.05.016
• Publication Date: 2018-05-31T23:39:23Z
• Authors: Rebecca Jordan; Izumi Fukunaga; Mihaly Kollo; Andreas T. Schaefer
• Keywords: 0301 basic medicine; 0303 health sciences; learning; sniffing; Patch-Clamp Techniques; Time Factors; Behavior, Animal; behavior; Olfactory Pathways; Olfactory Bulb; Article; context; Discrimination Learning; Mice; 03 medical and health sciences; olfactory bulb; Odorants; Animals; Learning; active sampling; olfaction

SummaryThe olfactory bulb (OB) is the very first site of odor information processing, yet a wealth of contextual and learned information has been described in its activity. To investigate the mechanistic basis of contextual modulation, we use whole-cell recordings to measure odor responses across rapid (<30 min) learning episodes in identified mitral/tufted cells (MTCs). Across these learning episodes, we found that diverse response changes occur already during the first sniff cycle. Motivated mice develop active sniffing strategies across learning, and it is this change of active sampling state that dynamically modulates odor responses, resulting in enhanced discriminability and detectability of odor representation with learning. Evoking fast sniffing in different behavioral states demonstrates that response changes during active sampling exceed those predicted from purely feed-forward input. Finally, response changes are highly correlated in tufted cells, but not mitral cells, indicating cell-type specificity in the effect of active sampling, and resulting in increased odor detectability in the tufted and enhanced discriminability in the mitral cell population over the rapid learning episodes. Altogether, we show that active sampling state is a crucial component in modulating and enhancing olfactory bulb responsiveness on rapid timescales.