Remarkably, after the induction of anesthesia, the spontaneous activity of granule cells virtually disappeared (Figure 3B, bottom). Furthermore, in contrast to mitral cells, anesthesia strongly reduced odor responses of granule cells (Figures 3C and 3D). This resulted in individual granule cells responding to fewer odors with weaker responses under anesthesia (Figures 3E and 3F). Both ketamine and urethane

caused similar decreases of odor-evoked activity in granule cells (Figure S2). In the awake state, many granule cells responded near the onset of odor stimulation (Figure 3G, left) and the temporal dynamics of granule cell responses did not appear to be strongly modulated by brain states (Figure 3G). Thus, both spontaneous and MEK inhibitor cancer odor-evoked activity of granule cells are much weaker Luminespib in the anesthetized state, indicating that the activity of local inhibitory circuits in the olfactory bulb is strongly enhanced during wakefulness. Having identified major differences in olfactory bulb circuits in the awake and anesthetized state, we next asked how odor experience shapes mitral cell odor representations in awake animals. We imaged mitral cell responses to a panel of seven structurally diverse odors applied on 2 successive days (eight trials of each odor/day, 4 s/trial for this and all subsequent experiments). Because we used naive mice that had never been tested with odors, we considered all tested odors

as “novel.” On the first day of testing, each odor activated a large subset of the simultaneously imaged mitral cell population. However, responses of the same mitral cell population to the same odors

1 day later were significantly different (Figure 4A). Examining the response of each mitral cell to each odor, we found that while only 4% of odor-cell pairs showed a significant increase in response magnitude, 27% showed Histamine H2 receptor a significant decrease (n = 3 mice, 151 mitral cells, p < 0.01, permutation test, 10,000 repetitions for this and all following analyses unless otherwise mentioned). Consequently, the same odors activated significantly lower proportions of mitral cells on day 2 compared to day 1 (day 1: 27.3% ± 3.2% versus day 2: 21.1% ± 2.5%, mean ± SEM; p < 0.01, paired permutation test, 10,000 repetitions). Thus, the population responses to each odor became weaker after only 1 day of brief odor experience, indicating that odor experience has a lasting effect on the way mitral cells represent olfactory information. We next considered whether the weakening of odor representations induced by odor experience reflects a nonspecific decrease in mitral cell responsiveness or is specific to the experienced odors. To address this, we assessed the difference between responses to two sets of odors (A and B, randomly chosen for each mouse), where set A odors were experienced daily for 7 days, while set B odors were only encountered on the first and last days (Figure 4B, n = 5 mice, 212 mitral cells).