Remarkably, mitral cell responses under anesthesia on day 7 were indistinguishable from those observed during anesthesia on day 1; thus, the selleck expression of the plasticity induced during wakefulness was blocked when tested under anesthesia (Figure 7B). These results indicate

that the expression of the experience-dependent plasticity of mitral cell responses depends critically on wakefulness. Although previous studies reported that odor-evoked mitral cell activity is enhanced under anesthesia (Adrian, 1950; Rinberg et al., 2006b), how odor coding by mitral cell ensembles differs in the awake and anesthetized state is unclear. In this study, we show that the transition from the awake to anesthetized brain state has a dramatic impact on how olfactory information is represented by ensembles of mitral cells. By imaging large populations of mitral cells in individual mice, we find that odor-evoked ensemble

activity is much sparser and more temporally dynamic in the awake state and that anesthesia increases the density of odor representations by broadening the odor tuning of mitral cells. Importantly, we also show that the sparse and temporally dynamic ensemble activity during wakefullness is more efficient for odor population coding: compared to anesthetized brain states, fewer mitral cell responses in the awake state are required for accurate odor discrimination. BVD-523 cell line The temporal

dynamics of mitral cell ensemble activity have been proposed to contribute to odor coding (Bathellier et al., 2008; Friedrich et al., 2004; Friedrich and Laurent, 2001; Laurent et al., 1996; Mazor and Laurent, 2005). Indeed, here we demonstrate that odor classification in the awake state improves gradually as odor representations develop over time. In contrast, the temporal dynamics of odor representations are reduced in the anesthetized brain state, which contributes to a reduction in the population coding efficiency. We note that the temporal resolution of our imaging approach (∼6.3 Hz) precludes the assessment of finer temporal features of mitral Ribonucleotide reductase cell responses (Bathellier et al., 2008; Cury and Uchida, 2010; Shusterman et al., 2011). Nevertheless, our results reveal a strong temporal component to mitral cell odor representations in awake animals, which may be underestimated in recordings under anesthesia. Variability in respiratory behavior in the awake state could contribute to the temporal dynamics of mitral cell responses (Carey and Wachowiak, 2011; Verhagen et al., 2007). However, as we discuss below, the opposite effects of anesthesia on mitral cells and granule cells make it unlikely that respiration variability can fully account for the changes and we suggest that actions of local inhibitory circuits probably play an important role.