g., Ojima et al., learn more 1984) and that odors generally evoke widely distributed PCx activity (Illig and Haberly, 2003). Recent experiments have shown that different odorants activate unique subpopulations of neurons distributed across the PCx without spatial preference (Stettler and Axel, 2009) and that the projections of individual glomeruli (Nagayama et al., 2010 and Sosulski et al., 2011) and of single mitral cells (Gosh et al., 2011) are broadly

distributed in the PCx. A second set of questions concerns how the odor features represented by MOB glomeruli are recombined in the cortex. How many different mitral cell inputs synapse on an individual PCx neuron? What are the numbers and distribution of glomeruli from which the inputs to a single neuron arise? Up until recently, there has been little direct evidence concerning MOB to PCx convergence. It is known that PCx neurons can respond to dissimilar odorants that are likely to activate nonoverlapping glomeruli (Rennaker et al., 2007 and Poo and Isaacson, 2009). Some PCx neurons respond to electrical stimulation only when more than one glomerulus is coactivated (Apicella et al., 2010), and odor mixtures can activate PCx neurons that are not activated by the components alone (Stettler and Axel, 2009). Although intracortical excitation

could contribute to some of these observations, it was recently shown that individual PCx neurons indeed receive anatomical connections from multiple broadly distributed mitral cells (Miyamichi et al., 2011). Further critical INCB024360 questions relate to more detailed features

of the integrative properties of individual PCx neurons. How strong are the individual functional inputs from each glomerulus? How many glomeruli connect to each PCx cell and how many inputs must be coactive for a PCx neuron to respond? Do inputs combine linearly or nonlinearly? In this issue of Neuron, Davison and Ehlers (2011) provide important new insight into these issues by using in vivo laser scanning glutamate uncaging to create distributed artificial patterns of glomerular activation in the MOB while recording in the PCx in anesthetized mice. With this approach, the authors were able to independently activate targeted locations within Parvulin the glomerular layer of the MOB with near single glomerular spatial resolution. Davison and Ehlers’s data address several aspects of the convergence and integration of MOB inputs by individual PCx cells. First, PCx neurons did not generally respond to single-site stimulation; PCx firing was only triggered reliably by joint activation of at least three uncaging sites. Moreover, for a given number of sites, PCx activation was specific to the pattern of those sites: each cell responded differentially to different spatial patterns and different PCx cells responded differentially to a particular pattern.