Odor Images and Tunes

while inhibiting others through these second messenger Odors in our natural and urban world (e.g., fragrances pathways, that single odorants can activate different of food, wines, flowers, or perfumes) are complex conductances in different neurons from the same aniblends of volatile compounds. Humans can distinguish mal, and that a single ORN can be excited by one odor hundreds of thousands of odors (Mori and Yoshihara, and inhibited by another (Dubin and Dionne, 1994; Ache, 1995). Yet,each blend generallyelicits a singularpercept 1994). These results begin to provide a mechanistic un(e.g., madeleines, St. Emilion, freesias, or N8 5) from derstanding for older descriptions of complex ORN which individual ingredients are hard, if not impossible, physiological responses to odors. They also indicate to segment and identify. In addition, odors (however that understanding the response of individual ORNs to complex) are easily learned, memorized, and recogodors will require the knowledge not only of the binding nized, emphasizing the great importance of chemical properties of the odor receptor protein(s) that they exsenses during our evolution. The olfactory system therepress, but also of the transduction cascades to which fore offers a remarkable opportunity to understand how they are coupled and of the possibly complex dynamics brains solve a class of complex and natural pattern recresulting from interacting excitatory and inhibitory pathognition problems. ways in the same neuron. The representation of olfactory signals has long been Second, physiological and histochemical results from thought to rely on distributed codes. Indeed, most invesmolluscs (Gelperin, 1994) and insects (Müller and Buchtigators agree that these spatiotemporal codes or ner, 1993; Elphick et al., 1995) indicate that, like in mamacross-fiber patterns involve the activity of ensembles of mals (Bredt et al., 1991; Breer and Shepherd, 1993), neurons, whose combined activity specifically encodes nitric oxide synthase is present in neurons (possibly each stimulus. But what precisely are the rules of this local) in the first olfactory relay. Gelperin (1994) further distributed representation? Are molecular signals spashowed that nitric oxide is necessary for the expression tially mapped in the brain? If so, how reliably and how of oscillatory dynamics in the olfactory circuits of the consistently? Is synchronous activity important? If so, mollusc Limax.