The process by which the mammalian nervous system represents the features of a sapid stimulus that lead to a perception of taste quality has long been controversial. The labeled-line (sparse coding) view differs from the across-neuron pattern (ensemble) counterpoint in proposing that activity in a given class of neurons is necessary and sufficient to generate a specific taste perception. This article critically reviews molecular, electro-physiological, and behavioral findings that bear on the issue. In the peripheral gustatory system, the authors conclude that most qualities appear to be signaled by labeled lines; however, elements of both types of coding characterize signaling of sodium salts. Given the heterogeneity of neuronal tuning functions in the brain, the central coding mechanism is less clear. Both sparse coding and neuronal ensemble models remain viable possibilities. Furthermore, temporal patterns of discharge could contribute additional information. Ultimately, until specific classes of neurons can be selectively manipulated and perceptual consequences assessed, it will be difficult to go beyond mere correlation and conclusively discern the validity of these coding models.

No abstract

No abstract

No abstract

1. The receptive field and topographic organization of single orosensory neurons located throughout the rostral division of the nucleus of the solitary tract (rNST) was studied by determining their responsiveness to gustatory stimulation of the entire oral cavity and to gustatory and mechanical stimulation of restricted oral regions. The rNST contained roughly equal numbers of two distinct populations of orosensory neurons, one responsive exclusively to oral mechanical stimulation (M neurons), the other to gustatory stimulation (G neurons). Some G neurons also responded to oral somatosensory stimuli, but usually less vigorously than to gustatory stimuli. The distribution of these two populations of rNST neurons was topographically organized: G neurons were centered anteriorly and medially to M neurons. 2. Eight of 44 G neurons responded only when the whole oral cavity was stimulated, but the remaining 36 cells responded to circumscribed stimulation of taste buds on the anterior tongue (AT), foliate papillae of the posterior tongue, nasoincisor ducts, retromolar mucosa (RM), or soft palate (SP). Overall, AT and SP stimulation were the most effective, and RM stimulation the least effective, for activating nucleus of the solitary tract (NST) G neurons. 3. Approximately half of the G neurons for which a receptive field could be defined (N = 36) responded to stimulation of a single taste receptor subpopulation, but the remaining neurons received convergent input from two or more taste bud groups. The receptive field configurations for convergent G neurons were orderly: convergence occurred preferentially between receptor subpopulations either within the anterior oral cavity (AO) or the posterior oral cavity (PO). An AO-PO distinction also was reflected in the topographic organization of gustatory responses. The mean location of neurons responding optimally to AO gustatory stimulation was more anterior in the NST, and also tended to be more lateral and ventral than the location of neurons that responded optimally to PO stimulation. 4. Forty-four rNST M neurons responded to innocuous mechanical stimulation of restricted areas of the tongue, palate, buccal mucosa, or periodontium. Stimulation of the hard palate and circumvallate papilla were most effective, whereas periodontal stimulation was least effective for activating these cells. 5. A majority (32 of 44) of rNST M neurons responded to stimulation of more than one of the oral sites tested.(ABSTRACT TRUNCATED AT 400 WORDS)