Information processing in the taste system of primates

ABSTRACT Analysis of the activity of single neurones in the gustatory pathways in primates (cynomolgus monkeys) shows that the tuning of neurones to the four prototypical stimuli 1·0moll−1 glucose, 1·0moll−1 NaCl, 0·001moll−1 quinine-HCl and 0·01 mol 1−1 HCl becomes sharper as information progresses through the taste system from the first central relay in the brainstem, the nucleus of the solitary tract, via the thalamus to the primary taste cortex in the frontal operculum and insula to reach a secondary cortical taste area in the caudolateral orbitofrontal cortex. Feeding monkeys to satiety with glucose has no effect on the gustatory responses of neurones in the nucleus of the solitary tract, the frontal opercular taste cortex or the insular taste cortex, but decreases the magnitude of the neuronal responses of orbitofrontal cortex taste neurones which respond to glucose to zero. The responses of orbitofrontal cortex taste neurones decrease to foods on which the monkey is fed to satiety, but continue to foods which have not just been eaten, that is, they reflect sensory-specific satiety, the phenomenon in which the pleasantness of the taste of a food and its acceptability, but not those of other foods, is decreased by eating that food to satiety. It has been found that the orbitofrontal cortex taste area or areas receives inputs from different modalities: single neurones with unimodal responses to taste (47 %), olfactory (12 %) and visual (10 %) stimuli are found in close proximity to each other. Moreover, some single neurones show multimodal convergence, responding, for example, to taste and visual inputs (17%), taste and olfactory inputs (10%), and olfactory and visual inputs (4%). Some of these multimodal single neurones have corresponding sensitivities in the two modalities: they respond best to sweet tastes (e.g. 1 mol 1−1 glucose), and respond more in a visual discrimination task to the visual stimulus which signifies sweet fruit juice than to that which signifies saline; or they respond to sweet taste and, in an olfactory discrimination task, to fruit juice odour. These results suggest that multimodal representations are formed in the orbitofrontal cortex secondary and related taste areas. It is proposed that tuning becomes sharper in the taste system through unimodal processing stages so that, after this processing, associations can be made to other modalities with minimal interference and maximum capacity in an association memory; and so that satiety can operate with some specificity, allowing responses to foods eaten to decrease without producing a decrease in responsiveness to other gustatory stimuli. It is suggested that, nevertheless, ensemble encoding is used because this allows the emergent properties of completion, generalization and graceful degradation to be generated in pattern association matrix memory neuronal networks used to build multimodal representations. It is suggested that, in the cerebral cortex, competitive learning is part of what occurs in neuronal networks in order to build the finely tuned ensemble-encoded representations required for association memories in multimodal cortical areas and the amygdala, and autoassociation memories in brain areas such as the hippocampus, to operate. It is also suggested that the backprojections to the cerebral cortex from the hippocampus and amygdala, and between adjacent areas of the cerebral cortex, are used to influence the storage of information in the cerebral cortex, as well as for recall, attention and dynamic top-down processing. Analysis of neuronal activity in the taste system thus leads to hypotheses about the principles of sensory analysis, and the taste system may be a useful model system for studying these principles.