Sensitivity of rat cortical neurons in distinguishing taste qualities by individual and correlative activities.

To examine the possibility that different taste qualities are represented by the correlative activity of cortical gustatory neurons, we made simultaneous recordings of neuron pairs during application of four basic tastes into the oral cavity of anesthetized rats and the following observations were made: (i) in 30 of 67 pairs of taste responsive neurons, peaks (troughs in a few cases) were produced in the cross-correlograms (CCs) during stimulation with some tastants; (ii) the correlative discharges occupied 6-8% of the total spikes discharged by individual neurons during taste stimulation and occurred, in a considerable number of cases, even during stimulation with tastants to which one or both of the component neurons of a pair were apparently non-responding (often sucrose and quinine); (iii) the number of tastants to which a neuron pair responded with a significant correlative activity was often greater than the number of tastants to which the component neurons of the same pair responded with significant changes in discharge rate; (iv) there was no significant difference between the correlative (formation of peaks or troughs in the CC) and individual (change in discharge rate of individual neurons) ways of coding in the sensitivity to distinguish between two taste qualities ranked to be adjacent on the basis of the number of spikes composing the response; and (v) the peaks or troughs appearing in two CCs during stimulations with two kinds of tastants were compared with regard to overlapping of their delay ranges and widths. The spikes in the non-overlapping portion of each peak (suppressed spike number in the case of troughs) are supposed to be able to contribute to two-taste discrimination: the correlated discharges occurring with a delay time that corresponds to the overlapping portion can in no way be judged differently, but the spikes falling in the non-overlapping portion may contribute to the differentiation. The ratio of the non-overlapping portion to the entire peak (or trough) was 0.35 on average. It is concluded that temporal coding of taste qualities seems to operate effectively in the gustatory cortex.