Classification of extracellularly recorded neurons by their discharge patterns and their correlates with intracellularly identified neuronal types in the frontal cortex of behaving monkeys.

Neurons in the cerebral cortex are not homogeneous. However, neuronal types have been ignored in most previous work studying neuronal processes in behaving monkeys. We propose a new method to identify neuronal types in extracellular recording studies of behaving monkeys. We classified neurons as either bursting or non-bursting, and then classified the bursting neurons into three types: (i) neurons displaying a burst of many spikes (maximum number of spikes within a burst; NSB max > or = 8) at a high discharge rate (maximum interspike interval; ISI max < 5 ms); (ii) neurons displaying a burst of fewer spikes (NSB max < or = 5) at a high discharge rate (ISI max < 5 ms); and (iii) neurons displaying a burst of a few spikes (NSB max < or = 7) at relatively long ISIs (ISI max > 5 ms). We found that the discharge patterns of the four groups corresponded to those of regular spiking (RS), fast spiking (FS), fast rhythmic bursting (FRB) and intrinsic bursting (IB) neurons demonstrated in intracellular recording studies using in vitro slice preparations, respectively. In addition, we examined correlations with the task events for neurons recorded in the frontal eye field and neuronal interactions for pairs of neurons recorded simultaneously from a single electrode. We found that they were substantially different between RS and FS types. These results suggest that neurons in the frontal cortex of behaving monkeys can be classified into four types based on their discharge patterns, and that these four types contribute differentially to cortical operations.

Modulation of V4 shifts from dependent to independent on feature during target selection.

We analyzed neuronal activities of visual area V4 of monkeys performing a delayed visual search task. To examine the temporal profile of factors influencing the neuronal activities, we conducted multiple regression analyses at 5 ms steps. During the period from 110 to 155 ms after the stimulus onset, there were neurons whose activity was suppressed when a target was presented near but beyond the neuron's receptive field (RF) compared to that when a target was within the RF. We referred this suppressive effect as an early period modulation. During the period from 155 to 280 ms after the stimulus onset, V4 activities were suppressed when a target was presented in any location outside of the neuron's RF. We referred this suppressive effect as a late period modulation. The magnitudes of the effect were equivalent across target locations when a target was beyond its RF. At the population level, while the modulation in the early period was correlated with stimulus selectivity, the modulation in the late period did not show such a correlation. These results suggest that V4 neurons have at least two distinct phases of modulations and those modulations contribute to select a target in the visual search task.

Responses of monkey prefrontal neurons during the execution of transverse patterning.

Recent functional imaging studies have suggested that the prefrontal cortex (PF) is engaged in the performance of transverse patterning (TP), which consists of 3 conflicting discriminations (A+/B-, B+/C-, C+/A-). However, the roles of PF in TP are still unclear. To address this issue, we examined the neuronal responses in 3 regions [the principal sulcus (PS), dorsal convexity (DC), and medial prefrontal cortex (MPF)] of the macaque PF during the performance of an oculomotor version of TP. A delayed matching-to-sample (DMS) task was used as a control task. The TP task-responsive neurons were most abundant in MPF. We analyzed the dependency of each neuronal response on the task type (TP or DMS), target shape (A, B, or C), and target location (left or right). Immediately after the choice cue presentation, many MPF neurons showed task dependency. Interestingly, some of them already exhibited differential activity between the 2 tasks before the choice cue presentation. Immediately before the saccade, the number of target location-dependent neurons increased in MPF and PS. Among them, many MPF neurons were also influenced by the task type, whereas PS neurons tended to show location dependency without task dependency. These results suggest that MPF and PS are involved in the execution of TP: MPF appears to be more important in the target selection based on the TP rule, whereas PS is apparently more related to the response preparation. In addition, some neurons showed a postsaccadic response, which may be related to the feedback mechanism.