Prefrontal cortex and spatial sequencing in macaque monkey

Summary 1. Single neuron activity was recorded from the prefrontal cortex of two macaque monkeys during the performance of a task involving spatial sequencing. The monkeys faced a panel displaying a central fixation point and three fixed targets (two lateral and one above the point of fixation). In the first phase of each trial, the three targets were turned on in random order: in the second phase, the animal had to press each target, still lighted, in the order of their illumination. Thus, successful performance of the task depended strongly on temporal memory. The animals were fitted with DC-EOG electrodes. 2. Three hundred and two task-related neurons were recorded in the superior arcuate area and caudal part of sulcus principalis. Among the cells whose pattern of activity appeared to be related to the sequencing task, five classes were distinguished: Visual tonic (VT), fixation, context, saccade related and visual phasic cells. In addition, a small number of cells appeared to be related to other aspects of the behavior, but not to the sequencing task. Our present analysis concentrates on two groups of sequencing task-related cells (VT and context cells). 3. The VT cells (35/302-11.5%) were recorded exclusively from the superior arcuate area. All VT cells increased their firing rate (sustained activation) during fixation of the central fixation point (FP) following onset of one of the three targets used, specific for a given cell (directional or spatial selectivity). In one group of VT cells, a shift in the eye position towards the specific peripheral target resulted in the return of the cells' firing rate to the pre-trial level. In the other group of VT cells, reset of the firing rate to pre-trial level was not related to the onset of fixation of the peripheral target. Sustained activation of the VT cells depended also on the sequential order of illumination of the specific target (temporal selectivity). In twenty-four cells (68.5% of VT cells) sustained activation was observed when the target came first in the sequence. Onset of the target in the second or third rank elicited either no response or only a short lasting phasic activation. In the remaining eleven cells (31.5% of VT cells), sustained activation was only observed when the target came second in a given sequence. The firing of the VT cells was correlated with the animals' performance of the task. On trials where the animals selected successive targets in an incorrect order, the temporal pattern of activation of VT cells was different from that in the correctly performed trials. Thus, the correct temporal encoding of a target appeared to be a prerequisite for the correct performance of a sequence. 4. The context cells (36.5%-16/302) were activated when the animal fixated a particular target during execution of the sequence and, like VT cells, were encountered exclusively in the penetrations through the superior arcuate area. Activation also depended on the state (illuminated vs extinguished or hit vs non-hit) of the non-fixated targets and/or on their time-relationships with respect to the fixated target hence context cells. 5. Properties of VT and context cells revealed in the present experiment are consistent with the idea that the superior arcuate area is involved in temporally and spatially extended structures of behavior. If so, the arcuate area would constitute a specialized part of prefrontal cortex implicated in construction of oculomotor plans.