Monkey Prefrontal Neuronal Activity Coding the Forthcoming Saccade in an Oculomotor Delayed Matching-to-Sample Task

Abstract

Hasegawa, Ryohei, Toshiyuki Sawaguchi, and Kisou Kubota. Monkey prefrontal neuronal activity coding the forthcoming saccade in an oculomotor delayed matching-to-sample task. J. Neurophysiol. 79: 322–333, 1998. To determine the role of the dorsolateral prefrontal cortex (PFC) in the selection of memory-guided saccadic eye movements, we recorded the activities of PFC neurons while macaque monkeys performed an oculomotor delayed matching-to-sample task. The task was designed to dissociate motor factors from visual factors in the selection and retention of the direction of the forthcoming saccade during delay periods after the visual cue but before the GO signal was presented. While the monkey fixated on a central fixation spot (FX period, 1 s), a sample cue (1 of 4 geometric figures) and a matching cue composed of two geometric figures were presented in succession (SC and MC periods, respectively, 0.5 s) with a brief delay (D1 period, 1 or 1.5 s). After another delay (D2 period, 1.5 s), the monkey made a saccade (GO period, <0.5 s) toward one of four locations (the goal) that had been indicated by the combination of the sample and matching cues in the MC period. We recorded the activities of 224 neurons in the periprincipal sulcal area of 3 hemispheres of 2 monkeys. Sixty-five neurons (29%) showed a significant increase in activity during the D2 period. Some of these also responded during other phases of the task (SC period, n = 32; D1, 22; MC, 53; GO, 47). Some of the activity during the D2(52/65, 80%) and GO (40/47, 85%) periods was associated with the direction of the forthcoming saccade (“direction selective”). Although most MC-period activities of D2 neurons were direction selective (38/53, 73%), a fraction of them (14/38) was also affected by both saccade direction and matching cue pattern. To compare quantitatively the contribution of motor (saccade direction) and visual (matching-cue pattern) factors to the activity of D2 neurons, we calculated directional and visual dependency indices (DDI and VDI) for each of the three periods (MC, D2, and GO). In both the D2 and GO periods, D2 neurons with high DDI values and low VDI values predominated. In the MC period, however, there was no significant difference between the distributions of DDI and VDI values. These findings suggest that PFC neurons store the direction of memory-guided saccades during a delay period before eye movement and that the same neurons may be involved in the decision-making process that underlies the selection of the saccade direction during the MC period.