Neuronal activity related to saccadic eye movements in the monkey's dorsolateral prefrontal cortex

Single-neuron activity was recorded from the prefrontal cortex of monkeys performing saccadic eye movements in oculomotor delayed-response (ODR) and visually guided saccade (VGS) tasks. In the ODR task the monkey was required to maintain fixation of a central spot throughout the 0.5-s cue and 3.0-s delay before making a saccadic eye movement in the dark to one of four or eight locations where the visual cue had been presented. The same locations were used for targets in the VGS tasks; however, unlike the ODR task, saccades in the VGS tasks were visually guided. Among 434 neurons recorded from prefrontal cortex within and surrounding the principal sulcus (PS), 147 changed their discharge rates in relation to saccadic eye movements in the ODR task. Their response latencies relative to saccade initiation were distributed between -192 and 460-ms, with 22% exhibiting presaccadic activity and 78% exhibiting only postsaccadic activity. Among PS neurons with presaccadic activity, 53% also had postsaccadic activity when the monkey made saccadic eye movements opposite to the directions for which the presaccadic activity was observed. Almost all (97%) PS neurons with presaccadic activity were directionally selective. The best direction and tuning specificity of each neuron were estimated from parameters used to fit a Gaussian tuning curve function. The best direction for 62% of the neurons with presaccadic activity was toward the contralateral visual field, with the remaining neurons having best directions toward the ipsilateral field (23%) or along the vertical meridian (15%). Most postsaccadic activity of PS neurons (92%) was also directionally selective. The best direction for 48% of these neurons was toward the contralateral visual field, with the remaining neurons having best directions toward the ipsilateral field (36%) or along the vertical meridian (16%). Eighteen percent of the neurons with postsaccadic activity showed a reciprocal response pattern: excitatory responses occurred for one set of saccade directions, whereas inhibitory responses occurred for roughly the opposite set of directions. Sixty PS neurons with saccade-related activity in the ODR task were also examined in a VGS task. Forty of these neurons showed highly similar profiles of directional specificity and response magnitude in both tasks, 13 showed saccade-related activity only in the ODR task, and 7 changed their response characteristics between the ODR and VGS tasks. The majority of PS neurons with saccade-related activity in the ODR task showed that activity only in conjunction with task-related saccades; comparable activity was not observed in association with saccadic eye movements made during the intertrial interval (ITI), even though both classes of saccades were performed in the dark. This context dependency was characteristic of both pre- and postsaccadic activity of PS neurons. Among 53 frontal eye field (FEF) neurons tested in the ODR task, 26 had saccade-related activity, with 42% of these having presaccadic responses. All saccade-related activity in FEF was directionally selective, and 72% of their best directions were toward the contralateral visual field. All six FEF neurons also examined in a VGS task showed similar responses across tasks. These results indicate that PS neurons participate in the control of purposive saccadic eye movements. Most saccade-related activity of PS neurons was postsaccadic, and such postsaccadic activity may be important in terminating the processes leading to memory-guided saccades. Their presaccadic activity, although a smaller proportion of the total, could help initiate and guide saccadic eye movements in the ODR task.