SACCADIC BURST NEURONS IN THE OCULOMOTOR REGION OF THE FASTIGIAL NUCLEUS OF MACAQUE MONKEYS

1. The discharge of 255 neurons in the fastigial nuclei of three trained macaque monkeys was investigated during visually guided saccades. Responses of these neurons were examined also during horizontal head rotation and during microstimulation of the oculomotor vermis (lobules VIc and VII). 2. One hundred and two units were characterized by bursts of firing in response to visually guided saccades. Ninety-eight of these (96.1%) were located within the anatomic confines of the fastigial oculomotor region (FOR), on the basis of reconstruction of recording sites. During contralateral saccades, these neurons showed bursts that preceded the onset of saccades (presaccadic burst), whereas, during ipsilateral saccades, they showed bursts associated with the end of saccades (late saccadic burst). They were hence named saccadic burst neurons. Sixty-one saccadic burst neurons (62.2%) were inhibited during microstimulation of the oculomotor vermis with currents < 10-mu-A. 3. All saccadic burst neurons were spontaneously active, and the resting firing rate varied considerably among units, ranging from 10 to 50 imp/s. The tonic levels of activity did not correlate significantly with eye position. 4. The presaccadic burst started 18.5 +/- 4.7 (SD) ms (n = 45) before the onset of saccades in the optimal direction (the direction associated with the maximum values of burst lead time, number of spikes per burst, and burst duration). Optimal directions covered the entire contralateral hemifield, although there was a slightly higher incidence in both horizontal and upper-oblique directions in the present sample. The duration of the presaccadic burst was highly correlated with the duration of saccade (0.85 less-than-or-equal-to r less-than-or-equal-to 0.97). 5. The late saccadic burst was most robust in the direction opposite to the optimal in each unit (the nonoptimal direction). Its onset preceded the completion of ipsilateral saccade by 30.4 +/- 5.9 ms. The lead time to the end of saccade was consistent among different units and was constant also for saccades of various sizes. Thus the late saccadic burst started even before the saccade onset when the saccade duration was < 30 ms. Unlike the presaccadic burst, its duration was not related to the duration of saccade. 6. Discharge rates of saccadic burst neurons were correlated neither to eye positions during fixation nor to the initial eye positions before saccades. 7. Eye-position units and horizontal head-velocity units were located rostral to the FOR. Thirteen eye-position units were found rostral to the saccadic burst neurons, and 5 of these units were inhibited during vermal stimulation. Of a total of 66 horizontal head-velocity units, 16 units increased discharges during the ipsilateral phase of sinusoidal head rotation (type I), and 50 units increased discharges during contralateral head rotation (type II). They were widely scattered in the rostral and central portions of the fastigial nucleus. 8. The present study has shown that the FOR is important for the processing of visuomotor information during visually guided saccades. The saccadic burst neurons showed a burst of firing, which may regulate the duration of a saccade in the contralateral direction, and a late saccadic burst, which may work as a brake during an ipsilateral saccade. These response features indicate that the cerebelllar output may carry the information necessary for controlling the amplitude of saccade.