Role of the caudal fastigial nucleus in saccade generation. II. Effects of muscimol inactivation.

1. We studied the effect of temporarily inhibiting neurons in the caudal fastigial nucleus in two rhesus macaques trained to make saccades to jumping targets. We placed injections of the gamma-aminobutyric acid (GABA) agonist muscimol unilaterally or bilaterally at sites in the caudal fastigial nucleus where we had recorded saccade-related neurons a few minutes earlier. 2. Unilateral injections (n = 9) made horizontal saccades to the injected side hypermetric and those to the other side hypometric (mean gain of 1.37 and 0.61, respectively, for 10 degrees target steps, and 1.26 and 0.81 for 20 degrees target steps; normal saccade gain was 0.96). Saccades to vertical targets showed a small but significant hypermetria and curved strongly toward the side of the injection. The trajectories and end points of all targeted saccades were more variable than normal. 3. After unilateral injections, centripetal saccades were slightly larger than centrifugal saccades (mean gains for ipsilateral saccades were 1.42 and 1.31, respectively, for 10 degrees target steps, and 1.37 and 1.15 for 20 degrees target steps). 4. Unilateral injections increased the average acceleration of ipsilateral saccades and decreased the acceleration of contralateral saccades. Injections decreased both the acceleration and deceleration of vertical saccades. 5. After dysmetric saccades, monkeys acquired the target with an abnormally high number of hypometric corrective saccades. Injection increased the average number of corrective saccades from 0.6 to 2.1 after 10 degrees horizontal target steps and from 0.8 to 2.1 after 20 degrees steps. The size of each successive corrective saccade in a series decreased, and the latency from the previous corrective saccade increased. 6. Bilateral injections (n = 2) of muscimol, in which we injected first into the left caudal fastigial nucleus and then, within 30 min, into the right, made all saccades hypermetric (mean gain for 10 degrees right, left, up, and down saccades was 1.18, 1.49, 1.43, and 1.10, respectively). Paradoxically, bilateral injection decreased both saccade acceleration and deceleration. Saccade trajectories and end points were more variable than normal. 7. To account for the effects of our injections, we propose that the activity of caudal fastigial neurons on one side normally helps to decelerate ipsilateral saccades and helps to accelerate contralateral saccades by influencing the feedback loop of the saccade burst generator in the brain stem. Without caudal fastigial activity the brain stem burst generator produces hypermetric, variable saccades. We therefore also propose that the influence of caudal fastigial neurons on the burst generator makes saccades more consistent and accurate.(ABSTRACT TRUNCATED AT 400 WORDS)