Changes in the control of arm position, movement, and thalamic discharge during local inactivation in the globus pallidus of the monkey

1. To examine the effect of disruption of basal ganglia output on limb stability and movement, muscimol was injected into the internal globus pallidus (GPi) of monkeys trained to make arm movements to visible or remembered targets in a two-dimensional workspace. 2. Injections of as little as 0.25 mu g muscimol at GPi sites at which pallidal neurons with arm movement activity had been re corded were followed by drift of the contralateral arm within <10 min. Drift was usually in the flexor direction. Injections at a few sites in or near the external pallidal segment sometimes were followed by extensor drift. 3. Drift was active (accompanied by activation of agonist muscles), but could be overcome by the animal, resulting in an oscillating movement off and on the required position. 4. The pallidal-receiving (PR) area of the thalamus was identified by recording the response of thalamic neurons to stimulation in the globus pallidus. The activity of 15 neurons identified as PR cells (n = 6) or within the PR region was recorded both before and after injection of muscimol into GPi. After the injection, the tonic discharge increased during the hold period in 47% of the cells studied. When postural drift also occurred, there was a close temporal correlation between the postinjection time at which drift occurred and the time at which the tonic discharge rate increased in thalamic neurons that were clearly related to arm movement. 5. The peak velocity of arm movements to visible or remembered visual target locations was decreased after injection of muscimol into GPi, sometimes with an increase in movement time. 6. The firing rate of PR thalamic neurons after injection of muscimol was also increased during the perimovement period. Because of the increase in the tonic discharge rate, however, the phasic movement-related change in activity could stay the same or even decrease. Postinjection changes in this movement-related phasic activity, however, were not necessarily coincident with changes in peak movement velocity. 7. Changes in reaction time were variable after injection of muscimol. In some cases it was increased, and in others decreased. The time of onset of phasic movement-related changes in the activity of PR neurons studied was not altered by the injection. 8. Our data indicate that the tonic inhibitory output of GPi, in particular to the cortical motor areas, is especially important in the maintenance of postural stability. In the absence of normal pallidal output, desired limb position can be achieved on the basis of either current or prior visual cues, but targeted movements are slowed.