Basal ganglia motor control. III. Pallidal ablation: normal reaction time, muscle cocontraction, and slow movement.

JOURNALOF NEUROPHYSIOLOGY Vol. 65, No. 2, February 199 I. Printed

in U.S.A.

Basal Ganglia Motor Control. III. Pallidal Ablation: Normal Reaction Time, Muscle Cocontraction, and Slow Movement J. W. MINK AND W. T. THACH Departments of Anatomy and Neurobiology, Neurology and Neurosurgery, and The McDonnell Center for the Study of Higher Brain Function, Washington University School of Medicine, St. Louis, Missouri 63110 SUMMARY

AND

CONCLUSIONS

1. Inactivation of the portions of globus pallidus pars interna (GPi) containing the greatest concentration of wrist-related neurons was achieved in two rhesus monkeys with microinjections of muscimol (temporary) and kainic acid (permanent). 2. After muscimol injection, there was onset within 30 s of 1) tonic and phasic coactivation of wrist flexors and extensors; 2) slightly greater activation of the flexors, giving a flexor bias in postural holds and the endpoint of movements; and 3) slowness of all movements with a prolonged movement time. Nevertheless, 4) movements made by lessening prior loaded muscle activity (to move in the direction of the load) were slower than movement made by increasing muscle activity (to move against the direction of the load). Despite marked slowing of all movements, there was 5) a normal reaction time for movement onset. Finally, there was 6) a reduced amplitude of most movements. Open room behavior included 7) spiraling contralateral to the lesion while walking. Effects were reproducible (12 injections), were apparent for 7-8 h and were usually completely gone by the next day’s testing. 3. After kainic acid injection, there was a period of mixed effects, followed by a period of permanent defects (observed for up to 24 days) that duplicated the temporary effects of muscimol. 4. By contrast, muscimol inactivation of the cerebellar dentate nucleus resulted in 1) a prolonged reaction time and 2) an increased variability of movement trajectory, but 3) without change in movement time or peak velocity. Open room behavior included overshoot in reaching for fruit with the forelimb ispilateral to the injection. 5. From the facts that normal pallidal neurons fire constantly, that pallidal neurons inhibit their target neurons, and that the muscimol effect was immediate, we conclude that the release of the target neurons from the tonic inhibition allowed them to fire in patterns that promoted a maintained state of cocontraction of agonist and antagonist muscles. From the fact that movement time was prolonged, we conclude that the maintained state of neural activity that caused the muscle cocontraction interfered with the commands for voluntary movement, which were generated by other mechanisms. From the fact that reaction time for movement onset was normal, we conclude that the pallidal neurons may play little or no role in the voluntary initiation of these movements, which are instead generated by other structures that include the anterior cerebral cortex and the lateral cerebellum.

INTRODUCTION

The idea that the basal ganglia somehow control normal posture (Marsden 1982; Wilson 1928) is based on the classic observation that patients with certain degenerative diseases of the basal ganglia, such as Parkinson’s and late Huntington’s, suffer from rigidly held abnormal body postures. 330

As to the mechanism of the normal control, the basal ganglia output was previously thought to project exclusively to structures in the brain stem that were normally involved in tonic neck, contact righting, and antigravity reflexes. Until recently (Penny and Young 198 1; Uno and Yoshida 1973, that output was assumed to be excitatory. Removal of the basal ganglia output was supposed to remove a tonic drive from the posture-generating brain stem structures. In time, a gradual increase in the excitability of the brain stem neurons, released from their normal excitatory control, was thought to give rise to the abnormal postures. As to the normal function of the basal ganglia, Wilson (1928) inferred that they help generate normal postures through a controlled action on the appropriate brain stem nuclei. The idea that the basal ganglia generate movement (Denny-Brown 1962; Marsden 1982) also arose from observations of patients with basal ganglia disease. Those conditions that produce rigid postures also produce a slowing of movement (bradykinesia) and, in advanced cases, a difficulty in moving at all (akinesia). These observations were proposed to be consistent with the idea that the basal ganglia contain the mechanisms not only for the generation of posture but also for the initiation of volitional movement. Other observations also held to be consistent with the idea are that lesions “high” in the basal ganglia (putamen and subthalamic nucleus) produce involuntary movements that, although spontaneous, have the appearance of normal coordinated willed movement. The interpretation was that destruction of controlling inputs releases a movement generator in the pallidum, such that the normal movements in its repertoire are abnormally displayed. Experimental work on lesioning the basal ganglia output at the level of the pallidum was, for a time, interpreted to support the above models. Denny-Brown (1962) reported that monkeys with large bilateral surgical lesions of the globus pallidus (GP) had a rigidly flexed posture. Flexed posture had also been seen in monkeys with large bilateral lesions of GP and substantia nigra produced by carbon disulfide (Richter 1945). These observations therefore suggested that lesions that decrease the basal ganglia output result in a persistently maintained abnormal flexed posture. The flexed posture of Parkinson’s disease in humans has been reported to be similar to that produced by bilateral pallidal ablation in monkeys; and, from the similarity of the behavioral deficits, it was inferred that this disease is also characterized by a reduced pallidal output (Denny-Brown 1962). In the reports cited above of large bilateral !esions of pallidum (Denny-Brown 1962; Richter 1945) and 6-OHDA le-

0022-3077/9 1 $1 SO Copyright 0 199 1 The American Physiological Society Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (144.173.006.094) on August 5, 2018. Copyright © 1991 American Physiological Society. All rights reserved.

PALLIDAL

INHIBITION

OF

sion of substantia nigra pars compacta (Filion 1979), monkeys have also shown, along with the maintained flexed posture, a slowing of movement and in the extreme case a difficulty in initiating movements. Recent work has raised questions and conflicts concerning the above models of basal ganglia function. The present paper addresses especially the results of pallidal lesion. First, it has since been argued that lesion of pallidal output does not produce rigidity of posture, slowed movement, and impaired movement initiation, as was previously claimed (Georgopoulos and DeLong 198 1). DeLong and Georgopoulos ( 198 1) point out that Denny-Brown’s bilateral lesions were very large, and suggest that the structures critical for producing the deficits lay beyond the globus pallidus pars interna (GPi). Unilateral pallidal inactivation has been reported to produce little or no observed defect in the posture of untrained monkeys (Kennard 1944; Ranson and Berry 1941) or even in those trained to move (Amato et al. 1978; Horak and Anderson 1984). Indeed, DeLong has recently proposed that the abnormal rigid postures are produced by an increased rather than a decreased pallidal output (DeLong 1990). Second, unilateral cooling or ablating of the GP has been reported to prolong the movement time without prolonging the reaction time or preventing movement initiation (Amato et al. 1978; DeLong and Coyle 1979; Horak and Anderson 1984). The same has been reported for patients with Parkinson’s disease (Evarts et al. 198 1). Yet this is countered by other studies of patients with Parkinson’s disease who show delayed reaction times as well as prolonged movement times (Pullman et al. 1988). Third, the above model would appear not to be consistent with new information that the output of the pallidum is inhibitory (Hikosaka and Wurtz 1983; Penny and Young 198 1; Uno and Yoshida 1975). Removal of tonic inhibition should produce an immediate rather than a delayed increase in the excitability of the target neurons. In this paper, we examine the effects of muscimol and of kainic acid injection into the GPi on four of the same tasks in the two animals in which the activity of GP neurons had been characterized previously (Mink and Thach 199 la,b). The present experiments were designed to look for the presence of postural rigidity, and especially at the time course of its development, if present. If the rigidity were to appear only late after a permanent pallidal inactivation (kainic acid), the occurrence would be consistent with supersensitivity that might occur after release from an excitatory, possibly executive, control. If the rigidity were to occur immediately after a temporary pallidal inactivation (muscimol), in which the excitatory effects of surgical, electrolytic, and kainic acid lesioning are obviated, the occurrence would be consistent with sudden release from tonic inhibition (Penny and Young 1981; Uno and Yoshida 1975). Because unilateral pallidal inactivation has produced little or no observed defect in the posture of untrained monkeys (Kennard 1944; Ranson and Berry 1941) or even in those trained to move (Amato et al. 1978; Horak and Anderson 1984), the maintenance of posture was carefully controlled and monitored by having the monkey maintain and change the angle of the wrist within 10’ of wrist arc against varying flexor and extensor torque loads, while re-

POSTURAL

MECHANISMS

331

cording electromyograph (EMG) from wrist flexors and extensor muscles. In the present experiments, maintained position and movement were both controlled at a single joint, so that the effects of pallidal inactivation could be seen on both movement and posture. EMG was monitored in agonist and antagonist muscles holding and moving the joint, so that muscular cocontraction might be detected that might not otherwise have altered the measured joint position. Moving pictures were also taken of the animal’s open room and cage behavior. The tasks were performed with constant torque loads opposing and assisting the direction of movement, such that a movement could be made either by contracting and shortening a preloaded muscle to carry the load to a final position or by relaxing and lengthening a preloaded muscle to let the load carry the limb to the final position. With this paradigm, we could dissociate those movement deficits that might be due to an abnormal increase of a maintained postural activity that cannot be turned off from an abnormal movement initiatory mechanism that cannot be turned on. As an added control, we determined in one monkey the effect of cerebellar dentate nucleus inactivation to provide a comparison with the effect of pallidal inactivation on both movement and posture at the one joint. Previous studies have suggested that dentate lesions prolong reaction time and delay the initiation of movement but do not disrupt movement time or posture (Meyer-Lohmann et al. 1977; Spidalieri et al. 1983; Thach 1975). Our observations, together with those of the preceding papers (Mink and Thach 199 la,b), lead us to accept the idea that the basal ganglia may control (through inhibition) normal postural mechanisms but to reject the idea that the basal ganglia initiate the movements that we studied. Instead, our results are consistent with the hypothesis that a role of the basal ganglia is to switch off the maintained motor activities, possibly including normal postural mechanisms, that would otherwise interfere with voluntary movement commands at central nodal points and the final common path. METHODS

Four tasks were used to determine the effect of inactivation or ablation of the pallidum or cerebellum: 1) visually guided step tracking (VisStep), 2) visually guided ramp-tracking (VisRamp), 3) visually guided sine-tracking (VisSine), and 4) self-paced rapid alternation (SelfSine). The details of the task design, training, surgery, and histology are described in detail in an accompanying paper (Mink and Thach 199 1a).

Surgery Chambers for single-unit recording had been implanted for these animals as described in the companion paper (Mink and Thach 199 1a). When the GP and cerebellar nuclei had been thoroughly mapped with the single-unit recording technique, a series of drug injections was made during performance of the tasks.

Muscimol Because a substantial portion of the striatopallidal and cerebellar corticonuclear afferents use y-aminobutyric acid (GABA) as its neurotransmitter, the GABA agonist muscimol was injected to

Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (144.173.006.094) on August 5, 2018. Copyright © 1991 American Physiological Society. All rights reserved.

332

J. W. MINK

AND W. T. THACH

A 13

FIG. 1. Location of muscimol and kainic acid injections into GP of the 1st monkey. Both muscimol injections and the 1 kainic acid injection in this monkey were made in the same location. Track left by the syringe is depicted as a solid black line; extent of gliosis resulting from the kainic acid lesion is indicated by cross-hatching. These injections were centered at the border of the ventromedial external segment and the dorsolateral part of the internal segment. GPi, globus pallidus internal segment; GPe, globus pallidus external segment.

inhibit reversibly the firing of the target neurons. All injections were made with a Hamilton or Glenco microsyringe that was stereotaxically positioned in the GP or cerebellar dentate nucleus in the region of the highest density of task-related neurons. Muscimol was mixed in physiological saline to a concentration of 8.8 mM, the pH was adjusted to 7.3, and the solution was injected through a microsyringe that was mounted on the recording chamber over the GP contralateral to the trained hand. One or 2 ~1 of muscimol was injected into the region of the pallidum with the highest density of task-related neurons. In the second monkey, muscimol injections were also made in the region of the cerebellar dentate nucleus, ipsilateral to the trained hand, where task-related neurons were recorded. A typical muscimol injection session consisted of 1) observing and tine filming the monkey before the injection while he reached for pieces of apple, sat, walked on all fours, stood up, and climbed on the cage rack; 2) recording EMG and wrist position and velocity before the injection while the monkey performed the tasks; 3) injection of muscimol in four equal quantities (0.25 or 0.5 ~1) with 2.5 min between each increment; 4) recording EMG, position, and velocity during task performance after the injection; and 5) observing and tine filming the monkey after the injection. Data

were recorded during one complete cycle through all of the tasks (96 trials). After muscimol injection, data were recorded during at least one and usually two complete cycles through the tasks. At least 2 days passed between muscimol injections to assure that there was no residual effect of the previous injection; no muscimol was injected until the behavior had returned to normal from the previous injection. Injections of 1 and 2 ~1 of normal saline were made into the pallidum of each monkey to control for nonspecific effect of fluid injection.

Kainic acid Permanent ablation of the GP was accomplished with the excitatory neurotoxin kainic acid (McGeer and McGeer 1978). Kainic acid was mixed in physiological saline to a concentration of 4.7 mM, the pH was adjusted to 7.3, and the solution was injected into the area of the pallidum where the effect of muscimol was the greatest. One microliter of kainic acid was injected into the GP of the first monkey and 2 ~1 of kainic acid were injected in the second monkey. Preceding kainic acid injection, EMG and wrist position and velocity were recorded daily for 1 wk to assure that task performance was stable. After kainic acid injection, EMG, position, and

Al3

Al2

11

3

FIG. 2. Location of muscimol and kainic acid injections in GP of the 2nd monkey. Five muscimol injections were made into the GP of this monkey, 2 at site I, 1 at site 2, and 2 at site 3. The kainic acid injection in this monkey was made at site 1. Track left by the syringe is depicted as a solid black line; extent of gliosis resulting from the kainic acid lesion is indicated by cross-hatching. Solid black region around the end of track 1 indicates a hole in the tissue that may have resulted from repeated injections. All injections were centered in the internal segment. GPi, globus pallidus internal segment; GPe, globus pallidus external segment.


PALLIDAL FLEXORS

/: *,’:;zzzzee.-.l-

OF

POSTURAL

EXTENSORS

LOADED I

INHIBITION

A

MECHANISMS

333

LOADED B _- *--.- - . - , - 7.v

f .: WRIST POSITION

I/

w---v-

/

WRIST VELOC IT Y --, k------t.JII

-4clv

--b---

--

ai*

EXTENSOR EMG

D WRIST POSIT ION

-- .-----.. --_--

-w-e-- -

WRIST VELOC I TY

FIG. 3. wrist position, velocity, and EMG in VisStep task performance before and after injection of muscimol into the pallidurn. A-D: data are time-aligned on the onset of wrist movement (center vertical line). Arrow represents onset of target movement. Abscissa scale is in milliseconds with time 0 indicating the start of movement. Solid traces represent data before injection of muscimol; dotted traces represent data after injection. Note: for the position traces, flexion is up; for the velocity traces, flexion is down. A: wrist flexion with torque load opposing flexion. Position and velocity were similar before and after inactivation of the pallidurn. Flexor EMG was tonically increased after the injection. B: wrist flexion with torque load opposing extension (assisting flexion). Peak velocity was reduced and movement time was prolonged after muscimol injection. C: wrist extension with torque load opposing flexion (assisting extension). Peak velocity was reduced and movement time was increased after muscimol injection. Flexor EMG was tonically increased after the injection. D: wrist extension with torque load opposing extension. Peak velocity was slightly reduced and movement time was slightly prolonged after muscimol injection. Flexor (antagonist) EMG was increased during the movement and in the final hold after injection. This figure illustrates the greater reduction of peak velocity when the movements were made by decreasing the activity of the loaded muscle (B and C) than when made by indreasing the activity of the loaded muscle (A and D).

EXT ENSOR EMG

! b


334

J. W. MINK 1

AND

j Control

m

Muscimol

3 FOIR

:

EDIR

:

EDIR

:

FLOAD 1 *

FDIR

:

ELOAD

, -

FIG. 4. Effect of pallidal inactivation on reaction time in VisStep task performance. The bars represent the average of reaction times after all 7 muscimol injections compared with the average of reaction times from the 7 preinjection controls. The pre- and postinjection reaction times were not significantly different (t test) at the P < 0.05 level. FDIR:FLOAD, flexion with the flexors loaded. EDIR:FLOAD, extension with the flexors loaded. FDIR:ELOAD, flexion with the extensors loaded. EDIR:ELOAD, extension with the extensors loaded.

W. T. THACH

tion of globus pallidus pars externa (GPe) and extending into the dorsolateral GPi, judged from the histological lesions made later with kainic acid at the same injection site (Fig. 1). Five muscimol injections were made over 3 wk in the second monkey, two of 1 ~1 and three of 2 ~1. All five injections in the second monkey were centered in GPi, as judged from the subsequently made kainic acid lesion site (Fig. 2). After each injection, motor deficits appeared within 30 s and persisted for 7- 12 h. After one 2-~1injection in the second monkey, a slightly flexed posture (described below) persisted for 24 h, but was gone within 36 h of the injection. In all cases,quantitative measuresof task performance returned to normal by 24 h after injection. The saline injections had no effect on task performance and no noticeable effect on behavior outside of the task. The results of all muscimol injections were qualitatively similar, and the data described below are from one representative injection in the second monkey (site ~20.3 in Fig. 2). TASK PERFORMANCE. After muscimol injection into the pallidum, movements in VisStep were slower. Fig. 3 shows wrist position and velocity and EMG of wrist flexors and extensors in VisStep after muscimol injection. The slownessof movement was reflected in an increased movement time principally attributable to decreased peak velocity, without any significant accompanying change in reaction time. The mean peak velocity over all directions and loads after this injection was decreasedfrom 3 16O/sto 275”/s (t =

II

velocity were monitored during task performance stored on computer disk every second day.

every day, and

Control

300

Muscimol

-r

-I-

Histology

250

After the brains were sectioned and stained with thionin, the injection sites and lesion extent were mapped and drawn on tracings of the GP. The dentate muscimol injections in the second monkey were also mapped. The injection sites were identified from the tracks caused by insertion of the syringe needle into the brain. The extent of the kainic acid lesion was determined by 1) loss of neuronal cell bodies and 2) infiltration of glia in the area surrounding the injection site. Because the extent of the muscimol injections could not be determined histologically, it was estimated from the tissue involved after injection of an identical volume of kainic acid.

200

iz -l-i Ic’3 0 -l-l -P&z

aJ u-l

150

V

ki E 100

Data analysis 50

The pre- and postinjection data were analyzed on the PDP-11 computer. Several variables were examined, including 1) reaction time, 2) movement time, 3) movement velocity, 4) movement amplitude, and 5) EMG patterns. In addition, the manipulating, reaching, sitting, walking, standing, and climbing behaviors outside of the task were observed and tine photographed to look for deficits that might not have been revealed in the wrist movement tasks. RESULTS

Inactivation

of the GP with muscimol

Two muscimol injection of 1 ~1 each were made 1 wk apart in the first monkey, centered at the ventromedial por-

0 t=DIA

:

FLDAD

EDIR

ELOAD

EDIR

:

FDIR

:

FIG. 5. Effect of pallidal inactivation on peak velocity in VisStep task performance. Bars represent the average of peak velocities after all 7 muscimol injections compared with the average of peak from the 7 preinjection controls. Peak velocity was consistently reduced by inactivation of the pallidum when movements were made by decreasing the activity of the loaded muscle (EDIR:FLOAD and FDIR:ELOAD). FDIR:FLOAD, flexion with the flexors loaded. EDIR:FLOAD, extension with the flexors loaded. FDIR:ELOAD, flexion with the extensors loaded. FDIR:ELOAD, flexion with the extensors loaded. *Difference significant at P < 0.05; **difference significant at P < 0~0 1. Peak velocity change for flexion with flexors loaded (FDIR:FLOAD) was not significant.


PALLIDAL FLEXORS

INHIBITION

OF POSTURAL

EXTENSORS

LOADED

MECHANISMS

335

LOADED

A

WRIST POSIT ION _c__

_v8/:s=_---

-*-

---y

WRIST VELOC I TY -400

-If00

EXTENSOR EMG

WRIST POSITION

---

WRIST VELOCITY

-

+

460

IL*

;OQQ

FIG. 6. Wrist position, velocity, and EMG in VisRamp task performance before and after injection of muscimol into the pallidum. Conventions same as in Fig. 3. A: wrist flexion with torque load opposing flexion. Position and velocity were similar before and after inactivation of the pallidum. Flexor EMG was increased during the ramp after the injection. B: wrist flexion with torque load opposing extension (assisting flexion). Peak velocity at the start of the ramp was slightly reduced and the initial hold position was 5” flexed compared with normal. Flexor EMG was tonically increased after the injection. C: wrist extension with torque load opposing flexion (assisting extension). Peak velocity at the start of the ramp was slightly reduced after muscimol injection. Flexor EMG was tonically increased after the injection. D: wrist extension with torque load opposing extension. Peak velocity at the start of the ramp was very slightly increased and initial hold and ramp position were 5” flexed compared with control. Flexor and extensor EMG were tonically increased throughout the task. This figure illustrates a tonic increase EMG in both directions of movement and with both loads.

FLEXOR EMG

EXTENSOR EMG

I


336

J. W. MINK

FLEXORS

AND W. T. THACH

LOADED

EXTENSORS

LOADED

WRIST POSIT ION

FIG. 7. Wrist position, velocity, and EMG in VisSine task performance before and after injection of muscimol into the pallidurn. A and B: data were aligned on the end of the trial, always occurring 500 ms after a flexion peak (vertical line). Arrow indicates occurrence of peak flexion. Abscissa scale is in milliseconds. Solid traces represent data before injection of muscimol; dotted traces represent data after injection. Note: for the position traces, flexion is up; for the velocity traces, flexion is down. A: alternating flexion and extension with torque load opposing flexion. Velocity was similar before and after inactivation of the pallidurn. Peak flexion was 5” more flexed after injection. Flexor EMG was increased after the injection. B: alternating flexion and extension with torque load opposing extension. Velocity similar before and after injection, but the peak flexion position was 5” more flexed after inactivation of the pallidurn. Flexor and extensor EMG were increased after the injection.

3.77, df = 65, P < 0.001). The mean reaction time was not significantly changed (249 ms before injection; 250 ms after injection; t = 0.17, df = 65, N.S.). Inspection revealed that the decreased peak velocity was not equal in all movement direction and load conditions. Peak velocity was consistently more decreased when the movement was made by decreasing the activity of the loaded muscle, i.e., flexion

with the extensors loaded (322”/s before and 27O”/s after muscimol; P < 0.05) or extension with the flexors loaded (32O”/s before and 283”/s after muscimol; P < 0.05). Peak velocity was also decreased during extension against an extensor load (from 3 10 to 28O”/s; P < O.OS), but during flexion against a flexor load was not significantly different (278”/s before and 29O”/s after injection). Figures 4 and 5


PALLIDAL

FLEXORS

INHIBITION

OF POSTURAL

LOADED

MECHANISMS

EXTENSORS

LOADED

FLEX0 EMG

EXTENSOR EMG

FIG. 8. Wrist position, velocity, and EMG in SelfSine task performance before and after injection of muscimol into the pallidurn. A and B: data were aligned on peak flexion (vertical line). Arrows indicate occurrence of peak extension. Abscissa scale is in milliseconds. Solid traces represent data before injection of muscimol; dotted traces represent data after injection. Note: for the position traces, flexion is up; for the velocity traces, flexion is down. A: alternating flexion and extension with torque load opposing flexion. Average wrist position was tonically more flexed by 10” after the injection. Peak velocity and peak-to-peak amplitude were similar before and after inactivation of the pallidurn. Flexor EMG was increased after muscimol, with broader peaks and greater activity during the trough phase. B: alternating flexion and extension with torque load opposing extension. Average wrist position was tonically more flexed by 10” after muscimol injection. Peak velocity was reduced on the average by 32”/s, but peak-to-peak amplitude on many single trials was similar before and after injection. Flexor EMG was increased after the injection.

summarize all seven injections of muscimol into the pallidum of the second monkey to show that lack of effect on reaction time (Fig. 4) and the lowest velocity when relaxing the loaded muscle (Fig. 5) were consistent.

In VisRamp (Fig. 6), there was no change in reaction time or movement time after muscimol injection, but the average peak velocity of the initial steplike component of the ramp was slightly reduced from 76 to 7O”/s. As in Vis-


J. W. MINK

338 Flexors

I -1200

I -600

1

-1200

t

1

-600

Loaded

I

I

I

0

600

1200

I

f

Extensors

0

I

600

1

1200

I -1200

I

-1200

Loaded

I

1

-600

t

I

-600

0

I

t

AND W. T. THACH

0

1 600

I

600

J

1200

1

1200

FIG. 9. Wrist position and velocity in VisStep task performance before and after kainic acid ablation of the GP. Solid traces represent data before ablation; dotted traces represent data from the 24th day after ablation. Top traces in each graph represent wrist position, the bottom traces represent wrist velocity. Other conventions same as in Fig. 3. A: flexion with flexors loaded. Position was similar before and after ablation of the pallidurn. Peak velocity was slightly decreased after the lesion. B: flexion with extensors loaded. Peak velocity was reduced and movement time was prolonged after ablation. C: extension with flexors loaded. Peak velocity was reduced and movement time was increased after ablation. Final hold position (wrist extended) was 5 O more flexed after the lesion. D: extension with extensors loaded. Peak velocity was slightly reduced and movement time was prolonged after ablation. Final hold position (wrist extended) was 5” more flexed than normal. This figure illustrates the greater reduction of peak velocity when the movements were made by decreasing the activity of the loaded muscle than when made by increasing the activity of the loaded muscle.

Step, after muscimol injection, the peak velocity was generally and consistently less when the movement was made by reducing the activity of the loaded muscle, but the difference was smaller and not statistically significant (t = 1.66, df = 30, N.S.). Despite the reduced peak velocity in the first 350 ms of the ramp movement, the velocity in the remainder of the ramp trajectory was normal. In the initial hold period of both VisStep and VisRamp, there was a tendency of the monkey’s wrist position to deviate in flexion. The flexor drift was exacerbated by loading the extensors and in some trials was so large as to move the cursor out of the target window despite the monkey’s apparent effort to hold it centered. The flexor drift was reduced by loading the flexor muscles and exacerbated by loading the extensors. In VisRamp, a 5’ flexor bias was seen during the initial hold when the extensors were loaded and during the ramp of extension movements with the extensors loaded (Fig. 6). After pallidal inactivation, the performance of VisSine was near normal (Fig. 7). Peak velocity was normal and the trajectory was also little changed except for a slight (5 “) flexor bias that was most evident at peak flexion. When errors occurred, they could be attributed to hyperflexion at both peak extension and peak flexion. In SelfSine, the primary effect of muscimol was the appearance of a flexor bias (Fig. 8). The average peak-to-peak

amplitude was less by 5O when the extensors were loaded but not changed when the flexors were loaded compared with normal. The mean position about which the alternating movement was made was also flexed by loo relative to normal. The average frequency of alternation decreased from 3.6 to 1.9 Hz (47%), but the average peak-to-peak velocity was only reduced by 3O”/s (8%). After muscimol injection, the EMG exhibited a pattern of increased activity in all tasks (Figures 3 and 6-8). In some cases EMG activity was increased in both muscle groups in a pattern of cocontraction. In other cases, there was increased activity in one muscle group with no change in the maintained limb position, and cocontraction was inferred. OPEN ROOMANDCAGE BEHAVIOR. After injection ofmuscimol into the GP, monkeys assumed a characteristic, abnormal posture of adduction and flexion at the shoulder, flexion at the elbow and wrist, and extension of fingers in the arm contralateral to the injection. This posture was maintained whenever the monkeys were immobile in their cages or in the open room, but both monkeys could move out of this posture to reach for bits of fruit. The ipsilateral arm exhibited normal postures and was used in preference to the contralateral when the monkeys reached for fruit,

---0---U----

-10

-6

-2

2

Days

6

After

10

KA

14

Agonist Loaded Antagonist Loaded

16

22

26

Injection

FIG. 10. Time course of reaction time in VisStep task performance after ablation of the GP. Reaction time was monitored for 25 days after the injection of kainic acid in the pallidum of the 2nd monkey. On the 12th day the maximum allowed time to make the movement (reaction time + movement time) was increased by 100 ms. After that time, there was a gradual increase in movement time that reached statistical significance only on the 18th and 25th days. During the period of 2- 14 days after ablation, the reaction time was unchanged, but the movement time was prolonged (compare with Fig. 11). Agonist Loaded includes data from both movements that were made by increasing the activity of the loaded muscle (flexion with flexors loaded and extension with extensors loaded). Antagonist Loaded includes data from both movements that were made by decreasing the activity of the loaded muscle (flexion with extensors loaded and extension with flexors loaded).


PALLIDAL

INHIBITION

OF

POSTURAL

339

MECHANISMS

key and both injections in the first monkey, abnormal turning was not observed, but the characteristic flexed posture of the contralateral limb and task deficits were present. Muscimol injections into the pallidum also impaired gait such that, on the side contralateral to the injection, the shoulder and elbow and the knee and hip were abnormally flexed in all phasesof the quadripedal walking. Passive manipulation of the arm contralateral to the injection revealed abnormally increased resistance to both flexion and extension of the arm. The increased resistance was judged not to be proportional to length (springlike) or velocity (viscous) of stretch, but was plastic (lead-pipe) in nature.

Ablation ---U--

Antagonist Loaded

I 1 l(

3

I -8

I

I

I

-2

I

2

Days

I

I

6

After

I

I

10

KA

I

14

I

I

16

I

I

22

I

26

Injection

FIG. 11. Time course of movement time in VisStep task performance after ablation of the GP. Movement time was monitored for 25 days after the injection of kainic acid in the pallidum of the 2nd monkey. Beginning on the 2nd day, the movement time for movements made by turning off the loaded antagonist was greater than control and significantly greater than for movements made by turning on the loaded agonist. On the 12th day the maximum allowed time to make the movement (reaction time + movement time) was increased by 100 ms. After that time, there was a gradual increase in movement time for both types of movement. Agonist Loaded includes data from both movements that were made by increasing the activity of the loaded muscle (flexion with flexors loaded and extension with extensors loaded). Antagonist Loaded includes data from both movements that were made by decreasing the activity of the loaded muscle (flexion with extensors loaded and extension with flexors loaded). The difference between agonist loaded movements and antagonist loaded movements remained significant until the monkey was killed.

held on to their leashes,or climbed. Although the ipsilateral arm was preferred, the monkeys could and did make normal movements with the contralateral arm, albeit with persistence of some of the abnormal postures. For example, normal monkeys grasped raisins by pinching with the thumb and forefinger and kept the other three fingers flexed. After inactivation of the GP, raisins were grasped with a normal-looking thumb and forefinger pinch, but the third, fourth, and fifth finger were extended, as in tea-cup etiquette. The strength of the pinch seemed normal, but extended fingers sometimes interfered with manipulation of the raisin. After three of the five muscimol injections (stte YIOS.1 and 3) the second monkey displayed a tendency to circle away from the side of the injection. The turning was characterized by deviation of eyes and head and rotation of the trunk to the contralateral side. When walking on all fours in the open room, the monkey would sometimes walk in repeated circles in the direction opposite the side of the injection. The turning was not obligatory, because the monkey could maintain positions, and, in the cage, no turning was observed. After the other two injections in the second mon-

ofthe GP with kainic acid

TASK PERFORMANCE. The effects of permanent pallidal ablation were qualitatively similar to, but quantitatively greater than, the effect of temporary inactivation with muscimol. In VisStep (Fig. 9) the peak velocity was reduced and the movement time increased, especially when the movement was made by decreasing activity of the loaded antagonist muscle (Fig. 9, B and C). Reaction time was normal except for an occasional inconsistent slight increase that began on the 12th day when the time allowed to make the movement was increased to permit the monkey to per-

Flexors

Loaded

Extensors

B

.

50

300”/

set

I

0

\

300”/ I

I

I

Loaded

Y-

set

I

J

1

1

-1200

t

0 400

1200

2000

1200

t

0

400

1200

2000

-1200

t

0 400

1200

2000

-1200

t

0 400

1200

2000

Extension

FIG. 12. Wrist position and velocity in VisRamp task performance before and after kainic acid ablation of the GP. Conventions same as in Fig. 9. Solid traces represent data before ablation; dotted traces represent data from the 24th day after ablation. A: flexion with flexors loaded. Position and velocity were similar before and after ablation of the pallidurn. B: flexion with extensors loaded. Position was normal and peak velocity was reduced by 5”/s after ablation (not visible at this magnification). C: extension with flexors loaded. Peak velocity reduced by 5 O/s and the ramp trajectory was slightly more flexed after ablation. Final hold position (wrist extended) was 2” more flexed than control after injection. D: extension with extensors loaded. Peak velocity was slightly increased and movement time was prolonged after ablation. Final hold position (wrist extended) was 5” more flexed than normal after the lesion. Amplitude of a preexisting tremor (that appeared during the course of multiple electrode penetrations into the cerebellum) was somewhat reduced in this task, but it was not known whether this was a direct affect of pallidal ablation or whether it was due to recovery from slight cerebellar damage.


340

J. W. MINK

Flexors

AND

W. T. THACH

Loaded

Extensors

Loaded

0 300°/

I

-500

-250

0

250

set

I

500

-500

-250

0

250

500

FIG. 13. Wrist position and velocity in SelfSine task performance before and after kainic acid ablation of the pallidurn. Solid traces represent data before ablation; dotted traces represent data from the 24th day after ablation. Other conventions same as in Fig. 8. Flexors loaded, the average wrist position was tonically more flexed by 35 O after the injection. Average peak velocity was reduced by 15O”/s and the peak-to-peak amplitude was reduced by 18” after pallidal ablation. Extensors loaded, the average wrist position was tonically more flexed by at least 60” and often exceed the measurable range after pallidal ablation. Peak velocity was increased by 2OO”/s and peak-to-peak amplitude was increased by 10” after ablation.

form the task successfully despite the prolonged movement time (Fig. 10). In both monkeys, movement time was increased beginning the second day after kainic acid injection and persisted until the monkeys were killed (14 days postlesion for the 1st monkey and 25 days postlesion for the 2nd monkey) (Fig. 11). Performance of VisRamp (Fig. 12) after pallidal ablation was also similar to that after inactivation with muscimol. The reaction time and total ramp movement time were unchanged, and the peak velocity was reduced by 5*/s, on average, when the movement was made by decreasing the activity of the loaded antagonist muscle. When the movement was made by increasing the activity of the loaded agonist muscle, the peak velocity was normal for flexion with the flexors loaded and actually increased by loo/s for extension with the extensors loaded. The performance of VisSine (not shown) after kainic acid ablation was equivalent to that described above for muscimol inactivation of the pallidurn. After pallidal ablation with kainic acid, the position of SelfSine (Fig. 13), as measured with the manipulandum, was flexed by 35’ with the flexors loaded and by 260” with the extensors loaded relative to control. When the extensors were loaded, the peak flexion often exceeded the measurable range of position. The average frequency of alternation was reduced from 3.5 to 2.4 Hz after ablation because of several periods in which the monkey stopped moving in midtrial. When the extensors were loaded, the movement ceased to be truly “alternating” because the wrist was tonically flexed for 70% of the time of each trial. Only with seemingly great effort did the monkey make fast individual extensions against the load and just as quickly return to the maintained flexed position. In this condition, the peak velocity increased, suggesting that the reduced frequency was not due to a deficit of velocity generation. It was not possible to make accurate quantitative comparisons of EMG before and after kainic acid ablation because the variability of EMG signal amplitude from session to

session was greater than the variability of EMG during the muscimol sessions. This was thought to be due to the variability of electrode placement and skin resistance from ses-

IP

I I

Location of muscimol injections in the dentate nucleus of the FIG. 14. 2nd monkey. Five muscimol injections were made into the GP of this monkey, 1 at site 1,2 at site 2, and 2 at site 3, as represented on a horizontal section through the deep cerebellar nuclei. F, n. fastigius; IP, n. interpositus; D, n. dentatus.


PALLIDAL FLEXORS

LOADED

INHIBITION

OF POSTURAL EXTENSORS

MECHANISMS

341

LOADED B

WRIST VELOC ITY

. LJ-

f

EXTENSOR EMG

D WRIST POSITI ON

WRIST VELOC I TY

FIG. 15. Wrist position, velocity, and EMG in VisStep task performance before and after injection of muscimol into the dentate nucleus. Conventions same as in Fig. 3. A: wrist flexion with torque load opposing flexion. Position and velocity were similar before and after inactivation of dentate. After the final position was reached, the wrist position deviated toward extension. After injection of muscimol, the flexor EMG peak was reached 100 ms later than control. B: wrist flexion with torque load opposing extension (assisting flexion). Velocity and position were near normal after dentate inactivation. C: wrist extension with torque load opposing flexion (assisting extension). Peak velocity and position were unchanged after dentate inactivation. D: wrist extension with torque load opposing extension. Peak velocity was normal after muscimol injection. Final position was more variable after injection and the average peak movement amplitude was increased by 3”. After injection, the extensor EMG reached a peak 100 ms later than control. Tonic levels of EMG were not changed. Compare this figure with the effect of muscimol inactivation of the pallidum shown in Fig. 3.


342

J. W. MINK

FLEXORS

AND

W. T. THACH

LOADED

EXTENSORS

LOADED

F 45”

F 45”

E 45”

E 45”

F~[G. 16. Increased reaction time in VisStep task performance after dentate ablation. Wrist position data in these graphs were aligned on the onset of target movement (cue to move). Solid trace is control, dotted trace is postmuscimol. Prolonged reaction time is reflected in the relatively later onset of movement after ablation.

sion to session. Yet, attempts to keep electrode position consistent by marking the location on the arm, and to minimize skin resistance by scrubbing the skin with alcohol and acetone, did not sufficiently reduce the day-to-day variability to allow reliable comparisons of EMG amplitude across sessions. That cocontraction was increased after ablation of the pallidum was inferred from three observations. First, the flexor-biased posture and deviations were similar to those seen after muscimol injection, which were accompanied by cocontraction. Second, the increased plastic resistance to passive manipulation of the arm after ablation was similar to that after inactivation with muscimol. Third, as was reported for muscimol injection, the deficit after ablation with kainic acid was greater when movements were made by decreasing EMG in the loaded muscle than when increasing EMG in the loaded muscle. The temporal patterns of EMG were consistent across sessions and, like those seen after muscimol injection, were not affected by pallidal ablation. OPENROOMANDCAGEBEHAVIOR. Theopenfieldandcage behavior after pallidal ablation was similar to that after inactivation with muscimol. Specifically, the I) flexed elbow, wrist, hip and ankle postures; 2) extended third, fourth, and fifth fingers during precision pinch; 3) flexed attitude of the contralateral limbs during locomotion; and 4) preferential use of the ipsilateral arm after ablation with kainic acid were qualitatively equivalent, but quantitatively more severe than those seenwith muscimol injections. These characteristics were consistently present beginning on the second day after the kainic acid injection. During the first 48 h after kainic acid injection, the monkeys made more aggressive gestures, had more turning, and were generally *more active in the open room; but, after 48 h, the behaviors were

indistinguishable from those described above for muscimol injections, with two exceptions. First, during walking on all fours, the second monkey circumducted the leg contralatera1 to the lesion. This was not seen after muscimol injections in either monkey, nor was it seenin the first monkey after pallidal ablation. Second, beginning on the secondday after the lesion, both monkeys were unable to insert their hands into the manipulandum without assistance. They were unable to extend the elbow to fit the hand through the side of the primate chair, and were unable to adopt the posture of extended wrist, extended and adducted fingers, and flexed thumb that was necessaryto insert the hand into the wedge-shaped manipulandum.

Inactivation

of the dentate nucleus with muscimol

Five muscimol injections of 1 ~1 each were placed in the dentate nucleus of the second monkey (Fig. 14). After each injection, motor deficits appeared within 2 min of withdrawal of the syringe and persisted for 7- 12 h. In every case, quantitative measuresof task performance returned to normal by 24 h after injection. Saline injections had no effect on task performance and no noticeable effect on behavior outside of the task. The results of all five muscimol injections were qualitatively similar, and the data described below come from one representative injection (site no. 2 in Fig. 14). TASKPERFORMANCE. In VisStep, reaction time was increased, but movement time and peak velocity were normal after muscimol injection into the dentate. Figure 15 showswrist position and velocity and EMG of wrist flexors and extensors in VisStep after muscimol injection. The reac-


PALLIDAL

INHIBITION

OF

-f

:

FLOAO

EOIR

:

FLOAO

FOIR

:

ELOAO

EOIR

343

MECHANISMS

amplitude (Fig. 20). The increased amplitude was accompanied by increased EMG in the loaded muscle, but the EMG in the unloaded muscle was not increased. Peak velocity was normal. In SelfSine, the primary effect of muscimol in the dentate was to slow the frequency (Fig. 2 1). The average frequency of alternation decreased from 4.2 to 2.2 Hz, and the peakto-peak velocity decreased by 2O”/s. The variation of frequency as measured by standard deviation of the period was increased from 99.5 to 459.2 ms, an increase of 46 1Yo. The standard deviation of the amplitude was consistently increased in both load conditions by 8O (Levine’s test, P < 0.05). The average peak-to-peak amplitude was less by 5O when the flexors were loaded, but was normal when the extensors were loaded (compared with normal). The mean position about which the alternating movement was made was unchanged. No increased cocontraction was seen, and, when the extensors were loaded, the flexor EMG decreased. OPEN ROOM AND CAGE BEHAVIOR. Inactivation of the dentate nucleus produced few deficits that were obvious in room or cage behavior. Walking on all fours, climbing, and leaping into the cage all appeared normal. The monkey grasped food with either hand, showing no preference. However, when the food was manipulated before it was eaten-

Control

FDIR

POSTURAL

:

ELOAO

17. Effect of dentate inactivation on reaction time in VisStep task performance. Bars represent the average of reaction times after 5 muscimol injections compared with the average of reaction times from the 5 preinjection controls. Postinjection reaction times were increased from control for both movement directions and loads. FDIR:FLOAD, flexion with the flexors loaded. EDIR:FLOAD, extension with the flexors loaded. FDIR:ELOAD, flexion with the extensors loaded. EDIR:ELOAD, extension with the extensors loaded. *Difference is significant at the P < 0.05 level 7**difference is significant at the P < 0.0 1 level. FIG.

Control

400

Muscimol

350

300

tion time was increased by 30-50 ms (mean increase of 35 ms) in all load and direction conditions (Figs. 16 and 17). The peak velocity was not decreased and actually increased by 4O”/s during flexion with the extensors loaded (Fig. 18). No consistent reduction or increase of movement amplitude was observed, but the variability of the final hold position increased by 10%. The increased variability appeared to be due to an overshoot followed by a correction once the cursor was in the final target position. The EMG exhibited no increase in cocontraction after muscimol injection, but in extension movements with the extensors loaded, the mean EMG peak was delayed by 50 ms. In VisRamp, reaction time was also increased by 30-50 ms (mean of 37 ms) after muscimol injection, but the average peak velocity was normal. Figure 19 shows wrist position and velocity and EMG in VisRamp before and after muscimol injection. No differences of movement time, movement amplitude, or position in the initial and final holds were seen after dentate inactivation. No consistent differences in the EMG patterns were seen after muscimol injection and, in particular, no agonist-antagonist cocontraction was observed. After dentate inactivation, the performance of VisSine was near normal, except for a 5’ increase in peak-to-peak

$ -l-l h 0 0 Otll dul F>

250

al YU fui! 200,

:50-

itooFOIR

:

FLOAO

EOIR

:

FLOAO

FOIR

:

ELOAO

EOIR

:

ELOAO

FIG. 18. Effect of dentate inactivation on peak velocity in VisStep task performance. Bars represent the average of peak velocities after all 5 muscimol injections compared with the average of peak from the 5 preinjection controls. Peak velocity was not significantly changed in 3 of the 4 conditions and was increased for flexion movements with the extensors loaded (FDIR:ELOAD). FDIR:FLOAD, flexion with the flexors loaded. EDIR:FLOAD, extension with the flexors loaded. FDIR:ELOAD, flexion with the extensors loaded. FDIR:ELOAD, flexion with the extensors loaded. *Difference significant at P < 0.05.


344

J. W. MINK FLEXORS

LOADED

AND

W. T. THACH

EXTENSORS

LOADED B --

WRIST POSITI ON

WRIST VELOCITY

.

-I:00

t

-400

4B0

1206

-1’00

-400

0 -1SX7

-40a+

- L004

-400

2d00

t

460

l‘xw

2000

440

I cm

&Je

J&A

1 2'ca

2000

EXTENSOR EMG.

-LO

-40e

+

I

*Be

120b

2000

I t

C

D

I

WRIST POSIT ION

FIG. 19. Wrist position, velocity, and EMG in VisRamp task performance before and after injection of muscimol into the dentate. Conventions same as in Fig. 3. A: wrist flexion with torque load opposing flexion. Position and velocity were similar before and after inactivation of the dentate. B: wrist flexion with torque load opposing extension (assisting flexion). Position and velocity were similar before and after injection. A small-amplitude slow tremor (2-3 Hz) appeared in this task condition after dentate inactivation. C: wrist extension with torque load opposing flexion (assisting extension). Velocity and position were similar before and after muscimol injection. Extensor EMG was slightly increased in the initial hold after dentate inactivation. II: wrist extension with torque load opposing extension. Position and velocity were similar before and after injection. Preexisting tremor was slowed bY -1 Hz.

------++-A---=---T

WRIST VELOCITY

de---

I

I -1200

-4e*

480

1;*

2009

-1’00

-.i00?

4tm

lsm

:&e

6


PALLIDAL

FLEXORS

INHIBITION

OF POSTURAL

MECHANISMS

LOADED

FLEXOR EMG

FIG. 20. Wrist position, velocity, and EMG in VisSine task performance before and after injection of muscimol into the dentate. Conventions same as in Fig. 7. A: alternating flexion and extension with torque load opposing flexion. Velocity was similar before and after inactivation of the pallidurn. Peak-to-peak amplitude was reduced by 5 Oafter injection. Flexor EMG was slightly increased and more irregular after dentate inactivation. B: alternating flexion and extension with torque load opposing extension. Velocity and position were similar before and after injection. Extensor EMG was increased after the injection.

a behavior that normally involved rapid to and fro rubbing of the piece of monkey chow with one hand while holding it with the other-the hand ipsilateral to the injection was always used for holding and not for rubbing. When normal, this monkey exhibited no hand preference for holding or rubbing. No gait abnormalities were observed. One other consistent deficit resulting from inactivation of the dentate was seen when the monkey reached for bits of fruit. When raisins or bit of apple were placed on the floor

or held in the experimenter’s hand, the monkey overreached for the fruit by -5 cm. When reaching for a raisin, the hand would extend beyond the fruit and then grasp it while bringing the hand back toward the body. No tremor of the arm or hand was observed during reaching movements. Passive manipulation of the arm ipsilateral to the injection did not reveal any obvious changes in muscle tone or resistance to stretch.


346

J. W. MINK

FLEXORS

AND

W. T. THACH

EXTENSORS

LOADED

LOADED

FL EXO EMG

EXTENSOR EMG

1

+

r!ss

T&a

0 -300

- c’

cl

SS

s&a

21. Wrist position, velocity, and EMG in SelfSine task performance before and after injection of muscimol into the dentate. Conventions same as in Fig. 8. A: alternating flexion and extension with torque load opposing flexion. Average wrist position was unchanged by muscimol injection. Peak velocity and peak-to-peak amplitude were similar before and after inactivation of the pallidum, but the frequency was slower after inactivation. Flexor EMG was reduced less after muscimol injection. B: alternating flexion and extension with torque load opposing extension. Average wrist position was tonically more flexed by 5” after muscimol injection. Peak velocity was reduced by 32”/s, but peak-to-peak amplitude was similar before and after injection. Frequency of alternation was reduced. Flexor EMG was reduced after the injection. FIG.

DISCUSSION

Pallidal inactivation releases immediately a maintained cocontraction of antagonist muscles without delaying initiation of movement Inactivation and ablation of the GP produced symptoms that can be summarized as 1) cocontraction of flexors and

extensors, tonic and phasic; 2) flexor bias during maintained posture, at peak flexion during sine-tracking, and during self-paced rapid alternation; 3) slowness of steptracking movements, more so when the movement was made by turning off the antagonist muscle than when turning on the agonist, and a reduced peak velocity in the steplike component of ramp-tracking; and 4) decreased am-


PALLIDAL

INHIBITION

OF

plitude of most movements, with the notable exception of self-paced rapid alternation under extensors load. These results cannot be explained by a deficit of timing of movement initiation, because the reaction times after pallidal inactivation and ablation in step- and ramp-tracking, with few exceptions, were normal. The normal reaction time with the prolonged movement time is consistent with several previous studies of unilateral pallidal inactivation or ablation in monkeys (Amato et al. 1978; DeLong and Coyle 1979; Horak and Anderson 1984; Hore et al. 1977; Hore and Vilis 1980). The normal reaction time of monkeys with pallidal lesions is also consistent with the late timing of GP unit discharge (Anderson and Horak 1985; Georgopoulos et al. 1983; Mink and Thach 199 lb; Mitchell et al. 1987), again suggesting that the initiation of movement is not a function of the basal ganglia. The movement deficits could not be attributed to alteration of amplitude generation, per se. Although some extensions in VisStep and VisRamp were smaller than normal by 5 O,flexions were normal in size and extensions were larger than normal in SelfSine after ablation. Neither could the deficits be attributed to a faulty velocity generation per se: not all of the movements were slower. In particular, the peak velocity during 1) step-tracking flexions under flexor load, 2) ramp-tracking extension under extensor load, and 3) rapid alternation under extensor load was normal or greater than normal. Finally, the deficits could not be attributed to nonspecific effects of muscimol or kainic acid on motivation, attention, or perception: in open room behavior, although reaching movements of the arm contralateral to the injection were abnormal, those of the ipsilateral arm were normal. The results appear to reflect a preferential inability to “turn off’ previously maintained muscle activity. Consistently the largest deficits of movement time and peak velocity were seen when movements were made by “turning off” the loaded antagonist muscles to let the torque load carry the limb. When movements were made by further “turning on” already active loaded agonist muscles, the movement time and peak velocity was normal for flexions (although somewhat slower for extensions). This last observation could be explained by the predominance of tonic activity in the flexors that would interfere with all extensions but would assist flexions made by further increasing the already high flexor activity. Because the flexion movements made by turning off the loaded extensors were slower than normal, it would appear that there is an inability to turn off loaded extensors as well as flexors. This is consistent with the observation of Wilson (1928), who noted that patients with basal ganglia disease have increased tone in both agonist and antagonist, but that it was more “difficult” to relax the hypertonic antagonist than to further contract the hypertonic agonist. The deficits in VisRamp and VisSine that followed inactivation or ablation of the pallidum were subtle. Although the peak velocity in the initial steplike movement of VisRamp was reduced in a pattern similar to that seen in VisStep, the decrease was small and not statistically significant. Furthermore, the velocity of the constant-velocity component of the ramp movement and throughout the

POSTURAL

MECHANISMS

347

sine-tracking movement was not detectably decreased by pallidal lesions. The trajectories in both tasks reflected a tonic flexor bias. This was seen in VisRamp as a flexor of&et of both extension ramp trajectories and final hold positions and in VisSine as a flexor bias of the peak flexion position. The effect of pallidal injections of muscimol and kainic acid on SelfSine was primarily a flexor positional offset. The SelfSine task can be described as having two components, a static and a dynamic. The static component would be the position about which the wrist alternates, and the dynamic component would be the alternation itself. After muscimol injection, the most obvious effect was on the static component, such that the alternating movement was performed from a more flexed position than normal. The dynamic amplitude and velocity were slightly reduced by 5’ and 3O”/s, respectively, but the static center position was more flexed by loo relative to normal. After ablation of the pallidum with kainic acid, the static flexion was qualitatively similar, but more pronounced, than after inactivation with muscimol. With the flexors loaded, the amplitude and velocity were reduced ( 18 O and 15O”/s, respectively), but the center position was abnormally flexed by 35O. When the extensors were loaded, the movement ceased to be truly “alternating” because of the statistically maintained flexion (Fig. 13B). With the extensors loaded, movement amplitude and velocity were greater by 10’ and 2OO”/s, respectively, than normal, but the center position was more flexed by 60°. These observations would suggest that after pallidal ablation rapid alternation was most impaired by a static flexor bias, a cocontraction with flexors more active than extensors. In the cage and open room, after inactivation or ablation of the pallidum, monkeys adopted a posture of the contralateral arm consisting of flexion at the shoulder, elbow, and wrist, with extension of the fingers. They also exhibited tonic flexion of the contralateral knee and hip. Inactivation of the monkey’s pallidum by cooling has been reported to produce some of the same task performance deficits that were produced with muscimol and kainic acid (Hore et al. 1977; Hore and Vilis 1980). They described increased cocontraction (greater in flexors); flexor deviation during attempts to maintain a constant position, which was reduced by loads opposing flexion and increased by loads opposing extension; and slowness of movement. Their descriptions of cage and open room deficits were also similar to those reported here. They attributed the deficits produced by cooling the pallidum to an inability to achieve a proper balance between antagonistic muscle pairs (Hore and Vilis 1980). Two findings common to both this study and those of Hore et al. (1977) and Hore and Vilis (1980) are of particular interest. First, inactivation of the pallidum had little or no effect on the constant-velocity phase of ramp movements in this or their experiments (Hore et al. 1977). Second, inactivating the pallidum reduced the size and frequency of rapid alternating step-movements, particularly in the extensor direction (Hore and Vilis 1980). Without visual feedback of target and arm position, the deficit was more pronounced. However, the deficits were qualitatively


348

J. W. MINK

AND

similar with or without visual feedback, and Hore and Vilis (1980) concluded that, although visual feedback could be used to reduce the flexor bias, it could not entirely eliminate the deficit. DeLong and Coyle ( 1979) concurred that movements were impaired with and without visual guidance. Their observation was similar to the present result that the flexor bias was more pronounced in SelfSine than in tasks where visual feedback was available (VisStep, VisRamp, and VisSine). The significance of the apparent ability of monkeys to compensate partially for some pallidal deficits by the use of visual feedback is not known. The present results agree with the pallidal kainate ablation results of Horak and Anderson (1984) in that the movement time is prolonged and the reaction time is normal. Nevertheless, the mechanism that they proposed to be responsible for the slowed movement is not supported by these experiments. The EMG recordings in their study indicated that the activity of the agonist muscles was reduced in amplitude and reached its peak later than in normals. On the basis of that observation, they concluded that monkeys with pallidal lesions have impaired ability to scale the magnitude or rate of increase of agonist EMG. Their observation contrasts with the present results after muscimol inactivation of the GP in that the EMG was increased in amplitude rather than decreased in both flexors and extensors. These differences might have two causes: 1) Horak and Anderson (1984) did not require their animals actively to maintain a posture, and it is not possible to determine whether they were impaired in reducing preexisting muscle activity; and 2) their published data were taken 30 min after kainic acid injection, during which time kainic acid is thought still to have excitatory effects (McGeer and McGeer 1978). Excitation of pallidum by kainic acid could be responsible for the decreased EMG amplitude, which would complement the increased EMG amplitude after inhibition of the pallidum by muscimol. The results of pallidal inactivation and ablation are consistent with the activity of single pallidal units recorded during the same tasks. Pallidal unit discharge frequency is more often and more intensely modulated in VisStep than in VisRamp or VisSine (Mink and Thach 199 1a). Further, neurons that were related to VisRamp had the greatest activity change during the initial step-like component of the task. The ablation deficits of the reduced velocity in VisStep and the slightly reduced peak velocity in the step-like component of VisRamp [but with normal constant velocity (ramp) movements] are negative mirror images of the normal patterns of pallidal discharge recorded in these tasks. An exception is the mismatch between unit activity and ablation deficits in SelfSine: few pallidal units were phasitally active (and those only weakly) during self-paced rapid alternating movements; yet, after pallidal inactivation, these movements were impaired. It may be that, because most of the pallidal units maintained a tonic discharge during SelfSine, the deficit in this task was more static than dynamic. The peak-to-peak velocity and amplitude were near normal, but the center position, about which the alternation was performed, was flexed relative to normal. In this explanation, the effect of ablation would be the removal of the tonic inhibition of a postural mechanism, which results

W. T. THACH

in a relatively position.

normal

movement

about

an abnormal

Dentate inactivation delays initiation of volitional movement without increasing postural rigidity The increased reaction time after inactivation of the dentate is consistent with previous studies of movement after inactivation or ablation of the dentate nucleus. The earliest quantitative studies of reaction time after cerebellar injury were those of Holmes ( 19 17, 1939) who demonstrated that reaction time could be prolonged by 0.1-0.2 s in human patients with cerebellar damage. More recently, it has been shown that inactivation of the dentate by cooling in monkeys trained to perform prompt elbow movements prolonged reaction time by 50- 150 ms (Meyer-Lohmann et al. 1977). Similar results have been found after electrolytic ablation of the dentate (Spidalieri et al. 1983; Thach 1975). The increased reaction time that follows dentate ablation is consistent with the finding that dentate unit activity precedes that of motor cortex (Thach 1975, 1978) and suggests that the lateral cerebellum participates in early components of movement initiation. Because dentate inactivation prolongs reaction time and not movement time, and pallidal inactivation prolongs movement time and not reaction time (this paper and Trouche and Beaubaton 1980) it follows that the dentate more properly controls initiation per se, and the pallidum exerts its influence only after the initiation.

Mechanisms

ofakinesia,

bradykinesia,

and choreoathetosis

A recent model proposes that, in disease of the basal ganglia, pallidal output increases to cause rigidity (DeLong 1990). Our data show just the opposite-that a decrease in pallidal output causes agonist-agonist cocontraction at rest (hence, rigidity). We propose that the pallidum normally increases its output to inhibit other central neural activity that can otherwise interfere with the voluntary movement, as first proposed by Penny and Young (1983). When pallida1 output is decreased by inactivation and cannot be increased, and the other central neural activity cannot be inhibited, then that activity may dominant common nodal points such as “upper and lower” motor neurons, and give rise to abnormal rigidity, bradykinesia, and akinesia. Normally, the nodal points are shared by many programs that alternately go through or are integrated. Such programs would include those initiating the types of movement here called voluntary, long loop stretch reflexes, tonic neck reflexes, antigravity attitudinal reflexes, and many others. The difficulty that the voluntary motor initiatory programs have in getting through could result either in a slowness of getting through (bradykinesia) or, in the extreme case, of not getting through at all (akinesia-not produced here). We are aware that this is a logical inference only and that the proof would consist of showing the signals of the competing programs at the nodal points in question and their correlation in degree with that of the behavioral abnormalities that their interference is presumed to have caused. We are aware also of the logical fallacy warned against by


PALLIDAL

INHIBITION

OF

POSTURAL

MECHANISMS

349

oral and caudal nuclei of the pontine reticular complex, each of which projects to the spinal cord (Rye et al. 1988). The substantiaInigra pars reticulata also has two targets that concern us here: the superior colliculus and the ventromedial and mediodorsal thalamus, which, in turn, project to the cerebral cortical frontal eyefields (Francois et al. 1984; Jones 1981). All GPi and substantia nigra pars reticulata projections are currently thought to be GABAergic and inhibitory. The basal ganglia thus output directly or indirectly to a variety of brain stem structures, as Wilson thought, as well as to thalamus and thence to cortex. The increased muscle activity that follo ws pallidal inactivation could be produced by any or all of spinal, brain stem, and transcortical reflexes. At a spinal level, Hore and Vilis (1980) reported that a short-latency EMG responseto muscle stretch was slightly increased in amplitude after cooling the pallidurn. At a brain stem level, Denny-Brown (1962) reported that after bilateral ablation of the pallidum, monkeys had increased forced body contact responses.At a cortical level, Tanji and Kurata (1982, 1985) recorded maintained movement-anticipatory signals in supplementary motor area and hypothesized that this was the source of the similar “set ” signals in motor cortex, which ultimately facilitated the long loop reflex through motor cortex. The hypothesis was supported when they cooled the supplementary motor area, abolished the set signal in motor cortex, and delayed the long loop reflex (Tanji et al. 1985). Although there are no accounts of the study of reflexes in patients with pure pallidal lesions, there is an extensi ve literature on the reflexes of patients with Parkinson’s disease (Berardelli et al. 1983; Rothwell et al. 1983; Tatton and Lee 1975). A consistent finding is that the short latency stretch reflex (MI) is normal in latency and amplitude, but the long latency stretch reflex (M2) is increased in amplitude (Berardelli et al. 1983; Rothwell et al. 1983; Tatton and Lee 1975). It has also been reported that the magnitude of the M2 in parkinsonian patients correlates with the degree of clinitally assessed rigidity (Mortimer and Webster 1978). Marsden and others (Rothwell et al. 1983) have suggestedthat the rigidity in Parkinson’s diseasemay in fact be due to the hyperactive M2, but the relative contributions of transcortical reflexes (Evarts 1973; Evarts and Tanji 1976; Tanji What does the pallidurn normally inhibit? Does the and Evarts 1976) multisynaptic spinal reflexes (Tracey et released abnormal posture have anything in common with al. 1980), and length-sensitive, slowly conducting type II normal posture? afferents (Matthews 1981) are not known. Tanji’s observaThe present data support a pallidal role in the inhibition tions and interpretation on the cooling of the supplemenof prior muscle activity such as to allow other mechanisms tary motor area are consistent with the observations of an to generate a volitional limb movement. Where in the ner- increased long loop reflex in Parkinson’s disease. If the vous system is the inhibition directed? symptoms of Parkinson’s diseaseresult from decreasedinhiThe GPi projects to two targets that concern us here. One bition of the pallidal target in thalamus, then the released is the oral division of the ventrolateral nucleus of the thala- thalamic cells should increase firing. This could drive supmus, which, in turn, projects to medial area 6 cerebral cor- plementary motor area cells to higher firing rates and could tex: the supplementary motor area (Jones, 198 1; Parent and provide a factitious “set” drive to motor cortex, thereby De Bellefeuille 1982; Schell and Strick 1984). The supple- causing the increase in the long loop reflex. mentary motor area projects to area 4 motor cortex, the red nucleus (parvo- and magnocellular portions), “pontine re- Switchfir selection: does the inhibitory output ofthe basal ticular nuclei,” and the spinal cord through the medullary ganglia increase and/or decrease to select which motor pyramids (Jiirgens 1984). The second target of GPi is the programs operate? pedunculopontine nucleus, which has been shown in rat to, We have thus far considered only the possibility of conin turn, project to the magnocellular, gigantocellular, and flict situations, in which one pattern of activity must be

Walshe (192 1, 1927) that the behavioral deficit produced by the neurological lesion is not necessarily the inverse of the normal function . of the unlesi.oned structure. He would have cautioned that, just becausepallidal ablation produces a maintained posture that cannot be turned off, the function of pallidum is not necessarily that of turning off the posture. Yet, the function of an element in a circuit is often revealed by the practical analytic maneuver of removing it. In this case, the know ‘n connective properties of the circuit, its timing, inhibitory nature, and the results of its removal seem to warrant this tentative conclusion. The above arguments apply only to rigidity, bradykinesia, and akinesia, the pathological reductions in movement. But any general theory of basal ganglia physiology must also account for the equally common pathophysiological symptoms of mov ‘ement excess. In this regard, part of the original reason for the proposal that the basal ganglia generate normal movements was the occurrence-after lesions of caudate, putamen, an.d subthalamic nucleus-of certain involuntary movements that, but for the fact that they were unwilled, appeared normal in pattern of performance (Carpenter et al. 1950; Denny-Brown 1962; Denny-Brown and Yanigasawa 1976). Can the present hypothesis that we put forward also offer an explanation for the occurrence of these movements in basal ganglia disease?Our observations are that the muscimol inactivation of the tonic inhibitory output of the GP immediately gives rise to muscular cocontraction with a flexor bias-flexor dystonia. If the inhibitory GABAergic outputs of the caudate and putamen onto the GP were to be thrown into periodic bursts of act ivity, the waxing-waning inhibition onto the pallidum would result in a waning-waxing ( 180’ out of phase) inhibition from the pallidum onto the generator of the flexor dystonia. A waxing-waning dystonia with a period equal to that of the striatal bursts would thereby constitute chorea and athetosis. The observation that bicuculline injection causes chorea when injected into the caudate and putamen (Crossman et al. 1984) is consistent with this model. The mechanism could be confirmed (or rejected) by microelectrode recordings of the discharge of cells in caudate, putamen, and pallidum during and after the bicuculline injections.


350

J. W. MINK

AND

W. T. THACH

This work was supported by National Institutes of Health Grants ROl turned off to allow another pattern to occur. Equally imNS- 12777, NS- 15070, and T32 GM-07200 and by The McDonnell Center portant may be the combined use of “automatic” postural the Study of Higher Brain Function. mechanisms and “voluntary” mechanisms to achieve max- forAddress for reprint requests: W. T. Thach, Dept. ofAnatomy and Neuroimum power (Fukuda 196 1; Hellebrandt et al. 1956). biology, Box 8 108, Washington University School of Medicine, 660 S. When human subjects lifted weights by flexing or extending Euclid Ave., St. Louis, MO 63 110. the wrist while facing toward or away from the active arm, it was found that head direction influenced wrist strength as if Received 15 November 1989; accepted in final form 11 October 1990. determined by the tonic neck reflex (Hellebrandt et al. 1956). When subjects faced toward the extending (lifting) REFERENCES arm, wrist extension was stronger than wrist flexion. When facing away, flexion was stronger than extension. When subjects not instructed to turn their heads began to fatigue, ALEXANDER, G. E. Selective neuronal discharge in monkey putamen reflects intended direction of planned limb movements. Exp. Brain Res. they nevertheless turned their heads in the direction that 67: 623-634, 1987. would engage the tonic neck reflex to support the voluntary AMATO,G.,TROUCHE, E., BEAUBATON, D., ANDGRANGETTO, A.Therole effort. This suggests that “voluntary” movements are of the internal pallidal segment in the initiation of a goal directed movestronger and more efficient if admixed with preexisting rement. Neurosci. Lett. 9: 159- 163, 1978. flex synergies. Fukuda (196 1) observed that athletes use ANDERSON, M. E. AND HORAK, F. B. Influence of the globus pallidus on arm movements in monkeys. III. Timing of movement-related informamotor patterns that resemble the tonic neck reflexes. Pallition. J. Neurophysiol. 54: 433-448, 1985. da1 output may be one way of combining, dissociating, or BERARDELLI, A., SABRA, A. F., AND HALLETT, M. Physiological mechainhibiting these synergies (see Denny-Brown 1962). nisms of rigidity in Parkinson’s disease. J. Neural. Neurosurg. Psychiatry In this paper, we have shown that inactivation and abla46: 45-53, 1983. tion of the GP produced behavioral abnormalities that can CARPENTER, M. B., WHITTIER, J. R., AND METTLER, F. A. Analysis of choreoid hyperkinesia in the rhesus monkey. Surgical and pharmacologibe summarized as I) cocontraction of flexors and extensors, cal analysis of hyperkinesia resulting from lesions of the subthalamic tonic and phasic, with flexor bias during maintained posnucleus of Luys. J. Comp. Neural. 92: 293-33 1, 1950. ture, at peak flexion during sine-tracking, and during self- CROSSMAN, A. R., SAMBROOK, M. A., AND JACKSON, A. Experimental hemichorea/hemiballismus in the monkey. Studies on the intracerebral paced rapid alternation; 2) slowness of step-tracking movesite of action in a drug-induced dyskinesia. Brain 107: 579-596, 1984. ments, more so when the movement was made by turning DELONG, M. R. Primate models of movement disorders of basal ganglia off the antagonist muscle than when turning on the agonist, origin. Trends Neurosci. 13: 28 l-285, 1990. with reduced peak velocity in the step like component of DELONG, M. R. AND ALEXANDER, G. E. Organization of basal ganglia. In: ramp-tracking; and 3) decreased amplitude of most moveDisease of the Nervous System, edited by A. K. Asbury, G. M. McKhann, and W. I. McDonald. Philadelphia, PA: Saunders, 1986, p. ments, with the exception of self-paced rapid alternation 380-393. under extensor load. We have reviewed previous suggesDELONG, M. R. AND COYLE, J. T. Globus pallidus lesions in the monkey tions that all these phenomena result from an ongoing produced by kainic acid: histologic and behavioral effects.Appl. Neuropathological activity of neural activity that cannot be physiol. 42: 95-97, 1979. turned off by the damaged nervous system. We have dis- DELONG, M. R. AND GEORGOPOULOS, A. P. Motor functions of the basal ganglia. In: Handbook ofPhysiology. The Nervous System. Motor Concussed the previous suggestions that this ongoing neural trol. Washington DC: Am. Physiol. Sot. 198 1, sect. 1, vol. II, p. activity may have something in common with the normal 1017-1062. processes that are used in maintaining postures. DENNY-BROWN, D. The Basal Ganglia. Oxford, UK: Oxford Univ. Press, We give reasons to reject models of basal ganglia roles in 1962. DENNY-BROWN, D. ANDYANAGISAWA, N.Theroleofthebasalgangliain the exclusive control of one movement mode as distinct the initiation of movement. In: The Basal Ganglia, edited by M. D. from others, of movement initiation, and of movement paYahr. New York: Raven, 1976, p. 115- 148. rameter specification. We offer instead a model wherein the EVARTS, E. V. Motor cortex reflexes associated with learned movement. inhibitory pallidal output increases in association with Science Wash. DC 179: 501-503, 1973. EVARTS, E. V., AND TANJI, J. Reflex and intended responses in motor movement to inhibit postural holding mechanisms that cortex pyramidal tract neurons of monkey. J. Neurophysiol. 39: 1069would otherwise interfere with the movement generated by 1080, 1976. other mechanisms, including the forebrain cerebral cortex EVARTS, E. V., TERAVAINEN, H., AND CALNE, D. B. Reaction time in and the lateral cerebellum. We suggest that basal ganglia Parkinson’s disease. Brain 104: 167- 186, 198 1. inhibition may be a general mechanism not only for switch- FILION, M. Effectsof interruption of the nigrostriatal pathway and of dopaminergic agents on the spontaneous activity of globus pallidus neurons ing “off’ one activity to allow another to operated unopin the awake monkey. Brain Res. 178: 425-44 1, 1979. posed, but also for combining in appropriate blends the FRANCOIS,C.,PERCHERON,G.,ANDYELNIK, J.Localization ofnigrostriaactivities of coactive motor programs. We attempt to extal, nigrothalamic and nigrotectal neurons in ventricular coordinates in plain how these normal control functions might have been macaques. Neuroscience 13: 6 l-76, 1984. mistaken for others in prior clinical, animal ablation, and FUKUDA, T. Studies on human dynamic postures from the viewpoint of postural reflexes. Acta Oto-Laryngol. 16 1: l-52, 196 1. single-unit studies. Finally, we outline a pathophysiological GEORGOPOULOS,A.P., DELONG, M.R., ANDCRUTCHER, M.D.Relation mechanism for the disorders of posture and movement that between parameters of step-tracking movements and single cell disresult from basal ganglia disease. charge in the globus pallidus and subthalamic nucleus of the behaving The authors thank C. Fineberg, G. Perry, M. Horne, and M. Freeman for assistance with computer and software design and development. We also express our appreciation to the Medical Scientist Training Program and Washington Universitv.

monkey. J. Neurosci. 3: 1586-1598, 1983. HELLEBRANDT, F. A., HOUTZ, S. J., PARTRIDGE, M. J., AND WALTERS, C. E. Tonic neck reflexes in exercises of stress in man. Am. J. Phys. Med. 35: 144-159,

1956.

HIKOSAKA, 0. AND WURTZ, R%.H. Visual and oculomotor functions of


PALLIDAL

INHIBITION

OF

monkey substantia nigra pars reticulata. IV. Relation of substantia nigra to superior colliculus. J. Neurophysiol. 49: 1285-130 1, 1983. HOLMES, G. The symptoms of acute cerebellar injuries due to gunshot injuries. Brain 40: 461-535, 1917. HOLMES, G. The cerebellum of man. Brain 62: l-30, 1939. HORAK, F. B. AND ANDERSON, M. E. Influence of globus pallidus on arm movements in monkeys. I. Effects of kainic acid-induced lesions. J.

Neurophysiol. 52: 290-304, 1984. HORE, J., MEYER-L• HMANN, J., AND BROOKS, V. B. Basal ganglia cooling disables learned arm movements of monkeys in the absence of visual guidance. Science Wash. DC 195: 584-586, 1977. HORE, J. AND VILIS, T. Arm movement performance during reversible basal ganglia lesions in the monkey. Exp. Brain Res. 39: 2 17-228, 1980. JONES, E. G. Functional subdivision and synaptic organization of the mammalian thalamus. Int. Rev. Physiol. 25: 173-245, 198 1. J~~RGENS, U. The efferent and afferent connections of the supplementary motor area. Brain Res. 300: 63-8 1, 1984. KENNARD, M. A. Experimental analysis of the functions of the basal ganglia in’ monkeys and chimpanzees. J. Neurophysiol. 7: 127-148, 1944. MARSDEN, C. D. The mysterious motor function of the basal ganglia: the Robert Wartenberg Lecture. NeuroZogy 32: 5 14-539, 1982. MATTHEWS, P. B. C. Muscle spindles: their message and their fusimotor supply. In: Handbook of Physiology. The Nervous System. Motor Control. Washington, DC: Am. Physiol. Sot., 198 1, sect. 1, vol. II, p. 189-228. MCGEER, E. G. AND MCGEER, P. L. Kainic Acid as a Tool in Neurobiology. New York: Raven, 1978. MEYER-L• HMANN, J., HORE, J., AND BROOKS, V. B. Cerebellar participation in generation of prompt arm movements. J. Neurophysiol. 40: 1038-1050, 1977. MINK, J. W. AND THACH, W. T. The effect of pallidal inactivation on trained wrist movements in the monkey. Sot. Neurosci. Abstr. 13: 982, 1987. MINK, J. W. AND THACH, W. T. Basal ganglia motor control. I. Nonexclusive relation of pallidal discharge to five movement modes. J. Neurophysiol. 65: 273-300, 1990a. MINK, J. W. AND THACH, W. T. Basal ganglia motor control. II. Late pallidal timing relative to movement onset and inconsistent pallidal coding of movement parameters. J. Neurophysiol. 65: 30 l-329, 1990b. MITCHELL, S. J., RICHARDSON, R. T., BAKER, F. H., AND DELONG, M. R. The primate globus pallidus: neuronal activity related to direction of movement. Exp. Brain Res. 18: 49 l-505, 1987. MORTIMER, J. A. AND WEBSTER, D. D. Relationships between quantitative measures of rigidity and tremor and the electromyographic responses to torque load perturbations in unselected normal subjects and parkinsonian patients. In: Progress in Clinical Neurophysiology. Cerebral Motor Control in Man: Long Loop Mechanisms, edited by J. E. Desmedt. Basel: Karger, 1978, p. 342-360. PARENT, A. AND DE BELLEFEUILLE, L. Organization of efferent projections from the internal segment of globus pallidus in primate as revealed by fluorescence retrograde labelling method. Brain Res. 245: 20 l-2 13, 1982. PENNY, J. B. AND YOUNG, A. B. GABA as the pallidothalamic neurotransmitter: implications for basal ganglia function. Brain Res. 207: 195- 199, 1981. PENNY, J. B. AND YOUNG, A. B. Speculations on the functional anatomy of basal ganglia disorders. Annu. Rev. Neurosci. 6: 73-94, 1983. PULLMAN, S., WATTS, R., JUNCOSE, J., CHASE, T., AND SANES, J. Dopaminergic effects on simple and choice reaction time performance in Parkinson’s disease. Neurology 38: 249-254, 1988. RANSON, S. W. AND BERRY, C. Observations on monkeys with bilateral

POSTURAL

MECHANISMS

351

lesions of the globus pallidus. Arch. Neural. Psychiatry 46: 504-508, 1941. RICHTER, R. Degeneration of the basal ganglia in monkeys from chronic carbon disulphide poisoning. J. Neuropathol. Exp. Neural. 4: 324-353, 1945. ROLLS, E. T. Information processing and basal ganglia function. In Hierarchies in Neurology, edited by C. Kennard and M. Swash. London: Springer-Verlag, 1989, p. 137- 138. ROTHWELL, J. C., OBESO, J. A., TRAUB, M. M., AND MARSDEN, C. D. The behavior of the long-latency stretch reflex in patients with Parkinson’s disease. J. Neurol. Neurosurg. Psychiatry 46: 35-44, 1983. RYE, D. B., LEE, H. J., SAPER, C. B., AND WAINER, B. H. Medullary and spinal efferents of the pedunculopontine tegmental nucleus and adjacent mesopontine tegmentum in the rat. J. Comp. Neural. 269: 3 15341, 1988. SCHELL, G. R. AND STRICK, P. L. The origin of thalamic inputs to the arcuate premotor and supplementary motor areas. J. Neurosci. 4: 539560, 1984. SCHWARZ, M., SONTAG, K.-H., AND WAND, P. Non-dopaminergic neurones of the reticular part of substantia nigra can gate static fusimotor action onto flexors in cat. J. Physiol. Land. 354: 333-344, 1984. SPIDALIERI, G., BUSBY, L., AND LAMARRE, Y. Fast ballistic arm movements triggered by visual, auditory, and somesthetic stimuli in the monkey. II. Effects of unilateral dentate lesion on discharge of precentral cortical neurons and reaction time. J. Neurophysiol. 50: 1359-l 379, 1983. TANJI, J. AND EVARTS, E. V. Anticipatory activity of motor cortex neurons in relation to direction of an intended movement. J. Neurophysiol. 39: 1062-1068, 1976. TANJI, J. AND KURATA, K. Comparison of movement-related activity in two cortical motor areas of primates. J. Neurophysiol. 48: 633-653, 1982. TANJI, J. AND KURATA, K. Contrasting neuronal activity in supplementary and precentral motor cortex of monkeys. I. Response to instructions determining motor responses to forthcoming signals of different modalities. J. Neurophysiol. 53: 129-141, 1985. TANJI, J., KURATA, K., AND OKANO, K. The effect of cooling of the supplementary motor cortex and adjacent cortical areas. Exp. Brain Res. 60: 423-426, 1985. TATTON, W. G. AND LEE, R. G. Evidence for abnormal long-loop reflexes in rigid parkinsonian patients. Brain Res. 100: 67 l-676, 1975. THACH, W. T. Timing of activity in the cerebellar dentate nucleus and cerebral motor cortex during prompt volitional movement. Brain Res. 169: 168-172, 1975. THACH, W. T. Correlation of neural discharge with pattern and force of muscular activity, joint position, and direction of the intended movement in motor cortex and cerebellum. J. Neurophysiol. 4 1: 654-676, 1978. TRACEY, D. J., WALMSLEY, B., AND BRINKMAN, J. ‘Long-loop’ reflexes can be obtained in spinal-monkeys. Neurosci. Lett. 18: 59-65, 1980. TROUCHE, E. AND BEAUBATON, D. Initiation of a goal-directed movement in the monkey. Exp. Brain. Res. 40: 3 1 l-32 1, 1980. U&o, M. AND YOSHIDA, M. Monosynaptic inhibition of thalamic neurons produced by stimulation of the pallidal nucleus in cats. Brain Res. 99: 377-380, 1975. WALSHE, F. M. R. On disorders of movement resulting from loss of postural tone, with special reference to cerebellar ataxy. Brain 44: 539-556, 1921. WALSHE, F. M. R. Discussion on “The significance of the voluntary element in the genesis of cerebellar ataxy.” Brain 50: 377-385, 1927. WILSON, S. A. K. Modern Problems in Neurology. London: Arnold, 1928.


Basal ganglia motor control. II. Late pallidal timing relative to movement onset and inconsistent pallidal coding of movement parameters.

The vestibular-basal ganglia connection: balancing motor control.

Movement disorders--limb movement and the basal ganglia.

Motor thalamus integration of cortical, cerebellar and basal ganglia information: implications for normal and parkinsonian conditions.

Action, time and the basal ganglia.

Opponent and bidirectional control of movement velocity in the basal ganglia.

Trigeminal-basal ganglia interaction: control of sensory-motor gating and positive reinforcement.

Effects of marijuana on human reaction time and motor control.

Basal ganglia--possible role in motor coordination and learning.

The cortico-pallidal projection: an additional route for cortical regulation of the basal ganglia circuitry.

Basal ganglia network by constrained spherical deconvolution: a possible cortico-pallidal pathway?

Movement Disorders Following Cerebrovascular Lesion in the Basal Ganglia Circuit.

Primate models of movement disorders of basal ganglia origin.

Two-phase model of the basal ganglia: implications for discontinuous control of the motor system.

Reward Based Motor Adaptation Mediated by Basal Ganglia.

Overlap of movement planning and movement execution reduces reaction time.

Time representation in reinforcement learning models of the basal ganglia.

Idiopathic Basal Ganglia Calcification Presented with Impulse Control Disorder.

Motor control mechanisms underlying human movement reproduction.

Variability in action: Contributions of a songbird cortical-basal ganglia circuit to vocal motor learning and control.

Is the reaction time-movement time relationship 'essentially zero'?

Movement deficits and neuronal loss in basal ganglia in TRPC1 deficient mice.

Preparation for movement in the cat. II. Unit activity in the basal ganglia and thalamus.

Movement disorders in astrocytomas of the basal ganglia and the thalamus.