The long-term effects of the neurotoxin 1-trichloromethyl-1,2,3,4-tetrahydro-beta-carboline (TaClo) on cognitive performance in rats.

The neurotoxin 1-trichloromethyl-1,2,3,4-tetrahydro-beta-carboline (TaClo) has been reported, both in vitro and in vivo models, to produce neurodegeneration and parkinsonian symptoms after prolonged exposure in rats. The aim of the present study was to investigate the effects of TaClo on the cognitive performance of rats. We used the COGITAT hole board system where rats can find hidden pellets by exploring the board. TaClo-treated rats found as many pellets as control rats treated with saline. Furthermore, their search was as efficient as that of control animals since there were no differences between the groups regarding explorative activity, visits to non-baited holes and time needed to find the pellets. These results suggest that there is no deficit in spatial memory following the chronic administration of TaClo to rats.

Cerebral oligemia and iron influence in cerebral structures--element of morbus parkinson models?

The consequences of short phases of restricted cerebral blood flow and iron enrichment of striatal tissues resulted in an animal model that could correspond to the basic features of a model for Parkinson's disease. An automatic and computerized hole-board offers simultaneous data on learning and cognitive memory capabilities, learning of distinct patterns of distributed food pellets found and eaten in a given time, switches between different locations of food in the holes and in different layout patterns. Wistar rats after 60 min of bilateral clamping of the carotid arteries (BCCA) under pentobarbital anesthesia received 1.5 microg FeCl3 injected one week after BCCA unilaterally into the ventrolateral striatum. The experiments showed that reduced cerebral blood flow and increased iron within the striatal tissue had the effect of retarding reactions. Rats after BCCA and iron need 180 s to find pellets deep inside holes that are distributed in a distinct pattern. During only 60 or 30s BCCA plus iron rats are no longer able to find the same number of pellets as over 180s. Rats after BCCA plus NaCl do not show such reduced success. These results point to the idea that cerebral oligemia and increased iron in the striatum stimulate the pathological symptoms of Parkinson's disease which need also more time to have reaction and success (see Fig. 5). The data covering abbreviated time-spans show how heavily the BCCA + Fe animals are dependent on longer times.

Unusual interactions of excitatory amino acid receptor agonists: alpha- and beta-kainate antagonize motor responses to N-methyl-D-aspartate in rodents.

The alpha- and beta-stereoisomers of kainate correspond sterically to the L- and D-isomers of glutamate. Alpha-Kainate is a potent excitant at a specific membrane receptor site (kainate receptor). Beta-Kainate has been proposed as a functional N-methyl-D-aspartate antagonist in vivo. Because of the structural similarities between the alpha- and beta-stereoisomers of kainate we have investigated the interactions of both compounds with N-methyl-D-aspartate-mediated excitation in two well established animal models for assessing the action of excitatory amino acids and their antagonists in vivo: determination of CD50 (convulsant dose) for myoclonic seizures in mice and electromyographic measurement of muscle tone in genetically spastic rats. We find that alpha-kainate and beta-kainate produce myoclonic seizures in mice when given intracerebroventricularly and increase the muscle tone in genetically spastic rats when given intrathecally. Alpha-Kainate is about 5000 times more potent than beta-kainate as a convulsant and about 1000 times more active than beta-kainate in increasing the muscle tone. The excitatory actions of alpha-kainate and of beta-kainate are blocked by gamma-D-glutamylaminomethylsulphonate, a preferential kainate/quisqualate antagonist, but not by (+/-)-2-amino-7-phosphonoheptanoate, a specific N-methyl-D-aspartate antagonist. Surprisingly, alpha-kainate and beta-kainate antagonize the myoclonic seizures and the increase in muscle tone produced by N-methyl-D-aspartate, and potentiate both the anticonvulsant and myorelaxant actions of (+/-)2-amino-7-phosphonoheptanoate. Quisqualate induces myoclonic seizures in mice after intracerebroventricular application and increases muscle tone in genetically spastic rats following intrathecal injection.(ABSTRACT TRUNCATED AT 250 WORDS)

Evaluation of the muscle relaxant properties of a novel beta-carboline, ZK 93423 in rats and cats.

The muscle relaxant action of ZK 93423 (6-benzyloxy-4-methoxymethyl-beta-carboline-3-carboxylate ethyl ester), a novel beta-carboline with agonistic properties at the benzodiazepine receptor, was examined by assessing its effect on the tonic electromyogram (EMG) activity of the gastrocnemiussoleus (GS) muscle of genetically spastic rats and on ventral root reflexes, presynaptic inhibition and fusimotor activity in the spinal cord of decerebrate cats. ZK 93423 (0.1-10.0 mg kg-1) depressed the tonic EMG activity in mutant rats in a dose-dependent manner. This effect was reversed by the benzodiazepine antagonist Ro 15-1788 (5.0 mg kg-1). Both ZK 93423 (0.5 mg kg-1) and diazepam (0.3 mg kg-1) enhanced the presynaptic inhibition of the GS muscle and associated dorsal root potentials in decerebrate cats in an almost identical manner. The actions of both drugs were reversed by Ro 15-1788 (5.0 mg kg-1). ZK 93423 (0.5 mg kg-1) and diazepam (0.3 mg kg-1) depressed the activity of both static and dynamic fusimotor neurones, as deduced from changes in the afferent responses of muscle spindle primary endings to low frequency (1 s-1) sinusoidal stretching. The depressant action of diazepam (0.3 mg kg-1) was weaker than that of ZK 93423 (0.5 mg kg-1). The actions of both drugs were reversed by Ro 15-1788 (5.0 mg kg-1). ZK 93423 (0.5 mg kg-1) failed to alter the magnitude of monosynaptic ventral root reflexes evoked by electrical stimulation of a flexor and an extensor nerve, whereas diazepam (0.3 mg kg-1) had a depressant effect on both types of monosynaptic reflexes, which was antagonized by Ro 15-1788 (5.0 mg kg-1). Neither ZK 93423 (0.5 mg kg-1) nor diazepam (0.3 mg kg-1) depressed polysynaptic ventral root reflexes evoked by electrical stimulation of a flexor and a cutaneous nerve. The present results demonstrate that ZK 93423 exerts a potent muscle relaxant action due to a specific interaction with benzodiazepine receptors. However, its profile of actions on spinal motor mechanisms is not identical to that of diazepam.

Altered pattern of immunohistochemical staining for glial fibrillary acidic protein (GFAP) in the forebrain and cerebellum of the mutant spastic rat.

The spastic rat is a neurological mutant of the Han-Wistar strain with prominent spasticity, tremor, and ataxia. Neurodegeneration is found in the CA3 sector of the hippocampus and in Purkinje cells of the cerebellum. We examined the forebrain and cerebellum of spastic rats for glial reactions by using immunolabelling for the astrocytic marker, glial fibrillary acidic protein (GFAP). First, a map of the GFAP-distribution was made representing a systematic series of frontal sections in controls. Reactive astrocytes with increased GFAP should occur in the areas with established neuronal degeneration, but they could also demarcate further regions with pathology in this rat strain. Since the baseline levels of GFAP-immunoreactivity differ between brain regions, control rats and clinically normal littermates served as controls to judge relative increases in major structures. In the CA3 sector and hilus of the dorsal hippocampus, a massive gliosis was detected. In the cerebellum, a patchy increase of GFAP labelling in Bergmann glia was found. Further increases of GFAP-labelling in reactive astrocytes occurred in fiber tracts, the ventral thalamic nuclei, medial geniculate nuclei, pontine region and optic layer of the superior colliculus. Inconsistent changes were noted in cortex and pallidum. No defects of glial labelling or malformations in glial architectonics were found. The reactive changes of astroglial cells in hippocampus and cerebellum are in proportion to the neuronal degeneration. The glial reactions in the other brain regions possibly reflect a reaction to fiber degeneration and incipient neuronal degeneration or functional alterations of glial cells in response to neuronal dysfunction.

Transient occlusion of rat carotid arteries increases formation of inositol phosphate. Evidence for a specific effect on ?(1)-receptors.

Rats which had both carotid arteries occluded (BCCA) for 24 min did not suffer histological damage to hippocampal tissue. However, BCCA treatment followed by transient normobaric hypoxia (reduced oxygen) for 30 min caused damage to the pyrimidal cells in the hippocampus. Intraventricular injection of [(3)H]inositol followed by BCCA + hypoxia treatment caused an increase in the release of [(3)H]inositol-mono-phosphate ([(3)H]Ins-1-P) and [(3)H]inositol-bis-phosphate ([(3)H]Ins-2-P) in both groups (BCCA and BCCA + hypoxia) but not in sham controls. These results suggest that BCCA or oxygen reduction stimulates phosphoinositide metabolism, causing a release of [(3)H]inositol triphosphate ([(3)H]Ins-3-P). Agonist-induced stimulation of inositol phosphates (InsPs) was also analysed in BCCA animals 3 and 14 days after their operations. Noradrenaline produced an increase in Ins-Ps in the hippocampus (but not in the frontal cortex) 14 days after BCCA. This effect did not occur for sham controls nor for carbachol and quisqualate stimulation. The results show a delayed effect of the ?(1)-receptors in the hippocampus of BCCA animals.

Effect of transient reduction of cerebral blood flow on membrane anisotropy and lipid peroxidation in different rat brain areas.

Light-microscopical studies revealed that oligemic hypoxia for 24 and 60 min as produced by bilateral clamping of the carotid arteries (BCCA) in normotension does not produce neuronal cell necrosis in the vast majority of rat brain. Less than 5% of cases showed a pattern of mild selective neuronal necrosis as would be expected in ischemia. However, significant changes in both lipid peroxidation (as measured by MDA formation) and membrane anisotropy (measured by DPH or TMA-DPH, respectively, as a fluorescence probe) in cortical and striatal, but not in hippocampal, membrane fractions could be measured in ex vivo studies. Twenty-four and 60 min of BCCA without reperfusion decreased lipid peroxidation in the cerebral cortex but not in the striatum. BCCA, either for 24 or 60 min, and 60 min of reperfusion produced no changes in lipid peroxidation in either structure. However, 24 and 60 min of BCCA followed by 14 days of reperfusion led to a significant increase in MDA formation in the striatum, while lipid peroxidation in the cortex was only increased after 60 min of BCCA. Cortical as well as striatal membrane anisotropy increased significantly 14 days later in rats submitted to BCCA for 24 or 60 min. The study shows an increased lipid peroxidation 2 weeks after a transient reduction in cerebral blood flow although no neuronal necrosis could be observed in general.

Stereotypy, locomotor and cataleptic effects produced by drugs influencing dopaminergic systems in a mutant strain of Wistar rats: a genuine model of basal ganglia dysfunction?

The mutant strain of Wistar rats carrying an autosomal recessive gene defect is characterized by a sequence of progressively developing behavioural alterations including hyperexcitability, tremor, olfactory and gustatory automatisms, bradykinesia, ataxia, rigidity, paresis and cachexia. The stereotypy and locomotor responses to increasing doses of apomorphine hydrochloride and D-amphetamine sulphate, and the catalepsy response to increasing doses of haloperidol were studied in mutant rats at the age of 6-7 weeks. In the mutants, both the stereotypy and locomotor responses to amphetamine were enhanced, while stereotypy and locomotor effects induced by apomorphine were unaltered. The cataleptic response to haloperidol was significantly diminished compared to controls. These findings indicate a derangement in the function of basal ganglia in the mutants.

Caudate nucleus and programming behaviour in cats: role of dopamine in switching motor patterns.

Cats were trained to walk on a specially designed treadmill: the cats were able to collect food pellets by switching motor patterns with or without the help of exteroceptive stimuli inherent to the treadmill. To study the involvement of the caudate nucleus in switching motor patterns cats received intracaudate bilateral injections of haloperidol. In addition, in a final series of experiments, EMG recordings of two antagonistic muscles, together with recordings of characteristic changes in the length of one muscle, were made before and after the haloperidol treatment. Haloperidol treatment resulted in a decreased number of motor patterns which were not directed by exteroceptive stimuli (non-exteroceptively directed motor patterns). This haloperidol-induced effect was dose-dependently counteracted by the additional intracaudate injections of apomorphine which per se remained ineffective. Haloperidol neither altered the number of food collecting attempts nor reduced the number of exteroceptively directed motor patterns. Furthermore, haloperidol did not affect the capacity to switch to proprioceptively directed motor patterns. Finally, haloperidol did not produce abnormalities in EMG and length signals recorded from hindlimb muscles. It is concluded that haloperidol selectively reduced the animal's capacity to 'programme non-stimulus directed motor behaviour'. The data are discussed in view of their significance for therapy of patients with basal ganglia disorders, such as patients suffering from Parkinson's disease.

Phaclofen antagonizes the depressant effect of baclofen on spinal reflex transmission in rats.

The action of phaclofen, the phosphonic acid derivative of baclofen, on baclofen-induced suppression of spinal reflex transmission was tested in anaesthetized rats. Intrathecal (i.th.) injection of phaclofen, 100 nmol, antagonized the depressant effect of baclofen, 2 nmol, on spinal Hoffmann (H)-reflexes and polysynaptic flexor reflexes but ha on the action of muscimol, 20 nmol. The antagonistic effect of phaclofen on baclofen-induced depression of H-reflexes was dose-dependent in doses ranging from 1 to 100 nmol. When administered alone, phaclofen, 100 nmol, was devoid of stimulatory or depressant effects on spinal reflexes. These results indicate that phaclofen specifically antagonizes the reflex suppressant action of baclofen. The lack of intrinsic action of phaclofen suggests that there is no endogenous tonic inhibition mediated by GABAB receptors under the present experimental conditions.

Paradoxical convulsant action of a novel non-competitive N-methyl-D-aspartate (NMDA) antagonist, tiletamine.

Intracerebroventricular (i.c.v.) injection of tiletamine, 0.001 mumol, a presumed non-competitive antagonist of N-methyl-D-aspartate (NMDA) receptors, protected mice from convulsions induced by NMDA and quinolinate, but not from those induced by excitatory amino acids interacting preferentially with non-NMDA receptors. At higher doses, however, tiletamine induced convulsions by itself. Tiletamine-induced convulsions were antagonized by the broad spectrum excitatory amino acid antagonist, gamma-D-glutamylamino-methylsulphonate (gamma-D-GAMS), and were potentiated by the competitive NMDA antagonist, 2-amino-7-phosphonohepatanoate (AP7). Intrathecal (i.t.) injection of tiletamine, 0.01-1.0 mumol, dose-dependently suppressed spinal flexor reflexes. Tiletamine, 0.01 and 0.1 mumol, failed to affect spinal Hoffman- (H-) reflexes, whereas tiletamine, 1.0 mumol, led to a 50% increase of the H-reflex amplitude. It is concluded that the anticonvulsant and reflex suppressant action of tiletamine are due to antagonism of NMDA receptor-mediated excitation. The convulsant effect of tiletamine and its excitatory effect on spinal H-reflexes at higher doses, however, appear to be mediated by non-NMDA receptors.

Behavioral actions of baclofen in the rat ventromedial thalamic nucleus: antagonism by delta-aminovalerate.

Baclofen, an agonist at GABAB receptors injected locally into the rat ventromedial thalamic nucleus (VM) induced catalepsy in a dose-dependent, stereospecific and site-specific way. Baclofen-induced catalepsy was accompanied by tonic activity in the electromyogram (EMG) recorded from the gastrocnemius muscle, which is considered to be a measure of limb rigidity. delta-Aminovalerate (delta-AVA) coadministered with baclofen into the VM prevented the development of both catalepsy and limb rigidity. delta-AVA injected alone into the VM was devoid of depressant or stimulating behavioral effects. In contrast to delta-AVA, bicuculline, a GABAA antagonist, failed to antagonize baclofen-induced catalepsy. On the other hand, catalepsy due to local injection of a GABAA agonist, muscimol into the VM was antagonized by bicuculline, but not by delta-AVA. The present results suggest that delta-AVA may be useful as an antagonist of central pharmacological actions of baclofen.

Motor actions of excitatory amino acids and their antagonists within the rat ventromedial thalamic nucleus.

The present study investigates the role of excitatory amino acid receptors within the rat ventromedial thalamic nucleus (VM) for the mediation of motor behaviour. For this purpose changes of motility were monitored after microinjections of excitatory amino acids and of various excitatory amino acid antagonists into the VM. N-Methyl-D-aspartate (NMDA) and kainate (KA), but not quisqualate (QA), led to a dose-dependent increase of exploratory activity. A specific NMDA antagonist (-)-2-amino-7-phosphonoheptanoate (-)-AP7), preferential non-NMDA antagonists, 1-(p-chlorobenzoyl)-piperazine-2,3-dicarboxylate (pCB-PzDA) and gamma-D-glutamylaminomethylsulphonate (GAMS), and a broad spectrum antagonist, kynurenate (KYN), induced catalepsy in a dose-dependent manner. The catalepsy induced by (-)-AP7 was antagonized by NMDA, but not by KA, the pCB-PzDA-induced catalepsy was blocked by KA, but not by NMDA and the KYN-induced catalepsy was reversed by either NMDA or KA. These data point to a role of both NMDA and KA receptors within the VM for the regulation of motility.

Substantia nigra and entopeduncular nucleus: supraspinal sites of the muscle relaxant action of tizanidine.

Microinjection of tizanidine into the substantia nigra pars reticulata or entopeduncular nucleus reduces muscle tone in genetically spastic rats. The effect of tizanidine is related to alpha 2-adrenergic mechanism since yohimbine, an alpha 2-adrenergic antagonist, and not prazosin, an alpha 1-adrenergic antagonist, attenuates the muscle relaxation produced by the drug. These results signify basal ganglia output stations as possible sites whereby tizanidine acting via alpha 2-adrenergic mechanism exerts its muscle relaxant action.

Substantia nigra and motor control in the rat: effect of intranigral alpha-kainate and gamma-D-glutamylaminomethylsulphonate on motility.

Bilateral microinjections of an excitatory amino acid, alpha-kainate (KA), 5-50 ng, into the substantia nigra pars reticulata (SNR) result in an increase in the muscle tone and catalepsy in rats. The preferential KA/quisqualate antagonist, gamma-D-glutamylaminomethylsulphonate (gamma-D-GAMS), 10 micrograms, blocks the actions of KA, 25 ng, when coadministered into the SNR. The chemical lesion of the caudate-putamen with 6-hydroxydopamine (6-OHDA) does not affect either increases in the muscle tone or catalepsy produced by KA, 25 ng, from the SNR. The lesion of the caudate-putamen with ibotenate moderately enhances the effect of KA, 25 ng, on the muscle tone. Microinjections of KA, 25 ng, into the substantia nigra pars compacta (SNC) do not increase the muscle tone and lead to significantly less pronounced catalepsy relative to that observed following the injections of KA into the SNR. Unilateral microinjections of KA, 10-50 ng, into the SNR elicit ipsilateral turning in rats in a dose- and time-dependent manner. Unilateral application of gamma-D-GAMS, 1-10 micrograms, into the SNR produces contralateral turning. The turning evoked by KA, 25 ng, or gamma-D-GAMS, 10 micrograms, is affected neither by 6-OHDA nor by ibotenate lesion of the caudate-putamen. These results demonstrate that excitatory neurotransmission in the substantia nigra participates in the regulation of the muscle tone and posture in rats.

Transient occlusion of rat carotid arteries decreases susceptibility to pilocarpine seizures.

Bilateral clamping of the carotid arteries (BCCA) for 24 min protects animals against convulsive effects of intraperitoneal injection of 300 mg/kg of pilocarpine when administered 14 days after surgery. Electrographic recordings from hippocampus, frontal cortex, striatum and substantia nigra demonstrate that (1) no spontaneous spiking activity is caused by the BCCA when observed for up to 10 days after surgery, (2) spiking activity can only be recorded in limbic structures but not in the frontal cortex after systemic administration of pilocarpine. These observations suggest that the propagation of seizure activity from subcortical limbic structures to the neocortex is affected by BCCA.

Muscular rigidity and delineation of a dopamine-specific neostriatal subregion: tonic EMG activity in rats.

In order to investigate the role of neostriatal dopamine receptors in muscular rigidity we have studied catalepsy and spontaneous muscle tone in rats receiving haloperidol injections into various parts of the neostriatum. It was found that low doses of haloperidol (250-750 ng/0.5 microliter) induced a tonic activity in the EMG of the gastrocnemius-soleus muscle when injected into the most rostral part of the neostriatum (A 8620-9650). This tonic EMG activity was found to be both dose-dependent and dopamine-specific: the haloperidol effect of 500 ng could be inhibited by 500 ng apomorphine. No muscular rigidity could be observed when haloperidol (500 ng/0.5 microliter) was injected further caudally into the neostriatum or into the medial part of the nucleus accumbens. With regard to the haloperidol-induced catalepsy as measured by means of the bar test, a similar distribution of effective and ineffective injection sites was observed; however, higher doses of haloperidol (1.5-2.5 micrograms/0.5 microliter) were required. Thus catalepsy and muscular rigidity were found to be closely related phenomena within the neostriatum. Finally, it is discussed how information relevant to tonic EMG activity is transmitted from the substantia nigra pars compacta to the superior colliculus and the ventromedial thalamus via the neostriatum and the substrate nigra pars reticulata.

Disturbed GABAergic transmission in mutant Han-Wistar rats: further evidence for basal ganglia dysfunction.

A mutant strain of Wistar rats which carries an autosomal gene defect is characterized by a progressively developing hyperexcitability, tremor, olfactory and gustatory movements, bradykinesia, ataxia and a pathologically increased muscle tone of hindlimbs which can be measured by recording tonic activity in the electromyogram (EMG) of the gastrocnemius-soleus muscle. The activity of the GABA synthesizing enzyme glutamic acid decarboxylase (GAD) and the receptor binding of GABA as estimated by [3H]GABA binding to synaptic membranes were examined in olfactory bulbs, frontal cerebral cortex, corpus striatum, hippocampus, thalamus, hypothalamus, tectum, substantia nigra, medulla oblongata, cerebellum, and pons of mutant rats. Mutant rats exhibit a lower activity of GAD in synaptosomal fractions of olfactory bulbs and substantia nigra whereas GAD activity within the pons was increased. The changes in the activity of GAD were accompanied by alterations in [3H]GABA binding to synaptic membranes: GABA binding was significantly elevated in the olfactory bulbs and the substantia nigra, but it was markedly reduced in the pons. The functional importance of impaired nigral GABAergic transmission in mutant rats was demonstrated by the fact that intranigral injection of the GABA agonist muscimol reduced the tonic extension of the hindlimbs as indicated by reduced tonic EMG activity of the gastrocnemius-soleus muscle, while intranigral injection of the GABA antagonist bicuculline increased the disturbance.

Drug-induced stereotypes and asymmetric behaviour after substantia nigra pars posterior (SNPP) lesions in cats.

Apomorphine- (Apo) and amphetamine- (Amph) induced behavioural phenomena were studied in cats which received either 6-hydroxydopamine (6-OHDA) or electrothermic lesions of the posterior part of the substantia nigra (SNPP). In this species, as in rats, both drugs evoked sterotypies (sniffing and head nodding) and an enhancement of locomotor activity. However, distinct differences between the reactions of cats and rats to both drugs were found, the most evident effect being the lack of sterotyped gnawing and licking. By correlating the data on behaviour, the histological examination showing the size and location of the lesions, and the dopamine (DA) content of the corresponding caudate nuclei, it is concluded that the fewer DA-specific neurons lesioned in the SNPP, the more pronounced was the ipsilateral asymmetric behaviour. We suggest, therefore, that the ipsilateral asymmetric behaviour in cats following APO- and AMPH-treatment is due to the destruction of a non-catecholaminergic output.

Blockade of excitatory neurotransmission in the globus pallidus induces rigidity and akinesia in the rat: implications for excitatory neurotransmission in pathogenesis of Parkinson's diseases.

Bilateral microinjections of the selective N-methyl-D-aspartate (NMDA) antagonist, (-)-2-amino-7-phosphonoheptanoate (AP7), 0.02-0.5 nmol, into the globus pallidus and ventral-posterior portions of the caudate-putamen result in an increase in the muscle tone (rigidity) and catalepsy (akinesia) in rats. NMDA blocked the actions of AP7 on motility in sensitive regions of the globus pallidus and caudate-putamen. Topographical differences in the action of AP7 in the striatum were detected in the dorsal-ventral and rostral-caudal direction. Microinjections of AP7 into the nucleus accumbens induced neither an increase in the muscle tone nor catalepsy in rats, while ventral regions of the caudate-putamen were sensitive to both actions of AP7. Microinjections of AP7 into the dorsal caudate-putamen resulted in a moderate or no increase in the muscle tone. AP7 failed to induce catalepsy from dorsal regions of the caudate-putamen. These data identify the globus pallidus and a defined subregion of the caudate-putamen as crucial sites where excitatory neurotransmission acts to regulate the final set-point of the respective output neurons providing modulation of the passage of motor information through the basal ganglia.