The Basal and Motor - Hindawi Publishing Corporationofanormalmotoroutput. Finally, thebasal ganglia, aswellasthe cerebellum, playanimportantrole in motor learning processes, albeit

NEURAL PLASTICITY VOLUME 10, NO. 1-2, 2003

The Basal Ganglia and Motor Control

Henk J. Groenewegen

Department ofAnatomy, Research Institute Neurosciences Vrije Universiteit, VU UniversityMedical Center, Amsterdam, The Netherlands

ABSTRACT

This paper briefly reviews the functionalanatomy of the basal ganglia and theirrelationships with the thalamocortical system.The basal ganglia, including the striatum,pallidum, subthalamic nucleus, and substantianigra, are involved in a number of parallel,functionally segregated cortical-subcorticalcircuits. These circuits support a wide range ofsensorimotor, cognitive and emotional-motivational brain functions. A main role of thebasal ganglia is the learning and selection of themost appropriate motor or behavioralprograms. The internal functional organizationof the basal ganglia is very well suited for suchselection mechanisms, both in development andin adulthood. The question of whetherclumsiness may be, at least in part, attributedto dysfunction of the basal ganglia is discussedin the context of the differential, comple-mentary, or interactive roles of the basalganglia and the cerebellum in the developmentof motor control.

KEYWORDS

cerebellum, dopamine, development

Reprint requests to: H.J. Groenewegen MD PhD,Department of Anatomy, VU University Medical Center(VUmc), Van der Boechorststraat 7, 1081 BT Amsterdam, TheNetherlands; e-mail: [email protected]

INTRODUCTION

Voluntary (intended) movements as well asunintentional movements are based on spatial andtemporal patterns of muscle contractions that areinitiated and coordinated by different structures inthe central nervous system. The fine-tuning ofthese structures and the neuronal networks that areinvolved in motor execution are essential for theexpression of adequate motor behavior. Most ofour skilled movements, as well as our complexbehaviors, have been learned in the course ofdevelopment and have to be ’maintained’ duringadult life. Although in this extended, lifelonglearning process extensive parts of the brain areimportant, in the case of skilled movements inparticular the cerebral cortex, the cerebellum, andthe basal ganglia have a crucial role. Clumsiness isa term associated in childhood with problems inthe learning and execution of skillful movements,the neuronal basis of which is, however, poorlyunderstood (Hadders-Algra, 2003). The knowledgeof normal structural and functional relations amongthe brain structures involved in motor functions isessential for insight into the etiology and patho-genesis of various movement disorders, includingclumsiness.

The execution of willed, intentional movementsoften requires the subtle, concerted action of themotor and sensory systems. Consider, for example,eye-hand coordination during the execution of finemanipulations with the hand and fingers. Complexmovements have to be learned and they have to bepracticed frequently. The correct execution of such

(C) 2003 Freund & Pettman, U.K. 107

108 H.J. GROENEWEGEN

movements is dependent upon the exact timing inthe activation and switching of motor programsthat have been imprinted in the brain during thecourse of motor learning processes (motor memory).In daily life, once a movement has been initiated,the execution of most of our willed, intentionalmovements occurs virtually automatically.

Intentional movements are in essence initiatedby the cerebral motor cortex that directly, orindirectly via local premotor circuits, reaches thebrain stem or spinal motor neurons that project tothe muscles. Before a motor signal descends fromthe motor cortex to the brain stem and spinal cord,however, several cortical and subcortical centers,including the basal ganglia and the cerebellum,have posed their influence on the motor cortex to’shape’ the final, descending signal. The basalganglia and the cerebellum exert their influence onthe final motor output pathways largely via thethalamus on the descending, corticobulbar andcorticospinal motor pathways that originate in themotor and premotor areas of the cerebral cortex. Inthis way, both the basal ganglia and the cerebellumhave an essential and distinctive role in theorganization (coordination, timing, and sequencing)of a normal motor output. Finally, the basal ganglia,as well as the cerebellum, play an important role inmotor learning processes, albeit in differentaspects and phases.

The present account briefly reviews thefunctional-anatomical aspects of the basal gangliain relation to the brain circuits in which they areinvolved. The arrangement (described below) of’intrinsic’ connections between basal gangliastructures and their functional-anatomical relationswith the (frontal) thalamocortical system, as wellas with a number of centers in the mesencephalon,lend support to the hypothesis that the basalganglia play a role in the facilitation of intended,desired motor programs and in the suppression ofunintended or competing ones. The basal gangliamight also play a role in the learning of skillful

movements and (complex) behavioral programs,both in early development as well as during laterlife. To what extent, if at all, clumsiness can beattributed to a malfunctioning of the basal gangliaor to disturbances that are related to a role of thebasal ganglia in the development of normal motorbehavior, remains an open question.

STRUCTURE AND FIBER CONNECTIONS OFTHE BASAL GANGLIA

Which structures belong to the basal ganglia?

The brain structures that are included in the’basal ganglia’ consist of the striatum, thepallidum, the subthalamic nucleus, and thesubstantia nigra. Each individual structure is, inthe human brain, constituted by macroscopicallydifferent subnuclei. Thus, the striatum includes thecaudate nucleus, putamen, and nucleus accumbens,the pallidum consists of an internal and an externalsegment, and the ventral pallidum. The subthalamicnucleus appears, macroscopically, as an undividedmorphological unity, but the substantia nigra has aclearly distinguishable pars compacta and parsreticulata. The ventral tegmental area (VTA),situated medially to the substantia nigra in therostral mesencephalon, can also be considered apart of the basal ganglia ’family’.

The reason for including the caudate nucleus,putamen, and nucleus accumbens collectively inthe striatum is that all three nuclei have similarhistological, neurochemical, and connectionalcharacteristics. The predominant striatal neuronalelement is the medium-sized, densely spiny neuronthat receives and integrates the bulk of striatalinputs from the cerebral cortex and thalamus andthat projects to the pallidum and the substantianigra (Fig. 1; Smith & Bolam, 1990; Gerfen &Wilson, 1996). Medium-sized spiny neurons containthe neurotransmitter gamma-amino butyric acid

BASAL GANGLIA AND MOTOR CONTROL 109

(GABA) and co-localize different neuropeptides.A minority of striatal neurons is formed by varioustypes of interneurons, among which arecholinergic and parvalbuminergic, as well as avariety of peptidergic neurons (Bolam et al,2000). Although the various parts of the striammmay transfer different types of information (on thebasis of inputs from functionally distinct areas ofthe cerebral cortex; see below), as a result of itsrelative histological and neurochemical uniformity,the way in which these different types ofinformation are ’processed’ in the striatum isprobably very similar. Likewise, distinct parts ofthe pallidal complex can transfer different types ofinformation, but the morphological characteristics

of the variots subnuclei of the pallidal complexare comparable. Most of the pallidum consists ofrelatively large aspiny neurons that contain theneurotransmitter GABA; the neuronal density ofthe pallidum is much lower than that of thestriatum. Pallidal neurons receive most of theirinput from the striatum and the subthalamicnucleus, and they project either to other basalganglia structures or to the thalamus or the brainstem outside the basal ganglia circuitry (Gerfen &Wilson, 1996).

In contrast to the striatum and the pallidumthe histological and connectional characteristics ofthe two parts of the substantia nigra are verydiftrent. In its histological, neurochemical, and

MSN

Cerebral cortexAmygdala

SNCVTA

Cholinergicinterneurons

MSN

Fig. 1" Schematic representation of a medium-sized spiny neuron (MSN) in the striatum. The position of differentinputs on the neuron are indicated: excitatory cortical or amygdaloid inputs end on the head of spines,dopaminergic terminals from the substantia nigra pars compacta (SNC) and ventral tegmental area (VTA)terminate in close association with the corticostriatal inputs on the spine necks or dendritic shafts. Inputs fromneighboring medium-sized spiny neurons and striatal interneurons terminate on the proximal parts of thedendrites. Adapted from Smith and Bolam (1990).


connectional characteristics, the pars reticulatamost resembles the internal segment of thepallidum, while the pars compacta contains thedopaminergic neurons that project to both thestriatum and the (pre)frontal cortex. The VTAcontains both dopaminergic and GABAergicneurons in about equal quantities and can beconsidered a mixture of elements comparable withthe pars compacta and pars reticulata of thesubstantia nigra. The subthalamic nucleus, situatedat the junction between the ventral diencephalonand the mesencephalon, is a compact homogeneousgroup of neurons that use glutamate as neuro-transmitter. The subthalamic nucleus is stronglyinfluenced by inputs from the pallidum, as well asby extrinsic cortical and thalamic inputs. It sendsglutamatergic projections to the pallidum and tothe pars reticulata of the substantia nigra, excitingthe output neurons of the basal ganglia (Gerfen &Wilson, 1996; Wise et al., 1996).

’Position’ of the basal ganglia in forebrain circuits

The striatum can be considered the main inputstructure of the basal ganglia in that the entirecerebral cortex, in a topographical manner,projects to the striatum while also the midline andintralaminar thalamic nuclei, the hippocampus andamygdala send fibers to the striatum. All thesestriatal inputs are excitatory (Parent & Hazrati,1995; Wise et al., 1996). The transfer of corticaland thalamic information through the striatum ismodulated by dopaminergic and serotonergicinputs from the pars compacta of the substantianigra and the mesencephalic raphe nuclei,respectively. The striatum is rich in dopamine D1and D2 receptors, whereas various types of sero-tonergic receptors are expressed, among which the5HT-2 receptors are most prominent (Gerfen &Wilson, 1996). The ventral parts of the striatum,including the nucleus accumbens, as well asventral parts of the caudate nucleus and putamen,

receive ’limbic’ inputs from the hippocampus andthe amygdala; this part of the striatum contains thehighest density of serotonergic receptors andexpresses, in addition to the D1 and D2 receptors,the dopamine D3 receptor (Groenewegen et al.,1996; Diaz et al., 1995). The electrophysiologicalproperties of medium-sized spiny output neuronsof the striatum are such that they depend onconvergent excitatory inputs (from the cerebralcortex and the thalamus) to become ’active’. Inview of the very intricate patterns of overlap andsegregation of inputs from (functionally) differentcortical areas, this dependence most probablyallows for a highly selective mechanism ofactivation of specific striatal regions orpopulations (ensembles) of striatal neurons, whichis important for an understanding of the functionsof the basal ganglia (Pennartz et al., 1994; Gerfen& Wilson, 1996; see also below).

The main output of the basal ganglia is derivedfrom the internal segment of the g|obus pallidus,the pars reticulata of the substantia nigra, and theventral pallidum. These structures predominantlyproject to the ventral anterior and mediodorsalthalamic nuclei and reach in this way the cerebralcortical areas in the entire frontal lobe. Also, thecentromedian-parafascicular thalamic nucleus isreached by pallidal inputs, and this thalamicstructure projects to the motor cortex as well as tothe striatum. In addition, pallidal and nigraloutputs reach the superior colliculus, themesencephalic reticular formation, and thepedunculopontine region, in this way influencingdescending brain stem projections to the spinalmotor apparatus (Parent & Hazrati, 1995; Gerfen& Wilson, 1996).

The external segment of the globus pallidusand the subthalamic nucleus have very limited, ifany, projections outside the basal ganglia circuitrybut are intensively interconnected with each other,as well as with other subnuclei of the pallidal andnigral complex. These structures form part of the

BASAL GANGLIA AND MOTOR CONTROL lll

so-called ’indirect pathway’, or rather ’indirectnetwork’ (Bolam et al., 2000), that is interposedbetween the striatum as the basal ganglia inputstructure and the output structures, namely theinternal pallidal segment, ventral pallidum, and

pars reticulata of the substantia nigra. A ’directpathway’ also exists between the basal gangliainput and output structures, which comprise thedirect striatopallidal and striatonigral projections(Fig. 2; Gerfen & Wilson, 1996). The prevailing

CEREBRAL CORTEX

/,"

/ ;output

dopapine

SPYN

Fig. 2: Schematic representation of the organization of basal ganglia-thalamocortical circuits. Direct and indirectpathways between the striatum and the internal segment of the globus pallidus (GPi) and the substantia nigrapars reticulata (SNr) are outlined. In the direct pathway the peptides substance P (SP) and dynorphin (DYN) areexpressed, together with the dopamine D receptor. In the indirect pathway, the projection from the striatum tothe external segment of the globus pallidus (GPe) contains the peptide enkephalin (ENK) and the dopamine D2receptor. This indirect pathway is further constituted by the GABAergic projection from GPe to the subthalamicnucleus (STN) and the glutamatergic STN projections to GPi/SNr. In the thalamus the ventral anterior (VA) andmediodorsal (MD) nuclei are reached by the basal ganglia outputs. Solid lines representing projections code forGABAergic pathways, dashed lines for glutamatergic pathways, unless otherwise indicated.

112 It.Jo GROENEWEGEN

notion is that the direct and indirect pathways arisefrom two different populations of striatal medium-sized, spiny projection neurons. The direct pathwayarises from striatal neurons that contain GABA,substance P and dynorphin as neurotransmitter/neuromodulator and that express the dopamine Dreceptor. The striatal neurons that give rise to theindirect pathway, which reaches the externalpallidal segment as a first way station in thismultisynaptic pathway, contain GABA andenkephalin, and these neurons express the dopamineD2 receptor (Fig. 2; Gerfen & Wilson, 1996).Subsequent projections in this indirect pathwayinclude the GABAergic projections from theexternal segment of the globus pallidus to thesubthalamic nucleus and, subsequently, the gluta-matergic projections from the subthalamic nucleusto the internal segment of the globus pallidus andthe pars reticulata of the substantia nigra (Fig. 2;Gerfen & Wilson, 1996; Parent & Hazrati, 1995).

The projection neurons in the basal gangliaoutput structures have the electrophysiologicalcharacteristic of being tonically active and in thisway exert a tonic inhibitory influence on thethalamus and the mesencephalon. Interestingly, thedirect and indirect striatal output pathways haveopposing effects on the output neurons in thepallidum and the substantia nigra. Thus, activity inthe direct striatal output pathway produces aninhibition of the tonically active pallidal and/ornigral output neurons, resulting in a disinhibitionof their target areas (Chevalier & Deniau, 1990).By contrast, a higher activi.ty in the ’indirectnetwork’, tbr example through the activity of theindirect striatal output pathway, is ’translated’ intoan increased activity of the excitatory subthalamicprojections to the basal ganglia outputneurons,leading to a stronger inhibition of the basal gangliatargets. If a higher activity in the (we)frontalthalamocoical systems is considered to beassociated with increased motor or cognitive/behavioral output of the brain, we can conclude

that the direct pathway facilitates, whereas theindirect pathway or network suppresses suchoutput. Noteworthy is that the subthalamic nucleusnot only receives a (tonic) inhibitory input fromthe external pallidal segment .(and in this way isdisinhibited during striatal activity) but also isprojected upon directly by excitatory cortical andthalamic fibers (Gerfen & Wilson, 1996; Feger etal., 1994). The cortical fibers originate mostly inthe frontal cortex, whereas the thalamic fibers arederived from the centromedian-parafascicularcomplex. This means that (parts of) the cerebralcortex, as well as the caudal intra|aminar thalamusplay a role in a stronger inhibition of the basalganglia target areas and, thereby, the suppressionof motor and/or cognitive outputs.

Via different types of dopamine receptors inthe two populations of striatal output neurons,dopamine has an opposing role on these outputpathways of the striatum. Via the dopamine D1receptor, the activity of the direct pathway isfacilitated, whereas the dopamine D2 receptorsuppresses the activity of the indirect pathway atthe level of the basal ganglia output neurons(Gerfen & Wilson, 1996). Therefore, higher striataldopamine levels result in a disinhibition of thebasal ganglia target areas, whereas lower dopamineconcentrations at the striatal level lead to astronger inhibition of the basal ganglia targets. Thelatter situation occurs in Parkinson’s disease and isassociated with bradykinesia and hypokinesia.

As indicated above, the projections frownfunctionally different parts of the cerebral cortexto the striatum are topographically organized. Theresult of this organization is that the striatum canbe subdivided into functionally different sectors,namely a sensorimotor sector receiving convergentinputs from motor, premotor, and sensory corticalareas; an associative sector receiving inputs frown(pre)frontal, temporal, and parietal associationcortical areas; and, finally, a ’limbic sector that isprojected upon by the hippocampus, amygdala,


and parahippocampal and orbitofrontal cortices(Fig. 3). Of course, no sharp boundaries existbetween these functionally different striatalsectors. The sensorimotor sector is located dorso-laterally in the striatum (including dorsal parts ofboth the caudate and putamen), whereas in

ventromedial direction, there is a gradual transitioninto the associative (including mainly the caudatenucleus, but also ventral parts of the putamen) and,subsequently, the ’limbic’ sector of the striatum(including the nucleus accumbens and the mostventral parts of both caudate and putamen). The

Sensory-motorcortices

Associationcortices

Limbic corticesAmygdala

Fig. 3" Schematic representation of the topographic arrangement of the projections from functionally different corticalareas to the striatum, i.e., the caudate nucleus (Caud), putamen (Put), and the nucleus accumbens (Acb). Thecorticostriatal projections globally ’dictate’ the functional subdivision of the striatum in a dorsolateral’sensorimotor’ part, an intermediate and medial ’cognitive’ part and a ventrolateral ’emotional/motivational’part. Note that the functional zonation of the striatum which follows the corticostriatal organization does notcomply with the macroscopic subdivision of the striatum in caudate nucleus, putamen and nucleus accumbens.


subsequent projections from the striatum to thepallidum and substantia nigra, as well as the basalganglia outputs to the thalamus, are also topo-graphically organized. On the basis of these point-to-point relations in the projections from thecerebral cortex to the basal ganglia, the intrinsicbasal ganglia connections and the output from thebasal ganglia to the thalamocortical system, theexistence of parallel, functionally segregated basalganglia-thalamocortical circuits has been putforward (Alexander et al., 1986; Groenewegen et al.,1990). All thalamocortical circuits, via differentthalamic nuclei, are directed to different areas ofthe frontal lobe, in this way emphasizing theimportance of the basal ganglia for the wide arrayof motor, executive, and emotional-motivationalfunctions subserved by the frontal lobe, whichincludes primary motor, premotor, and associative(cognitive) and ’limbic’ prefrontal cortical areas.

Convergence and segregation of information

Next to a parallel arrangement of connectionswithin the basal ganglia and the basal ganglia-thalamocortical circuitry, convergence of infor-mation at several levels is likewise characteristicfor the organization of basal ganglia circuits andessential for the understanding of basal gangliafunctioning. For example, at the level of thestriatum, intricate patterns of overlap and segre-gation exist between the afferents from differentcortical and subcortical sources. Because themedium-sized striatal output neurons, due to theirelectrophysiological membrane properties (seeabove), are difficult to excite and need strongconvergent input from excitatory inputs to becomeactive, they are excellent ’coincident detectors’(Fig. 1; Houk et al., 1995). Thus, primarily on thebasis of’coincident’ activity of different excitatoryinputswhich may be derived from differentcerebral cortical areas, midline- or intralaminar

thalamic nuclei, the amygdala, or hippocampusspecific striatal neuronal populations can becomeactive and produce a specific pattern of outputthrough direct and/or indirect pathways. Thespatial and temporal aspects of this coincidentafferent activity determines the location and theidentity of the striatal population (’ensemble’;Pennartz et al., 1994) that becomes active at aparticular moment. At the level of the striatum,individual cortical areas can have multiple (small)areas of termination and provide in this way amechanism through which various differentcombinations between afferent inputs can beestablished (Flaherty & Graybiel, 1991). In moregeneral functional terms, such arrangements arepre-eminently suited for the activation of aparticular output on the basis of a specific set ofinputs, namely, the detection of a particularsensory, motor, cognitive, or emotional context.

The integration of different streams of infor-mation can play a role not only in the striatum butalso in other basal ganglia structuresnamely, thepallidum, the substantia nigra, and the subtha|amicnucleus. In particular the substantia nigra has beenproposed to play an important role in integratingfunctionally different streams of information thatinfluence motor and behavioral output (Haber et al.,2000). Nevertheless, the striatum can be consideredthe main locus of integrative aspects of basalganglia functioning.

FUNCTIONAL ASPECTS OF THE BASALGANGLIA

In the preceding paragraphs, haveemphasized, based on the functional-anatomicalorganization of the basal ganglia, that these brainstructures play a role in a wide array of frontallobe functions, ranging from sensorimotor andcognitive to emotional-motivational behavioralfunctions. Yet, the specific contribution of the


basal ganglia to these functions has not beendiscussed. Most considerations on the functionalrole of the basal ganglia refer to the functionaldeficits that are caused by lesions of thesestructures, either ’natural’ in the course of adisease or experimentally induced. From theextensive body of literature on this subject, we canconclude that such lesions lead either to a paucityand/or slowness of movements (hypo- andbradykinetic movement disorders) or to the releaseof unintentional movements, including choreatic,athetotic, dyskinetic, and dystonic movements(hyperkinetic movement disorders) (DeLong, 1990;Marsden & Obeso, 1994). Globally considered, thehypo- and hyperactivity in the expression of basalganglia disorders might be understandable fromthe above-discussed arrangement of direct andindirect pathways within the basal gangliacircuitry. An overactivity of the direct pathwayrelative to the indirect pathway would lead to adisinhibition of the thalamocortical system andthereby the uncoordinated or unsupervised releaseof motor output or cognitive processing. Bycontrast, the relative overactivity of the indirectpathway or network would ultimately lead to anoverinhibition of the thalamocortical system, inthis way preventing or impeding motor output orcognitive processing (DeLong, 1990; Gerfen &Wilson, 1996). Also widely accepted, however, isthat the basal ganglia are not crucial for theinitiation of movements. This conclusion is basedmainly on the observations that the electrophysio-logical activity of basal ganglia structures occursrelatively late in the initiation phase of amovement (Mink, 1996). Thus, while intentionalmovements can still be executed, even when thefunctions of the basal ganglia are impaired, thesestructures appear to be important for the qualityand correctness of a movement or a behavioral act.

Therefore, the question of what the basalganglia exactly contribute to the execution ofmovements and behavioral acts still remains open.

Interpretations based on lesions or interferenceswith, usually extensive parts of basal gangliastructures, might not reveal the real functional roleof the basal ganglia under physiological conditions.Several hypotheses have been put. forward, none ofwhich has been definitely established. Thus, earlierhypotheses have entertained the idea that the basalganglia are responsible for the automatic executionof learned movement sequences (Marsden, 1987;Marsden & Obeso, 1994). Whether the basalganglia are the site of ’storage’ of motor programsor whether these programs are laid down in thecerebral cortex remained open, whereas the basalganglia are considered crucial for ’calling up’ theseprograms and the switching between them.Hikosaka (1994) suggested that the basal ganglia areresponsible for the suppression and releasemechanisms on such innate movements as loco-motion, mastication, and so on via descendingprojections to the brain stem, as well as for similarmechanisms on complex, learned movements viathe projections to the thalamocortical system.Hikosaka (1994) also proposed a role for the basalganglia in the learning phase of complex move-ments, specifically also for the dopaminergic systemand in his proposal, motor programs are ’stored’ inthe cerebral cortex. Following the phase of learningof motor programs, the basal ganglia would have arole only in the ’activation’ of these programs, or inelements ofthose, in a particular context.

A recent hypothesis (Mink, 1996), taking intoaccount the detailed knowledge of the basal gangliacircuitry outlined above, is more specific about therole of the basal ganglia in motor behavior. Thishypothesis states that the basal ganglia are crucialfor the facilitation of desired movements and thesuppression of unwanted, competing movements(Mink, 1996). The basal ganglia, with their strongand tonic inhibitory input on the thalamic and mes-encephalic target areas, act as a general brake on theexpression of motor and behavioral output. At themoment that it is ’decided’ in the prefrontal and


premotor cortices to execute a motor program, thisinformation is sent to the striamm, as well as in acorollary manner to the subthalamic nucleus. Theactivation of striatal neurons that give rise to thedirect output pathway to either the internal segmentof the globus pallidus or to the substantia nigra parsreticulata leads to a disinhibition of thethalamocortical system that, in ram, provides the

final output for the desired movement (Fig. 4). The’parallel’ cortical excitation of the subthalamicnucleus leads to a higher activity of the basalganglia output neurons not concerned with theintended movement or complex of movements,resulting in the suppression of potentially competingmotor output (Mink, 1996). Future experiments willhave to test this attractive hypothesis for its validity.

Striatum

GPi/SNr

Cerebral Cortex

ThalamusMidbrain

Other motor Other motorprograms programs

Desired motor programsFig. 4: Schematic representation of the interactions between the cortex cerebri, the striatum, the internal segment of the

globus pallidus (GPi) or the reticular part of the substantia nigra (SNr) and the subthalamic nucleus (STN) thatare hypothesized to lead to the selection of desired motor programs while suppressing other motor programs.Lines representing GABAergic projections have solid arrowheads, those using excitatory neurotransmitters haveopen arrowheads. Slightly adapted from figure 14 in Mink (1996).

BASAL GANGIIA AND MOTOR CONTROL 117

Possible, different roles for dopamine

As indicated above, levels of dopamine in thestriatum determine, via their actions on dopamineD1 and D2 receptors, the output activity of thebasal ganglia. In general terms, low levels ofdopamine cause a strong inhibitory output of thebasal ganglia to the thalamocortical system and tothe brain stem. Low striatal dopamine levels are,therefore, associated with a paucity of movementsas well as with cognitive and emotional/motivational behavior. By contrast, high levels ofstriatal dopamine result in a low activity of theinhibitory basal ganglia outputs and, consequently,a ’disinhibition’ of its targets. This situation iscorrelated with a facilitation of movements andcognitive/behavioral acts. In this way, dopamineappears to have a ’gating role’ at the level of thestriatum. This aspect of the role of dopamine in thebasal ganglia is most probably correlated withvariations in the tonic, relatively low rate of firingof dopamine neurons (Gerfen & Wilson, 1996;Schultz, 2002).

Another, more differentiated role fordopamine has been hypothesized in the realm ofthe learning of movements and behavioral acts. Onthe basis of behavioral studies combined with invivo electrophysiology of the dopaminergicsystem, Schultz and colleagues (for review, seeSchultz, 2002) elucidated important aspects of therole of dopamine in guiding behavior in primates.Dopaminergic neurons in the ventral mesen-cephalon show phasic activations following theencounter of the animal with novel stimuli,particularly in relation to the presentation ofprimary reward. The phasic activations occurbetween 100 and 300 rnsec following the stimulusand are short-lasting. Neurons in the medialsubstantia nigra pars compacta and the VTA,projecting to ventral striatal regions, tend to showa stronger activation than those in more lateralparts of the substantia nigra. Furthermore, in well-

controlled experimental tasks of behaviorallearning, such as Pavlovian or instrumentallearning, the phasic activity of the dopaminergicneurons gradually shifts from the primary rewardto conditioned stimuli that predict the primaryreward. Interestingly, dopamine neurons show adepression in their activity if an expected reward,predictive on the basis of the previous learningprocess, does not appear (Schultz, 1998).Noteworthy is that the activations of the dopamineneurons are not related to any aspect of themovements that animals have to make during thetasks underlying the behavioral learning (Schultz,2002).

The effect of a brief, phasic activation of thedopaminergic system, which at the striatal levelleads to a spatially rather general release ofdopamine, can be interpreted as follows. Dopamineterminals are present on striatal medium-sized spinyneurons and, at the ultrastructural level, terminatevery close to the corticostriatal terminals on thespines of these neurons (Fig. 1; Smith & Bolam,1990). As discussed above, the cortico-striatalsystem is very precisely organized, and fibers fromdifferent although functionally related cortical areasconverge on the same neuron, or on small groups ofneurons (’ensembles’). These neurons will beactivated only under specific circumstances, namelywhen a sufficient number of their excitatorycorticostriatal afferents are active. Striatal medium-sized spiny neurons are therefore thought to be ideal’context detectors’ (Houk et al., 1995). The effect ofthe release of dopamine under these circumstancesmight be twofold.

First, dopamine release might have animmediate focusing or attentional effect thatcould enable or facilitate the output of aparticular ensemble of striatal neurons at theexpense of other striatal outputs. In this way,dopamine, on the basis of rewarding or errorsignals, could have a gating or ’instructional’role. Such processes could contribute to the


selection mechanisms that are thought to playa role at the level of the striatum (Pennartz etal., 1994; Redgrave et al., 1999).Second, dopamine may lead to plastic synapticchanges, building and shaping striatal outputmodules that, on the basis of a learningprocess, will be selectively and preferentiallyactive in a particular context. In this way,dopamine would have a transient role in theprocess of reward-based motor and behaviorallearning (Hikosaka, 1994; Houk et al., 1995;Schultz, 2002).

CONCLUSIONS

As discussed above, the basal ganglia appearto have a role in a wide range of sensorimotor,cognitive, and behavioral processes that areclosely associated with the executive and motorfunctions of the (pre)frontal cortex. The selectionof motor or behavioral programs, or elementsthereof, appropriate for a particular context, mightbe one of the primary functions of the basalganglia (Mink, 1996; Redgrave et al., 1999).Plasticity in the basal ganglia circuitry and learningprocesses are important fundaments for thesefunctions. In particular, the ventral striatum mightbe crucial for the learning and execution ofreward-related behavior, whereas the dorsal striatumis important for stimulus-response behavior (habits).

The development of the human basal ganglia,in particular the dopaminergic system, takes placeover an extended period of time and most probablyincludes at least the first three decades of life(Segawa, 2000). The activity of tyrosinehydroxylase, as a marker for the dopaminergicsystem, and the expression of dopamine receptorsvary significantly in this extended age period.Certain variables may reach ’adult’ levels only inthe fourth decade. The effects of lesions of thebasal ganglia are also dependent upon the age at

which they occur. Thus, whereas lesions of thedopaminergic system in early life lead to dystonicsymptoms, after the third decade such lesionsrather result in parkinsonistic symptoms (Segawa,2000). Another disease in which the basal gangliaare most likely affected and with onset in child-hood is Tourette’s syndrome. This syndrome ischaracterized by multiple chronic tics that may beconsidered as the expression of unwanted,stereotyped movements or (fragments of) behavioralacts. Within the context of the hypothesis of Mink(1996), namely, that the basal ganglia play animportant role in the release of desired and in thesuppression of unwanted movements, the symptomsin Tourette’s syndrome can be interpreted as theresult of a defective suppression mechanism in thebasal ganglia (Mink, 2001).

These two examples of basal ganglia disordersoccurring in childhood illustrate different aspectsof the role of the basal ganglia in the control ofmovements. In dystonianamely, focal or moregeneralized prolonged contraction of (groups of)muscles leading to a disturbed posturenormalskillful movements are hampered by the unwantedcontractions. In Tourette’s syndrome, fragments ofin principal normal movements are expressedbeyond the control of the patient. In both cases,however, the mechanism to select a movement or amotor program is defective.

The question whether clumsiness, at least inpart, may be attributed to a defective role of thebasal ganglia in the process of the learning andexecution of skillful movements is at presentdifficult to answer. It seems very likely thatdisturbances in the development of the cerebellumplay an important role in clumsiness(Gramsbergen, 2003). Clumsiness, however, is avery heterogeneous concept in which variousdifferent aspects of motor control, includingsensory feedback, can be involved. As far asdeficient motor programming is involved, the basalganglia probably play a role (see above). When


inappropriate timing of contractions ’within’ amore global motor program is affected, thecerebellum can be more involved. Both the basalgang|ia and the cerebellum undergo an extendedperiod of postnatal development, in humansspanning almost two decades. Factors that affectthe normal development of the brain in thisextended developmental period can affect thefunctions of both structures. An important issuefor future research might be the more precisedetermination of the relative contribution of thebasal ganglia and the cerebellum to the learning ofskillful movements. Clearly both structurescontribute to motor skill learning (Hikosaka et al.,2002), but the relative contribution of differentbrain structures remains to be elucidated. Althoughthe roles of the cerebellum and the basal gangliamay be largely complementary, an interestingquestion is to what extent interactions betweenthese structures, for example at the level of thebrain stem or the thalamus, are contributing to thelearning of skillful movements (Stein and Aziz,1999; Doya, 2000; Hikosaka et al., 2002).Hopefully, this approach will provide more insightinto the deficits in the central control of movementsthat are associated with clumsiness.

REFERENCES

Alexander GE, DeLong MR, Strick PL. 1986. Parallelorganization of functionally segregated circuitslinking basal ganglia and cortex. Ann RevNeurosci 9: 357-381.

Alexander GE, DeLong MR, Strick PL. 1986. Parallelorganization of functionally segregated circuitslinking basal ganglia and cortex. Ann RevNeurosci 9: 357-381.

Bolam JP, Hanley, JJ, Booth PAC, Bevan MD. 2000.Synaptic organisation of the basal ganglia. J Anat196: 527-542.

Chevalier G, Deniau JM. 1990. Disinhibition as abasic process in the expression of striatal functions.Trends Neurosci 13" 277-280.

DeLong MR. 1990. Primate models of movementdisorders of basal ganglia origin. Trends Neurosci13: 281-285.

Diaz J, Levesque D, Lammers CH, Griffon N, MartresMP, Schwartz JC, SokoloffP. 1995. Phenotypicalcharacterization of neurons expressing the dopa-mine D3 receptor in the rat brain. Neuroscience65:731-745.

Doya K. 2000. Complementary roles of basal gangliaand cerebellum in leaming and motor control.Curr Opin Neurobiol 10: 732-739.

Feger J, Bevan M, Crossman AR 1994. The projectionsfrom the parafascicular thalamic nucleus to thesubthalamic nucleus and the striatum arise fromseparate neuronal populations: a comparison withthe corticostriatal and corticosubthalamic efferentsin a retrograde fluorescent double-labeling study,Neuroscience 60" 125-132.

Flaherty AW, Graybiel AM. 1991. Corticostriataltransformations in the primate somatosensorysystem. Projections from physiologically mappedbody-part representations. J Neurophysiol 66:1249-1263.

Gerfen CR, Wilson CJ. 1996. The basal ganglia. In:Swanson LW, Bj6rklund A, H6kfelt T, eds,Handbook of Chemical Neuroanatomy, Vol. 12.Integrated Systems of the CNS. Part III.Amsterdam, The Netherlands: Elsevier; 371-468.

Gramsbergen AA. 2003. Clumsiness and disturbedcerebellar development: Insights from animalexperiments. Neural Plast 10: 00-00.

Groenewegen HJ, Berendse HW, Wolters JG, LohmanAHM. 1990. The anatomical relationship of theprefrontal cortex with the striatopallidal system,the thalamus and the amygdala: evidence for aparallel organization. Prog Brain Res 85" 95-118.

Groenewegen HJ, Wright CI and Beijer AVJ. 1996.The nucleus accumbens: gateway for limbicstructures to reach the motor system? In: HolstegeG, Bandler R, Saper CB, eds, The EmotionalMotor System. Prog Brain Res 107:485-511.

Haber SN, Fudge JL, McFarland NR. 2000. Striato-nigrostriatal pathways in primates form anascending spiral from the shell to the dorsolateralstriatum. J Neurosci 20: 2369-2382.

Hadders-Algra, M. 2003. Developmental coordinationdisorder: is clumsy motor behavior caused by alesion of the brain at early age. Neural Plasticity10: 39-50.


Hikosaka O. 1994. Role of basal ganglia in control ofinnate movements, learned behavior andcognition--a hypothesis. In: Percheron G,McKenzie JS, Feger J, eds, The Basal Ganglia.IV. New Ideas and Data on Structure andFunction. New York, NY, USA: Plenum Press;589-596.

Hikosaka O, Nakamura K, Sakai K, Nakahara H.2002. Central mechanisms of motor skill learning.Curr Opin Neurobiol 12:217-222.

Houk JC, Adams JL, Barto AG. 1995. A model howthe basal ganglia generate and use neural signalsthat predict reinforcement. In: Houk JC, Davis JL,Beiser DG, eds, Models of InformationProcessing In the Basal Ganglia. Cambridge,Massachusetts, USA: MIT Press; 117-130.

Marsden CD. 1987. What do the basal ganglia tellpremotor cortical areas? Ciba Found Symp 132:282-300.

Marsden CD, Obeso JA. 1994. The functions of thebasal ganglia and the paradox of stereotaxicsurgery in Parkinson’s disease. Brain 117: 877-897.

Mink JW. 1996. The basal ganglia: Focused selectionand inhibition of competing motor programs.Prog Neurobiol 50:381-425.

Mink JW. 2001. Basal ganglia dysfunction inTourette’s syndrome: a new hypothesis. PediatrNeurol 25: 190-198.

Parent A, Hazrati L-N. 1995. Functional anatomy ofthe basal ganglia. I. The cortico-basal ganglia-thalamo-cortical loop. Brain Res Rev 20:91-127.

Pennartz CMA, Groenewegen HJ and Lopes da SilvaFH. 1994. The nucleus accumbens as a complexof functionally distinct neuronal ensembles: Anintegration of behavioural, electrophysiologicaland anatomical data. Prog Neurobiol 42: 719-761.

Redgrave P, Prescott TJ, Gurney K. 1999. The basalganglia: a vertebrate solution to the selectionproblem? Neuroscience 89:1009-1023.

Schultz W. 1998. Predictive reward signal of dopamineneurons. J Neurophysiol 80: 1-27.

Schultz W. 2002. Getting formal with dopamine andreward. Neuron 36:241--263.

Segawa M. 2000. Development of the nigrostriataldopamine neuron and the pathways in the basalganglia. Brain Dev 22: S 1-$4.

Smith AD, Bolam JP. 1990. The neural network of thebasal ganglia as revealed by the study of synapticconnections of identified neurones. TrendsNeurosci 13: 259--265.

Stein JF, Aziz TZ. 1999. Does imbalance betweenbasal ganglia and cerebellar outputs cause move-ment disorders. Curr Opin Neurol 12: 667-669.

Wise SP, Murray EA, Gerfen CR. 1996. The frontalcortex-basal ganglia system in primates. Crit RevNeurobiol 10:317-356.

Submit your manuscripts athttp://www.hindawi.com

Neurology Research International

Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

Alzheimer’s DiseaseHindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

International Journal of

ScientificaHindawi Publishing Corporationhttp://www.hindawi.com Volume 2014


BioMed Research International


Research and TreatmentSchizophrenia

The Scientific World JournalHindawi Publishing Corporation http://www.hindawi.com Volume 2014


Neural Plasticity


Parkinson’s Disease


Research and TreatmentAutism

Sleep DisordersHindawi Publishing Corporationhttp://www.hindawi.com Volume 2014


Neuroscience Journal

Epilepsy Research and TreatmentHindawi Publishing Corporationhttp://www.hindawi.com Volume 2014


Psychiatry Journal


Computational and Mathematical Methods in Medicine

Depression Research and TreatmentHindawi Publishing Corporationhttp://www.hindawi.com Volume 2014


Brain ScienceInternational Journal of

StrokeResearch and TreatmentHindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

Neurodegenerative Diseases


Journal of

Cardiovascular Psychiatry and NeurologyHindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

Documents

The Basal and Motor - Hindawi Publishing Corporationofanormalmotoroutput. Finally, thebasal ganglia, aswellasthe cerebellum, playanimportantrole in motor learning processes, albeit