Building action repertoires: memory and learning functions of the basal ganglia

What are the computations of the cerebellum, the basal ganglia and the cerebral cortex?

The classical notion that the cerebellum and the basal ganglia are dedicated to motor control is under dispute given increasing evidence of their involvement in non-motor functions. Is it then impossible to characterize the functions of the cerebellum, the basal ganglia and the cerebral cortex in a simplistic manner? This paper presents a novel view that their computational roles can be characterized not by asking what are the "goals" of their computation, such as motor or sensory, but by asking what are the "methods" of their computation, specifically, their learning algorithms. There is currently enough anatomical, physiological, and theoretical evidence to support the hypotheses that the cerebellum is a specialized organism for supervised learning, the basal ganglia are for reinforcement learning, and the cerebral cortex is for unsupervised learning.This paper investigates how the learning modules specialized for these three kinds of learning can be assembled into goal-oriented behaving systems. In general, supervised learning modules in the cerebellum can be utilized as "internal models" of the environment. Reinforcement learning modules in the basal ganglia enable action selection by an "evaluation" of environmental states. Unsupervised learning modules in the cerebral cortex can provide statistically efficient representation of the states of the environment and the behaving system. Two basic action selection architectures are shown, namely, reactive action selection and predictive action selection. They can be implemented within the anatomical constraint of the network linking these structures. Furthermore, the use of the cerebellar supervised learning modules for state estimation, behavioral simulation, and encapsulation of learned skill is considered. Finally, the usefulness of such theoretical frameworks in interpreting brain imaging data is demonstrated in the paradigm of procedural learning.

Sequence generation in arbitrary temporal patterns from theta-nested gamma oscillations: a model of the basal ganglia-thalamo-cortical loops

A computational model that is able to generate sequences at arbitrary rates in a given serial order is presented for the cortico-basal ganglia (BG)-thalamic neural circuitry. Upon generating a sequence, this model stores information on the serial order of components in a cortical buffer by means of theta-nested gamma frequency oscillations observed experimentally in cortico-striatal neurons. This model assumes the existence of at least two functionally different classes of striatal spiny neurons. One class of striatal projection neurons (S-cells) select the first component in the cortical buffer through a temporal winner-take-all mechanism implemented by lateral inhibition. The inhibition should last for at least a few hundred milliseconds. In reality, it may be mediated by GABA(B) receptors at the presynaptic terminals of the cortico-striatal projection. The other class of striatal projection neurons (M-cells) retain the currently executed component in a cortico-BG-thalamic loop, for which the strong nonlinearity in transitions between up and down states of striatal neurons is crucial. For sequence generation at the level of striatum, the cortical neurons encoding the component selected for execution are inactivated by the feedback from the activated cortico-BG-thalamic loop. This model predicts that the transition to next component is triggered by a single external signal, i.e. the subthalamic input to the globus pallidum. This input gives a neural substrate for adjusting the rate of sequence generation.

Development of topography within song control circuitry of zebra finches during the sensitive period for song learning

Refinement of topographic maps during sensitive periods of development is a characteristic feature of diverse sensory and motor circuits in the nervous system. Within the neural system that controls vocal learning and behavior in zebra finches, axonal connections of the cortical nucleus lMAN demonstrate striking functional and morphological changes during vocal development in juvenile males. These circuits are uniquely important for song production during the sensitive period for vocal learning, and the overall size of these brain regions and their patterns of axonal connectivity undergo dramatic growth and regression during this time. Axonal connections to and from lMAN are topographically organized in adult males that have already learned song. We wondered whether the large-scale changes seen in lMAN circuitry during the time that vocal behavior is being learned and refined could be accompanied by the emergence of topographic mapping. However, results presented herein demonstrate that most of these song-control circuits show the same broad patterns of axonal connectivity between subregions of individual nuclei at the onset of song learning as seen in adult birds. Thus, coarse topographic organization is not dependent on the types of experience that are crucial for vocal learning. Furthermore, this maintenance of topographic organization throughout the period of song learning is clearly not achieved by maintenance of static axonal arbors. In fact, because the volumes of song-control nuclei are growing (or regressing), topography must be maintained by active remodeling of axonal arbors to adapt to the changes in overall size of postsynaptic targets. A salient exception to this pattern of conserved topography is the projection from lMAN to the motor cortical region RA: this pathway is diffusely organized at the onset of song learning but undergoes substantial refinement during early stages of song learning, suggesting that remodeling of axonal connections within this projection during the period of vocal learning may signify the production of increasingly refined vocal utterances.

Role of [corrected] nigrostriatal dopamine system in learning to perform sequential motor tasks in a predictive manner

Neurons in the primate striatum and the substantia nigra pars compacta change their firing patterns during sensory-motor learning. To study the consequences of nigrostriatal dopamine depletion for learning and memory of motor sequences, we used a neurotoxin, 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP), to deplete dopamine unilaterally in the striatum of macaque monkeys either before or after training them on sequential push-button motor tasks. We compared the monkeys' performance with the arms ipsilateral and contralateral to dopamine depletion. During training and retraining on the tasks, we measured initial and serial movement times and reaction times for the push button movements, electromyographic patterns of arm and orofacial muscle activity during button pushing and reward licking, and saccadic eye movements during the button push sequences. With the arm ipsilateral to the side of dopamine depletion, each monkey showed progressive shortening of movement times and initial and serial reaction times, and each developed consistent strategies of hand-orofacial and hand-eye coordination in which single button push movements were linked efficiently to succeeding movements so that performance of the whole sequence became predictive. These patterns did not develop for contralateral arm performance in this monkey treated with MPTP before training. With the arm contralateral to dopamine depletion, the monkey showed significant quantitative deficits in all parameters measured except initial reaction times. Movement times and serial reaction times were longer than those for the ipsilateral arm; anticipatory saccadic eye movements were not well time-locked to individual button pushes made with the contralateral hand; and push and licking movements were not smoothly coordinated. This monkey further showed striking differences in performance when using the ipsilateral and contralateral arms in switch trial tests in which reward was delivered unexpectedly one button early. He continued to make movements to the previously rewarded button with the ipsilateral arm but showed no such automatic movements when he used his contralateral arm. For the monkey treated with MPTP after training, performance on the push-button task was skilled for both arms before dopamine depletion, but the unilateral dopamine depletion produced deficits in contralateral arm performance for all parameters measured, again excepting initial reaction times. With retraining, however, his performance with the contralateral arm improved. We conclude that the striatum and its nigrostriatal afferents function in the initial learning underlying performance of sequences of movements as single motor programs. The nigrostriatal system also operates during the retrieval of these programs once learning is accomplished, but lesions of the nigrostriatal system spare the ability to relearn the previously acquired programs.

Corticostriatal and corticosubthalamic input zones from the presupplementary motor area in the macaque monkey: comparison with the input zones from the supplementary motor area

The presupplementary motor area (pre-SMA) is a cortical motor-related area which lies in the medial wall of the frontal lobe, immediately anterior to the supplementary motor area (SMA). This area has been considered to participate in the control of complex forelimb movements in a way different from the SMA. In an attempt to analyze the patterns of projections from the pre-SMA to the basal ganglia, we examined the distributions of pre-SMA inputs in the striatum and the subthalamic nucleus and compared them with the SMA input distributions. To detect morphologically the terminal fields from the pre-SMA and the forelimb region of the SMA, anterograde tracers were injected into such areas that had been identified electrophysiologically in the macaque monkey. Corticostriatal inputs from the pre-SMA were distributed mainly in the striatal cell bridges connecting the rostral aspects of the caudate nucleus and the putamen, as well as in their neighboring striatal portions. These input zones were located, with no substantial overlap, rostral to corticostriatal input zones from the SMA forelimb region. Corticosubthalamic input zones from the pre-SMA were almost localized in the medial aspect of the nucleus, where corticosubthalamic inputs from the SMA forelimb region were also distributed predominantly. However, the major terminal fields from the pre-SMA were centered ventrally to those from the SMA. The present results indicate that the corticostriatal and corticosubthalamic input zones from the pre-SMA appear to be segregated from the SMA-derived input zones. This implies the possibility of parallel processing of motor information from the pre-SMA and SMA in the cortico-basal ganglia circuit.

Paradoxical locomotor behavior of dopamine D1 receptor transgenic mice

The behavioral effects of augmenting dopamine D1 receptor expression in the brain were investigated in mice incorporating additional copies of the mouse D1 receptor gene. Two transgenic lines showed increases in brain D1 receptor binding sites, which were greatest in extrastriatal regions. The full D1 agonist SKF 81297, when administered systemically to control animals, stimulated a dose-dependent increase in locomotor activity. In contrast, in D1 receptor overexpressing transgenic mice, this drug caused a marked suppression of locomotion due to a decrease in the frequency of movement initiation. Amphetamine and cocaine induced comparable locomotor activation in both transgenic animals and their control littermates. In the transgenic animals, D1 agonist-induced rearing and climbing behaviors were suppressed. However, on rotarod testing, the agonist-treated transgenic and control mice performed comparably, indicating that sensorimotor coordination was unaffected. These studies demonstrate that altering the levels of D1 receptor expression reverses the effects of D1 agonism on locomotor initiation and rearing.

A critical role of the nitric oxide/cGMP pathway in corticostriatal long-term depression

High-frequency stimulation (HFS) of corticostriatal glutamatergic fibers induces long-term depression (LTD) of excitatory synaptic potentials recorded from striatal spiny neurons. This form of LTD can be mimicked by zaprinast, a selective inhibitor of cGMP phosphodiesterases (PDEs). Biochemical analysis shows that most of the striatal cGMP PDE activity is calmodulin-dependent and inhibited by zaprinast. The zaprinast-induced LTD occludes further depression by tetanic stimulation and vice versa. Both forms of synaptic plasticity are blocked by intracellular 1H-[1,2,4]oxadiazolo[4, 3-a]quinoxalin-1-one (ODQ), a selective inhibitor of soluble guanylyl cyclase, indicating that an increased cGMP production in the spiny neuron is a key step. Accordingly, intracellular cGMP, activating protein kinase G (PKG), also induces LTD. Nitric oxide synthase (NOS) inhibitors N(G)-nitro-L-arginine methyl ester hydrochloride (L-NAME) and 7-nitroindazole monosodium salt (7-NINA) block LTD induced by either HFS or zaprinast, but not that induced by cGMP. LTD is also induced by the NO donors S-nitroso-N-acetylpenicillamine (SNAP) and hydroxylamine. SNAP-induced LTD occludes further depression by HFS or zaprinast, and it is blocked by intracellular ODQ but not by L-NAME. Intracellular application of PKG inhibitors blocks LTD induced by HFS, zaprinast, and SNAP. Electron microscopy immunocytochemistry shows the presence of NOS-positive terminals of striatal interneurons forming synaptic contacts with dendrites of spiny neurons. These findings represent the first demonstration that the NO/cGMP pathway exerts a feed-forward control on the corticostriatal synaptic plasticity.

Periventricular leucomalacia and preterm birth have different detrimental effects on postural adjustments

Postural adjustments during sitting on a moveable platform were assessed by means of multiple surface EMGs of neck, trunk and leg muscles and kinematics in three groups of children, aged 1 1/2-4 1/2 years. The first group consisted of 13 preterm children (born at a gestational age of 25-34 weeks), whose neonatal ultrasounds had shown distinct lesions of the periventicular white matter (PWM). The second group was the preterm control group, consisting of 13 preterm children with normal neonatal brain scans, matched to the PWM group with respect to gestational age at birth, birth weight, sex and age of postural assessment. The third group was formed by 13 healthy children born at term and matched to the PWM group with respect to sex and age at examination. In addition to the postural assessment an age-specific neurological examination was carried out. Three of the children of the PWM group developed a cerebral palsy syndrome, nine showed minor neurological dysfunction and one child was neurologically normal. In the preterm control group one child showed minor neurological dysfunction, while the remaining 12 children of this group and all children of the full-term group were neurologically normal. The postural assessment revealed that preterm birth was associated with two types of postural dysfunction. One dysfunction was related to the presence of a PWM lesion and consisted of a limited repertoire of response variation. The other dysfunction was not related to the presence of a PWM lesion, but to preterm birth itself. It consisted of a change in the ability to modulate the postural responses. Preterm children showed a higher sensitivity to platform velocity than full-term children, and they lacked the capacity to modulate EMG amplitude with respect to initial sitting position.

Schizophrenia--a disorder of the corollary discharge systems that integrate the motor systems of thought with the sensory systems of consciousness

BACKGROUND

In spite of intensive research, no causal anatomical lesion has been found in schizophrenia. It may instead be caused by malfunctioning circuits in the corollary discharge, feed forward (CD-FF) systems of thought.

AIMS

To integrate with the CD-FF hypothesis recent data showing that subcortical motor systems participate in thinking.

METHODS

We review CD-FF concepts in relation to recent evidence that 'motor' brain structures participate in cognitive processing.

RESULTS

Malfunctioning of CD-FF systems that integrate thinking and consciousness could produce auditory hallucinations, delusions and disorganised thought.

CONCLUSIONS

We hypothesise that the pathophysiology of schizophrenia lies in integrative circuits of basal ganglia, thalamus and frontal cortex. Fruitful research directions would include elucidation of CD-FF circuits at even higher brain levels, the behaviour of these circuits during dreaming, and their responses to late maturational events including synaptic elimination.

Activation of M1-like muscarinic receptors is required for the induction of corticostriatal LTP

We studied the role of endogenous acetylcholine (Ach) in corticostriatal long-term potentiation (LTP). The muscarinic receptor antagonists scopolamine and pirenzepine fully prevented the induction of LTP suggesting that the activation of M1-like muscarinic receptors is a crucial event in the conditioning phase of this form of synaptic plasticity.

Planning and executing an action in Parkinson's disease

We evaluated the possible impairment in planning and executing an action in patients with Parkinson's disease (PD). The action considered in the present study was formed by two successive motor acts: reaching-grasping an object (first target) and placing it on a second target of the same shape and size. We examined the effects of extrinsic properties of the second target (that is, distance) on the various kinematic phases of reaching-grasping movements. Distance, position, and size of both stimuli were randomly varied across the experimental session. Movements were executed with and without visual control of both targets and arm. The performance of six patients with PD was compared with an age-matched control group. The kinematics of the initial phase of reaching was influenced by position and size of the first target and by distance of the second target in both patients and control subjects. In particular, peak acceleration was higher for farther position of the second target. However, in the subsequent phase patients, differently from control subjects, removed the effects of the second target distance by modifying their reaching kinematics. This was obtained by varying the duration of the acceleration phase. In summary, the patients reprogrammed the reaching component by taking into account only the properties of the first target. The decreasing influence of second-target distance on reaching kinematics of patients was more evident during movements executed under visual control. Moreover, their movements executed without visual control were slowed down from the beginning. The second target affected the grasping kinematics only of the control subjects. Globally, these results indicate that PD patients are able to compute the general program of an action that takes into account extrinsic properties of the final target. However, the finding that PD patients reprogrammed the movement during its execution suggests a decay of the program during its time course, that is, basal ganglia can be involved in storing the plan of an action and in controlling its correct execution.

What is the role of dopamine in reward: hedonic impact, reward learning, or incentive salience?

What roles do mesolimbic and neostriatal dopamine systems play in reward? Do they mediate the hedonic impact of rewarding stimuli? Do they mediate hedonic reward learning and associative prediction? Our review of the literature, together with results of a new study of residual reward capacity after dopamine depletion, indicates the answer to both questions is 'no'. Rather, dopamine systems may mediate the incentive salience of rewards, modulating their motivational value in a manner separable from hedonia and reward learning. In a study of the consequences of dopamine loss, rats were depleted of dopamine in the nucleus accumbens and neostriatum by up to 99% using 6-hydroxydopamine. In a series of experiments, we applied the 'taste reactivity' measure of affective reactions (gapes, etc.) to assess the capacity of dopamine-depleted rats for: 1) normal affect (hedonic and aversive reactions), 2) modulation of hedonic affect by associative learning (taste aversion conditioning), and 3) hedonic enhancement of affect by non-dopaminergic pharmacological manipulation of palatability (benzodiazepine administration). We found normal hedonic reaction patterns to sucrose vs. quinine, normal learning of new hedonic stimulus values (a change in palatability based on predictive relations), and normal pharmacological hedonic enhancement of palatability. We discuss these results in the context of hypotheses and data concerning the role of dopamine in reward. We review neurochemical, electrophysiological, and other behavioral evidence. We conclude that dopamine systems are not needed either to mediate the hedonic pleasure of reinforcers or to mediate predictive associations involved in hedonic reward learning. We conclude instead that dopamine may be more important to incentive salience attributions to the neural representations of reward-related stimuli. Incentive salience, we suggest, is a distinct component of motivation and reward. In other words, dopamine systems are necessary for 'wanting' incentives, but not for 'liking' them or for learning new 'likes' and 'dislikes'.

Cerebellar contributions to motor timing: a PET study of auditory and visual rhythm reproduction

The perception and production of temporal patterns, or rhythms, is important for both music and speech. However, the way in which the human brain achieves accurate timing of perceptual input and motor output is as yet little understood. Central control of both motor timing and perceptual timing across modalities has been linked to both the cerebellum and the basal ganglia (BG). The present study was designed to test the hypothesized central control of temporal processing and to examine the roles of the cerebellum, BG, and sensory association areas. In this positron emission tomography (PET) activation paradigm, subjects reproduced rhythms of increasing temporal complexity that were presented separately in the auditory and visual modalities. The results provide support for a supramodal contribution of the lateral cerebellar cortex and cerebellar vermis to the production of a timed motor response, particularly when it is complex and/or novel. The results also give partial support to the involvement of BG structures in motor timing, although this may be more directly related to implementation of the motor response than to timing per se. Finally, sensory association areas and the ventrolateral frontal cortex were found to be involved in modality-specific encoding and retrieval of the temporal stimuli. Taken together, these results point to the participation of a number of neural structures in the production of a timed motor response from an external stimulus. The role of the cerebellum in timing is conceptualized not as a clock or counter but simply as the structure that provides the necessary circuitry for the sensory system to extract temporal information and for the motor system to learn to produce a precisely timed response.

Expectation of reward modulates cognitive signals in the basal ganglia

Action is controlled by both motivation and cognition. The basal ganglia may be the site where these kinds of information meet. Using a memory-guided saccade task with an asymmetric reward schedule, we show that visual and memory responses of caudate neurons are modulated by expectation of reward so profoundly that a neuron's preferred direction often changed with the change in the rewarded direction. The subsequent saccade to the target was earlier and faster for the rewarded direction. Our results indicate that the caudate contributes to the determination of oculomotor outputs by connecting motivational values (for example, expectation of reward) to visual information.

Somatotopical organization of striatal activation during finger and toe movement: a 3-T functional magnetic resonance imaging study

The present study aimed at determining the distribution and somatotopical organization of striatal activation during performance of simple motor tasks. Ten right-handed healthy volunteers were studied by using a 3-T whole-body magnetic resonance unit and echo planar imaging. The tasks consisted of self-paced flexion/extension of the right fingers or toes. Motor activation was found mainly in the putamen posterior to the anterior commissure (10 of 10 subjects) and the globus pallidus (6 subjects), whereas the caudate nucleus was activated in only 3 subjects, and in a smaller area. Thus, performance of a simple motor task activated the sensorimotor territory of the basal ganglia. Within the putamen, there was a somatotopical organization of the foot and hand areas similar to that observed in nonhuman primates. These data suggest that functional magnetic resonance imaging can be used to study normal function of the basal ganglia and should therefore also allow investigation of patients with movement disorders.

Modifications of reward expectation-related neuronal activity during learning in primate striatum

This study investigated neuronal activity in the anterior striatum while monkeys repeatedly learned to associate new instruction stimuli with known behavioral reactions and reinforcers. In a delayed go-nogo task with several trial types, an initial picture instructed the animal to execute or withhold a reaching movement and to expect a liquid reward or not. During learning, new instruction pictures were presented, and animals guessed and performed one of the trial types according to a trial-and-error strategy. Learning of a large number of pictures resulted in a learning set in which learning took place in a few trials and correct performance exceeded 80% in the first 60-90 trials. About 200 task-related striatal neurons studied in both familiar and learning conditions showed three forms of changes during learning. Activations related to the preparation and execution of behavioral reactions and the expectation of reward were maintained in many neurons but occurred in inappropriate trial types when behavioral errors were made. The activations became appropriate for individual trial types when the animals' behavior adapted to the new task contingencies. In particular, reward expectation-related activations occurred initially in both rewarded and unrewarded movement trials and became subsequently restricted to rewarded trials. These changes occurred in parallel with the visible adaptation of reward expectations by the animals. The second learning change consisted in decreases of task-related activations that were either restricted to the initial trials of new learning problems or persisted during the subsequent consolidation phase. They probably reflected reductions in the expectation and preparation of upcoming task events, including reward. The third learning change consisted in transient or sustained increases of activations. These might reflect the increased attention accompanying learning and serve to induce synaptic changes underlying the behavioral adaptations. Both decreases and increases often induced changes in the trial selective occurrence of activations. In conclusion, neurons in anterior striatum showed changes related to adaptations or reductions of expectations in new task situations and displayed activations that might serve to induce structural changes during learning.

Crossed reduction of human motor cortex excitability by 1-Hz transcranial magnetic stimulation

Electrophysiological studies have shown that 1-Hz repetitive transcranial magnetic stimulation (rTMS) of the primary motor area (M1) can produce a local decrease in excitability. Functional imaging data suggest that this change may be bilateral. In normal subjects, we measured motor evoked potential (MEP) amplitude at a series of stimulation intensities in the contralateral M1 before and after 15 min of active or sham rTMS at just above the MEP threshold. The slope of the curve relating MEP amplitude and stimulation intensity was decreased in the unstimulated hemisphere by active but not sham rTMS. This demonstrates that rTMS can condition cortical excitability at a distance of one or more synapses and suggest that decreased excitability to TMS is a correlate of decreased blood flow and metabolism.

The basal ganglia and chunking of action repertoires

The basal ganglia have been shown to contribute to habit and stimulus-response (S-R) learning. These forms of learning have the property of slow acquisition and, in humans, can occur without conscious awareness. This paper proposes that one aspect of basal ganglia-based learning is the recoding of cortically derived information within the striatum. Modular corticostriatal projection patterns, demonstrated experimentally, are viewed as producing recoded templates suitable for the gradual selection of new input-output relations in cortico-basal ganglia loops. Recordings from striatal projection neurons and interneurons show that activity patterns in the striatum are modified gradually during the course of S-R learning. It is proposed that this recoding within the striatum can chunk the representations of motor and cognitive action sequences so that they can be implemented as performance units. This scheme generalizes Miller's notion of information chunking to action control. The formation and the efficient implementation of action chunks are viewed as being based on predictive signals. It is suggested that information chunking provides a mechanism for the acquisition and the expression of action repertoires that, without such information compression would be biologically unwieldy or difficult to implement. The learning and memory functions of the basal ganglia are thus seen as core features of the basal ganglia's influence on motor and cognitive pattern generators.

Coding of serial order by neostriatal neurons: a "natural action" approach to movement sequence

The neostriatum controls behavioral sequencing, or action syntax, as well as simpler aspects of movement. Yet the precise nature of the neostriatums role in sequencing remains unclear. Here we used a "natural action" approach that combined electrophysiological and neuroethological techniques. We identified neostriatal neurons that code the serial order of natural movement sequences of rats. During grooming behavior, rats emit complex but highly predictable species-specific sequences of movements, termed "syntactic chains." Neuronal activity of 41% of cells in the dorsolateral and ventromedial neostriatum coded the sequential pattern of syntactic chains. Only 14% coded simple motor properties of grooming movements. Neurons fired preferentially during syntactic chains compared with similar grooming movements made in different sequential order or to behavioral resting. Sequential coding differed between the dorsolateral and ventromedial neostriatum. Neurons in the dorsolateral site increased firing by 116% during syntactic chains, compared with only a 30% increase by neurons in the ventromedial site, and dorsolateral neurons showed strongest coding of grooming syntax by several additional criteria. These data demonstrate that neostriatal neurons code abstract properties of serial order for natural movement and support the hypothesis that the dorsolateral neostriatum plays a special role in implementing action syntax.

Thalamic deactivation during early implicit sequence learning: a functional MRI study

Previous research has implicated the striatum in implicit sequence learning. However, imaging findings have been inconsistent with regard to activity within the thalamus during performance of such tasks. Contemporary models of cortico-striato-thalamic circuitry suggest opposing influences on thalamic activity; suppression of thalamic activity is mediated by the indirect pathway and enhancement is mediated by the direct pathway. Using functional magnetic resonance imaging, we studied activity within human thalamus during early and late phases of an implicit sequence learning task known to reliably recruit the striatum. Significant deactivation (decreased signal relative to a baseline condition) was observed within the thalamus during early implicit learning. This finding is consistent with models of cortico-striato-thalamic function and specifically supports a profile of early 'thalamic gating' via the indirect pathway.

Rapid plasticity of human cortical movement representation induced by practice

The process of acquiring motor skills through the sustained performance of complex movements is associated with neural plasticity. However, it is unknown whether even simple movements, repeated over a short period of time, are effective in inducing cortical representational changes. Whether the motor cortex can retain specific kinematic aspects of a recently practiced movement is also unknown. We used focal transcranial magnetic stimulation (TMS) of the motor cortex to evoke isolated and directionally consistent thumb movements. Thumb movements then were practiced in a different direction. Subsequently, TMS came to evoke movements in or near the recently practiced direction for several minutes before returning to the original direction. To initiate a change of the TMS-evoked movement direction, 15 or 30 min of continuous training were required in most of the subjects and, on two occasions, as little as 5 or 10 min. Substantially smaller effects followed more direct stimulation of corticofugal axons with transcranial electrical stimulation, pointing to cortex as the site of plasticity. These findings suggest that the training rapidly, and transiently, established a change in the cortical network representing the thumb, which encoded kinematic details of the practiced movement. This phenomenon may be regarded as a short-term memory for movement and be the first step of skill acquisition.

Gene targeting and the biology of learning and memory

The general goal of genetic studies of learning and memory is to develop and test theories that explain the animal's behavior in neuroanatomical, neurophysiological, cellular, and molecular terms. In this review we describe the role that gene targeting and other transgenic techniques have had in the study of mammalian learning and memory. We focus especially on the hippocampus, a brain structure that is thought to be central to the processing and temporary storage of complex information. We also discuss the main issues that confront this young field, as well as our vision for its future.

Physiological aspects of information processing in the basal ganglia of normal and parkinsonian primates

There are two views as to the character of basal-ganglia processing - processing by segregated parallel circuits or by information sharing. To distinguish between these views, we studied the simultaneous activity of neurons in the output stage of the basal ganglia with cross-correlation techniques. The firing of neurons in the globus pallidus of normal monkeys is almost always uncorrelated. However, after dopamine depletion and induction of parkinsonism by treatment with 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP), oscillatory activity appeared and the firing of many neurons became correlated. We conclude that the normal dopaminergic system supports segregation of the functional subcircuits of the basal ganglia, and that a breakdown of this independent processing is a hallmark of Parkinson's disease.

Current controversies and future directions in basal ganglia research. Integrating basic neuroscience and clinical investigation

This article discusses current controversies and future directions in basal ganglia research, detailing behavioral aspects, anatomic models, neurochemistry, pharmacology, and diagnostic methods as well as surgical techniques. A neuroethologic perspective is highlighted. Furthermore, the relevant literature pertaining to contemporary molecular approaches such as brain microinjections of embryonic or genetically modified cells, for therapeutic purposes and the use of transgenic and knockout animals.

Neuroimaging and neuropsychology of the striatum. Bridging basic science and clinical practice

Neuroimaging and neuropsychology are complementary disciplines that provide powerful means for assessing the structure and function of corticostriatal systems. Findings from four model basal ganglia disorders--OCD, TS, HD, and PD--are reviewed. This survey is intended as a vehicle for illustrating the breadth of current clinical and research applications, as well as the potential for future advances. The perspectives brought by neuroimaging and neuropsychology serve as a natural bridge from the basic neuroscience to the clinical practice articles in this issue.

Neuroanatomy of the basal ganglia

This article deals with the neuroanatomic aspects of the basal ganglia with regard to different neurotransmitter systems and to different diseases. A general scheme of these circuits with the overall distinction between limbic-associative and motor components and circuits is presented.

Effect of electrical stimulation of the cerebral cortex on the expression of the Fos protein in the basal ganglia

The protein Fos is a transcription factor which is quickly induced in response to a variety of extracellular signals. Since this protein is expressed in a variety of neuronal systems in response to activation of synaptic afferents, it has been suggested that it might contribute to activity-dependent plasticity in neural networks. The present study investigated the effect of cortical electrical stimulation on the expression of Fos in the basal ganglia in the rat, a group of structures that participate in sensorimotor learning. Results show that the repetitive application of electrical shocks in restricted areas of the cerebral cortex induces an expression of Fos mostly confined to the striatum and the subthalamic nucleus. The induction which can be elicited from different cortical areas (sensorimotor, auditory and limbic areas) does not require particular temporal patterns of stimulation but rather depends on the total number of shocks delivered during a given period of time. Moreover, it appears to be rather independent of the number of spikes discharged by the activated cells. In the striatum, the distribution of immunoreactive neurons is precisely delineated and conforms to the known topographical organization of stimulated corticostriatal projections. As demonstrated using a variety of double labelling techniques (combination of the immunocytochemical detection of Fos with the autoradiography of mu opioid receptors, calbindin immunocytochemistry, in situ hybridization of preproenkephalin and preprotachykinin A messenger RNAs), striatal neurons which express Fos are mostly localized in the matrix compartment and concern equally enkephaline and substance P containing efferent neurons. In the subthalamic nucleus, Fos expression evoked by cortical stimulation is also confined to discrete regions of the nucleus, the localizations corresponding to the primary projection site of the stimulated cortical cells. These results indicate that in addition to its phasic synaptic influence on the basal ganglia, the cerebral cortex could exert a long-term effect on the functional state of this system via a genomic control. Since the basal ganglia are involved in sensorimotor learning and motor habit formation, it is tempting to speculate that the activity-dependent Fos induction at corticostriatal and subthalamic synapses may contribute to consolidate the functionality of the neuronal networks activated during the completion of given motor tasks.

Localized changes in immediate-early gene regulation during sensory and motor learning in zebra finches

A complex neural system controls birdsong learning, but its organization is not understood, nor is it known why learning only occurs during a critical period in adolescence. Here, we analyzed developmental regulation in zebra finches of zenk, an immediate-early gene (IEG) implicated in memory consolidation. Basal expression was elevated within auditory telencephalon (specifically, within the caudomedial neostriatum [NCM]) during song acquisition. Expression could be further induced by song playbacks 30 days after hatching but not at 20 days nor in juveniles reared in severe isolation. Singing itself induced zenk in song production nuclei, including Area X, even in adults. Within a compartment of the robust nucleus of the archistriatum (RA), however, this response dwindled as singing matured. These results suggest that the onset of sensory memory storage may be regulated in part at NCM, and motor plasticity may be regulated at RA.

Synaptology of the nigrostriatal projection in relation to the compartmental organization of the neostriatum in the rat

The patch-matrix organization of the striatal complex, which is fundamental to the structural and functional organization of the basal ganglia, is characterized on the basis of both connections and neurochemistry. In order to determine whether differences in the connections and neurochemistry are reflected in differences in synaptic organization, we examined the synaptology of the dopaminergic nigrostriatal projection in the patch-matrix complex of the rat. Three approaches were used. First, deposits of the anterograde tracer, biotinylated dextran amine, were placed in the substantia nigra. Sections of perfuse-fixed neostriatum were then processed to reveal anterogradely-labelled nigrostriatal axons and calbindin-D28k immunoreactivity, a marker for the patch-matrix complex. Secondly, sections of perfuse-fixed neostriatum were immunolabelled to reveal both tyrosine hydroxylase, a marker for dopaminergic structures and calbindin-D28k. Labelled axons in the patches and the matrix were examined at both the light and the electron microscopic levels. Finally, in order to test for the presence of fixed GABA in sub-type of anterogradely-labelled terminals in the neostriatum, ultrathin sections were immunolabelled by the post-embedding immunogold method. Based on morphological analysis, anterogradely-labelled nigrostriatal axons were divided into two types (Type I and Type II). The density of tyrosine hydroxylase labelling in the neostriatum prevented the classification of immunolabelled nigrostriatal axons. The Type I anterogradely-labelled axons and tyrosine hydroxylase-positive axons were found both in the patches and in the matrix. They both formed symmetrical synapses with spines, dendrites and occasionally somata. The morphology, dimensions, type of synaptic specialization and the distribution of postsynaptic targets of axons labelled by both methods were similar in the patches and the matrix. The Type I anterogradely-labelled axons were immunonegative for GABA. The Type II anterogradely-labelled axons were GABA-immunopositive, were found only in the matrix and were only present in those animals in which retrograde labelling was observed in the globus pallidus, they are thus not part of the dopaminergic nigrostriatal projection. It is concluded that although the patch-directed and matrix-directed dopaminergic projections from the ventral mesencephalon arise from different populations of dopaminergic neurons, their innervation of neurons in the patches and matrix is similar. The anatomical substrate, and therefore probably also the mechanism, for dopaminergic modulation of the flow of cortical information through the striatal complex in essentially the same in the patch and in the matrix sub-divisions of the striatal complex.

The primate striatum: neuronal activity in relation to spatial attention versus motor preparation

The primate basal ganglia are known to be involved in the initiation and control of visually guided movements. However, the precise role of these structures is not clear, partly because most neurophysiological studies have not dissociated neuronal activity related to visuomotor processing from that reflecting other aspects of behaviour, such as shifts of spatial attention. Moreover, the way the basal ganglia function together with the frontal cortex during movement initiation and execution is still a matter of debate. In an effort to clarify these issues, we recorded single neurons from the striatum (caudate nucleus and putamen) in two rhesus monkeys trained to perform a conditional visuomotor task, and compared their properties with those of the frontal cortex. The experimental paradigm was designed to distinguish neuronal activity associated with shifts of attention from that reflecting motor preparation. In a given trial, an identical visual stimulus could serve as a cue for the reorientation of spatial attention or as a cue for establishing a motor set depending on when it occurred during that trial. Additional aspects of the paradigm were designed to identify neurons whose activity differed when various stimulus configurations instructed the same action (stimulus effect), as well as neurons whose activity differed when two different actions were instructed by the same stimulus (movement effect). The majority of cells (60%) were preferentially active after instructional cues, 38% discharged preferentially after attentional cues, and the remaining 2% of cells discharged equally after both types of cue. Neurons active after instructional cues were further analysed for stimulus and movement effects. During movement preparation, the activity of the vast majority of striatal cells (putamen, 81%; caudate, 76%) varied significantly when different stimuli instructed the same action. Likewise, when different movements were instructed by the same stimulus, preparatory activity of a majority of cells (putamen, 92%; caudate, 82%) changed. Consequently, a substantial proportion of cells showed combined stimulus and movement effects. Comparison of these neuronal properties with those of the dorsal premotor cortex showed significantly higher proportions of cells in the striatum whose activity reflected sensory or sensorimotor processing. These results suggest that the basal ganglia are involved in shifting attentional set and in high-order processes of movement initiation, including the linking of sensory information with behavioural responses.

Dopamine D3 receptor mutant mice exhibit increased behavioral sensitivity to concurrent stimulation of D1 and D2 receptors

The dopamine D3 receptor is expressed primarily in regions of the brain that are thought to influence motivation and motor functions. To specify in vivo D3 receptor function, we generated mutant mice lacking this receptor. Our analysis indicates that in a novel environment, D3 mutant mice are transiently more active than wild-type mice, an effect not associated with anxiety state. Moreover, D3 mutant mice exhibit enhanced behavioral sensitivity to combined injections of D1 and D2 class receptor agonists, cocaine and amphetamine. However, the combined electrophysiological effects of the same D1 and D2 agonists on single neurons within the nucleus accumbens were not altered by the D3 receptor mutation. We conclude that one function of the D3 receptor is to modulate behaviors by inhibiting the cooperative effects of postsynaptic D1 and other D2 class receptors at systems level.

Encoding behavioral context in recurrent networks of the fronto-striatal system: a simulation study

This research addresses the hypothesis that behavioral context is encoded in recurrent networks of the fronto-striatal system. Behavioral context influences the processing of subsequent brain events, including responses to sensory inputs, thus providing a basis for context-dependent behavior. We define context-dependent behavior as the adaptive ability to produce the appropriate response to a given stimulus, dependent upon the context in which it appears. Behavioral context can change with a time-scale on the order of seconds to tens of seconds or more. This suggests a flexible mechanism that encodes context via an ensemble of neural activation that will appropriately influence the processing of subsequent sensory stimuli. We present a functional model of context encoding in recurrent connections of the fronto-striatal system with simulation results that correspond closely to empirical data. Neuronal activity in monkeys that perform a context-dependent task indicate that the prefrontal cortex and striatum participate differentially in this kind of context encoding. Likewise, simulated neurons in our model of the fronto-striatal system, which performs the context-dependent task, display task-related activity remarkably similar to that found in monkey frontal cortex and striatum, supporting our hypothesis.

Local release of GABAergic inhibition in the motor cortex induces immediate-early gene expression in indirect pathway neurons of the striatum

The neocortex is thought to exert a powerful influence over the functions of the basal ganglia via its projection to the striatum. It is not known, however, whether corticostriatal effects are similar across different types of striatal projection neurons and interneurons or are unique for cells having different functions within striatal networks. To examine this question, we developed a method for focal synchronous activation of the primary motor cortex (MI) of freely moving rats by local release of GABAergic inhibition. With this method, we monitored cortically evoked activation of two immediate-early gene protein products, c-Fos and JunB, in phenotypically identified striatal neurons. We further studied the influence of glutamate receptor antagonists on the stimulated expression of c-Fos, JunB, FosB, and NGFI-A. Local disinhibition of MI elicited remarkably selective induction of c-Fos and JunB in enkephalinergic projection neurons. These indirect pathway neurons, through their projections to the globus pallidus, can inhibit thalamocortical motor circuits. The dynorphin-containing projection neurons of the direct pathway, with opposite effects on the thalamocortical circuits, showed very little induction of c-Fos or JunB. The gene response of striatal interneurons was also highly selective, affecting principally parvalbumin- and NADPH diaphorase-expressing interneurons. The glutamate NMDA receptor antagonist MK-801 strongly reduced the cortically evoked striatal gene expression in all cell types for each gene examined. Because the gene induction that we found followed known corticostriatal somatotopy, was dose-dependent, and was selectively sensitive to glutamate receptor antagonists, we suggest that the differential activation patterns reflect functional specialization of cortical inputs to the direct and indirect pathways of the basal ganglia and functional plasticity within these circuits.

In vivo activity-dependent plasticity at cortico-striatal connections: evidence for physiological long-term potentiation

The purpose of the present study was to investigate in vivo the activity-dependent plasticity of glutamatergic cortico-striatal synapses. Electrical stimuli were applied in the facial motor cortex and intracellular recordings were performed in the ipsilateral striatal projection field of this cortical area. Recorded cells exhibited the typical intrinsic membrane properties of striatal output neurons and were identified morphologically as medium spiny type I neurons. Subthreshold cortical tetanization produced either short-term posttetanic potentiation or short-term depression of cortically-evoked excitatory postsynaptic potentials. When coupled with a postsynaptic depolarization leading the membrane potential to a suprathreshold level, the tetanus induced long-term potentiation (LTP) of cortico-striatal synaptic transmission. Induction of striatal LTP was prevented by intracellular injection of a calcium chelator suggesting that this synaptic plasticity involves an increase of postsynaptic free calcium concentration. Contrasting with previous in vitro studies our findings demonstrate that LTP constitutes the normal form of use-dependent plasticity at cortico-striatal synapses. Since excitation of striatal neurons produces a disinhibition of premotor networks, LTP at excitatory striatal inputs should favor the initiation of movements and therefore could be critical for the functions of basal ganglia in motor learning.

Differential expression of transcription factors in the accumbens of an animal model of ADHD

Transcription factors have been used as neuronal markers in the nucleus accumbens (ACB) of male juvenile spontaneously hypertensive rats (SHR), an animal model of attention-deficit hyperactivity disorder (ADHD), to trace putative neural substrates. In SHR, immunocytochemistry and PC-assisted image analysis showed lower expression of pan-fos, c-fos, zif/268 in the shell, and the c-fos and zif/268 in the core, with an increased level of Jun-B in the core. The differential lower basal expression of transcription factors in the ACB of an animal model of ADHD implies a reduced number of modules and might represent a neural substrate of the attention deficits seen in SHR and children with ADHD at low motivational levels.

Mnemonic functions of the basal ganglia

A synthesis of older and recent work on mnemonic functions of the basal ganglia in rats, monkeys and humans emphasizes a reciprocal relationship of the caudate nucleus and putamen with the cerebral cortex, which mediates the memory of consistent relationships between stimuli and responses (sometimes called habits) that often involve relationships between the individual and its environment (egocentric memory). Evidence at several levels of analysis (including neuroplastic synaptic changes, activity of single neurons, and behavioral changes caused by lesions or neurochemical manipulations) implicate dopamine release from nigro-striatal neurons in the reinforcement, or strengthening, of neural representations in the basal ganglia.

Network models of the basal ganglia

Over the past year, a number of conceptual and mathematical models of the basal ganglia and their interactions with other areas of the brain have appeared in the literature. Even though the models each differ in significant ways, several computational principles, such as convergence, recurrence and competition, appear to have emerged as common themes of information processing in the basal ganglia. Simulation studies of these models have provoked new types of questions at the many levels of inquiry linking biophysics to behavior.

A neural substrate of prediction and reward

The capacity to predict future events permits a creature to detect, model, and manipulate the causal structure of its interactions with its environment. Behavioral experiments suggest that learning is driven by changes in the expectations about future salient events such as rewards and punishments. Physiological work has recently complemented these studies by identifying dopaminergic neurons in the primate whose fluctuating output apparently signals changes or errors in the predictions of future salient and rewarding events. Taken together, these findings can be understood through quantitative theories of adaptive optimizing control.

Synaptic integration of functionally diverse pallidal information in the entopeduncular nucleus and subthalamic nucleus in the rat

To determine the principles of synaptic innervation of neurons in the entopeduncular nucleus and subthalamic nucleus by neurons of functionally distinct regions of the pallidal complex, double anterograde labeling was carried out at both light and electron microscopic levels in the rat. Deposits of the anterograde tracers Phaseolus vulgaris-leucoagglutinin and biotinylated dextran amine were placed in different functional domains of the pallidal complex in the same animals. The tracer deposits in the ventral pallidum and the globus pallidus gave rise to GABA-immunopositive projections to the entopeduncular nucleus, the subthalamic nucleus, and the more medial lateral hypothalamus that were largely segregated but overlapped at the interface between the two fields of projection. In these regions the proximal parts of individual neurons in the entopeduncular nucleus, lateral hypothalamus, and subthalamic nucleus received synaptic input from terminals derived from both the ventral pallidum and the globus pallidus. Furthermore, the analysis of the afferent synaptic input to the dendrites of neurons in the subthalamic nucleus that cross functional boundaries of the nucleus defined by the pallidal inputs, revealed that terminals with the morphological and neurochemical characteristics of those derived from the pallidal complex make synaptic contact with all parts of the dendritic tree, including distal regions. It is concluded that functionally diverse information carried by the descending projections of the pallidal complex is synaptically integrated by neurons of the entopeduncular nucleus, lateral hypothalamus, and subthalamic nucleus by two mechanisms. First, neurons located at the interface between functionally distinct, but topographically adjacent, projections could integrate diverse information by means of the synaptic convergence at the level of the cell body and proximal dendrites. Second, because the distal dendrites of neurons in the subthalamic nucleus receive input from the pallidum, those that extend across two distinct domains of pallidal input could also provide the morphological basis of integration.

Hemispheric asymmetry of movement

Studies in brain-damaged patients indicate that the left hemisphere in right-handers is specialized for controlling cognitive-motor tasks in both arms. Recent functional imaging data support this conclusion, with the finding that ipsilateral, as well as contralateral, movements activate the left, but not the right, motor cortex or association areas of either hemisphere. Future studies must aspire to identify the mechanisms for this asymmetry.

Functional and pathophysiological models of the basal ganglia

Because of new data, anatomical and functional models of the basal ganglia in normal and pathological conditions (e.g. Parkinson's and Huntington's diseases) have recently come under greater scrutiny. An update of these models is clearly timely, taking into consideration not only changes in neuronal discharge rates, but also changes in the patterning and synchronization of neuronal discharge, the role of extrastriatal dopamine, and expanded intrinsic and input/output connections of these nuclei.

Structure and function of declarative and nondeclarative memory systems

This article reviews recent studies of memory systems in humans and nonhuman primates. Three major conclusions from recent work are that (i) the capacity for nondeclarative (nonconscious) learning can now be studied in a broad array of tasks that assess classification learning, perceptuomotor skill learning, artificial grammar learning, and prototype abstraction; (ii) cortical areas adjacent to the hippocampal formation, including entorhinal, perirhinal, and parahippocampal cortices, are an essential part of the medial temporal lobe memory system that supports declarative (conscious) memory; and (iii) in humans, bilateral damage limited to the hippocampal formation is nevertheless sufficient to produce severe anterograde amnesia and temporally graded retrograde amnesia covering as much as 25 years.

A neostriatal habit learning system in humans

Amnesic patients and nondemented patients with Parkinson's disease were given a probabilistic classification task in which they learned which of two outcomes would occur on each trial, given the particular combination of cues that appeared. Amnesic patients exhibited normal learning of the task but had severely impaired declarative memory for the training episode. In contrast, patients with Parkinson's disease failed to learn the probabilistic classification task, despite having intact memory for the training episode. This double dissociation shows that the limbic-diencephalic regions damaged in amnesia and the neostriatum damaged in Parkinson's disease support separate and parallel learning systems. In humans, the neostriatum (caudate nucleus and putamen) is essential for the gradual, incremental learning of associations that is characteristic of habit learning. The neostriatum is important not just for motor behavior and motor learning but also for acquiring nonmotor dispositions and tendencies that depend on new associations.