The primate motor thalamus

Repetitive transcranial magnetic stimulation of the motor cortex attenuates pain perception in complex regional pain syndrome type I

In complex regional pain syndrome (CRPS) many clinical symptoms suggest involvement of the central nervous system. Neuropathic pain as the leading symptom is often resistant to therapy. In the present study we investigated the analgesic efficiency of repetitive transcranial magnetic simulation (rTMS) applied to the motor cortex contralateral to the CRPS-affected side. Seven out of ten patients reported decreased pain intensities. Pain relief occurred 30 s after stimulation, whereas the maximum effect was found 15 min later. Pain re-intensified increasingly 45 min after rTMS. In contrast, sham rTMS did not alter pain perception. These findings provide evidence that in CRPS I pain perception can be modulated by repetitive motor cortex stimulation.

Involvement of Human Thalamic Neurons in Internally and Externally Generated Movements

Several anatomical studies support the existence of recurrent neural pathways from cortical motor areas to the thalamus via basal ganglia and back to the cortex. Neuronal responses to internally and externally generated sequential movements have been studied in the motor and premotor cortex of monkeys, but the involvement of subcortical motor structures such as the thalamus have not been studied in monkeys or humans. We examined the activity of neurons during a sequential button press task in motor thalamus of parkinsonian as well as chronic pain patients undergoing implantation of deep brain stimulating electrodes. Single and dual microelectrode recordings were carried out during an internally generated task with a memorized sequence (MEM) and an externally driven task with the sequence given during task performance (follow). Average histograms of neuronal firing were constructed for each task and aligned with respect to visual cues (ready, go) or button presses (P1, P2, P3). Sequential movements were monitored with surface electromyography and hand accelerometry, and cell responses were divided into movement-defined epochs for ANOVA and post hoc means testing. Of 52 neurons tested, 31 were found to have task-related responses and 10 were task-selective with 4 responding preferentially to MEM and 7 responding preferentially to follow (1 was both). Complex responses were found including preparatory, delay period, and phase- and task-specific activity. These kinds of responses suggest a role of the thalamus in both internally and externally cued arms movement and provide some evidence for a role in sequential movements.

Subcortical afferents to the lateral mediodorsal thalamus in cynomolgus monkeys

The mediodorsal (MD) nucleus of the thalamus has long been known to provide the principal source of subcortical input to the primate prefrontal cortex, as well as to other areas of the frontal lobe that are thought to contribute to higher-order cognitive functions. In this study, we used injections of retrograde tracers in the lateral portion of the monkey MD to assess the locations of labeled cells in subcortical structures. Three main patterns were identified in the distribution of subcortical connections. We found that the claustrum, superior colliculus and ventral midbrain regions were heavily labeled in the cases with injections in caudoventral MD. In these cases, labeled cells were also found in either the periaqueductal gray or zona incerta, depending on the specific case. In one case with an injection in anterodorsal MD, labeled cells were most numerous in the structures of the ventral midbrain, especially the ventral tegmental area. Finally, the claustrum and superior colliculus contained the largest percentage of labeled subcortical cells in cases with injections in ventrolateral MD. These three patterns of subcortical label corresponded to three equally distinctive trends in the distribution of MD connections with the cortex in these same cases [J Comp Neurol 473 (2004) 107]. Very few labeled cells were found in other areas such as the amygdala, globus pallidus and deep cerebellar nuclei, suggesting that pathways leading from these structures to dorsolateral and dorsomedial frontal cortices are not likely to include the lateral divisions of MD. In concert, these findings show that particular locales within lateral MD receive distinct profiles of subcortical afferents, and project into specific neocortical domains, suggesting that these different sites within lateral MD may participate in functionally distinct circuits of information processing.

The role of the thalamus in motor control

Two characteristics of the thalamus--its apparently simple relay function and its daunting multinuclear structure--have been customarily viewed as good reasons to study something else. Yet, now that many other brain regions have been explored and neurophysiologists are turning to questions of how larger circuits operate, these two characteristics are starting to seem more attractive. First, the relay nature of thalamic neurons means that recording from them, like tapping into a wire, can reveal the signals carried by specific circuits. Second, the concentration of like relay neurons into nuclei means that inactivating or stimulating them can efficiently test the functions of the circuits. Recent studies implementing these principles have revealed pathways through the thalamus that contribute to generating movements and to monitoring one's own actions (corollary discharge).

Tremor-related activity of neurons in the 'motor' thalamus: changes in firing rate and pattern in the MPTP vervet model of parkinsonism

The pathophysiology of parkinsonian tremor remains a matter of debate with two opposing hypotheses proposing a peripheral and a central origin, respectively. A central origin of tremor could arise either from a rhythmic activity of the internal segment of the globus pallidus (GPi) or from a structure such as the thalamus, outside the basal ganglia. In this study, single-unit recordings were performed in three 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine-treated monkeys within the GPi and within three territories of the motor thalamus (delimited by their afferent inputs from the GPi, the substantia nigra and the cerebellum, respectively). For each recorded neuron, we compared the variations in firing rate and pattern in tremor and no tremor periods. Tremor either occurred spontaneously or was induced by external stimulation. When the animals entered into a tremor period we observed: (i) an increase in the mean firing rate in about half of the recorded neurons of the motor thalamus; and (ii), a change from an irregular to a rhythmic discharge within the range of tremor frequency (5-7 Hz) in about 10% of the recorded neurons of the motor thalamus (pallidal and cerebellar territories) and the GPi. Most of the thalamic neurons that exhibited a rhythmic discharge during tremor were found to be sensitive to external stimulation. Because the changes in firing rate occurred predominantly in the motor thalamus and not in the GPi, and because a fast rhythmic discharge of 10-15 Hz was frequently observed in the GPi and not in the motor thalamus, we conclude that thalamic activity is not a simple reproduction of basal ganglia output. Moreover, we suggest that thalamic processing of basal ganglia outputs could participate in the genesis of tremor, and that this thalamic processing could be influenced by sensory inputs and/or changes in attentional level elicited by external stimulation.

Distribution and morphology of nigral axons projecting to the thalamus in primates

This study presents an analysis of the distribution and organization pattern of axons originating from the substantia nigra pars reticulata and projecting to the thalamus in monkeys. Biotin dextran amine was iontophoretically injected into different parts of the substantia nigra pars reticulata of monkeys (African green monkeys and macaques). In whatever part of the substantia nigra the injection was made, numerous axonal endings were found to be distributed within different thalamic regions: the ventral anterior nucleus and mainly its magnocellular part, the most ventromedial part of the ventral lateral nucleus, and the mediodorsal and parafascicular nuclei. Moreover, the nigrothalamic projection appeared to be topographically organized. Ten anterogradely labeled axons were reconstructed from serial sections. The axons terminated in three to six terminal fields per axon located in a relatively small portion of only one thalamic region. These terminal fields were variable in size and comprised 4-43 very thin, varicose branches. They consisted either of different axonal branches of the same axon or of different axons and covered 10-31 thalamic cell bodies. These findings demonstrate that the overall morphological organization of individual nigral axons is complex and allows single axons to influence thalamic neurons via a combination of divergent, convergent, and amplification processes.

Comparison of the basal ganglia in rats, marmosets, macaques, baboons, and humans: volume and neuronal number for the output, internal relay, and striatal modulating nuclei

This study compares the basal ganglia of rats, marmosets, macaques, baboons, and humans. It uses established protocols to estimate the volume and number of neurons within the output nuclei (internal globus pallidus, IGP; and nondopaminergic substantia nigra, SNND), two internal relay and modulating nuclei (subthalamic nucleus, STh; and external globus pallidus, EGP), and a modulator of the striatum (dopaminergic substantia nigra, SND). Nuclear boundaries were defined by using immunohistochemistry for striatal afferents. Total numbers of Nissl-stained and parvalbumin-immunoreactive neurons were calculated by using the fractionator technique. Comparisons between species were standardized relative to brain mass (rats < marmosets < macaques < baboons < humans). The EGP consistently had more neurons relative to the IGP, STh, and SND, which had similar neuronal numbers within each species. The SNND had proportionally more neurons in rats than in primates (especially humans). The distribution of SND neurons varied substantially between rats and primates (very few ventrally located neurons in rats) with humans containing fewer SND neurons than other primates. The reduction in SND neurons in humans suggests less dopaminergic regulation of the basal ganglia system compared with other species. The consistency in the number of IGP neurons across all species, combined with the reduction in SNND neurons in humans, suggests a greater emphasis on output pathways through the IGP and that there are proportionally more STh and EGP neurons in humans.

Two types of projection neurons in the internal pallidum of primates: single-axon tracing and three-dimensional reconstruction

The axonal projections of the internal pallidum (GPi) in cynomolgus monkeys (Macaca fascicularis) were studied by labeling small pools of neurons with biotinylated dextran amine. Fifty-two axons were reconstructed entirely from serial sections with a camera lucida. Two types of projection neurons were identified in the GPi on the basis of their target sites. The abundant and centrally located type I neurons gave rise to a long axonal branch that descended directly to the pedunculopontine tegmental nucleus, where it arborized discretely. Other branches ascended to the thalamus and broke into 10-15 thinner collaterals that ran through most of the ventral anterior nucleus, where they terminated as typical plexuses. About half of these axons gave rise to collaterals that arborized in both components of the centre médian/parafascicular thalamic complex. The less numerous and peripherally located type II neurons had an axon that climbed the rostral thalamic pole, coursed along the stria medullaris, and arborized profusely within the lateral habenular nucleus, which stood out as the most densely innervated pallidal target. Some type II axons provided collaterals to the anterior thalamic nuclei. A small proportion of axons of both types had branches that crossed the midline and terminated in contralateral GPi target structures. Three-dimensional reconstruction showed that type I axons arborized principally along the sagittal plane. These data reveal that GPi neurons of type I act through a widely distributed axonal network upon thalamic and brainstem premotor neurons, whereas type II neurons act in a much more focused manner upon lateral habenular neurons.

Antiparkinsonian and behavioral effects of inactivation of the substantia nigra pars reticulata in hemiparkinsonian primates

Altered activity in one of the output nuclei of the basal ganglia, the internal segment of the globus pallidus, is known to play an important role in the generation of parkinsonism. These inactivation studies tested the hypothesis that altered activity in the second major output nucleus of the basal ganglia, the substantia nigra pars reticulata (SNr), also contributes to parkinsonian motor signs. To this end, three rhesus monkeys were rendered hemiparkinsonian by intracarotid injections of MPTP. The animals then received intra-SNr injections of the GABA(A) receptor agonist muscimol to inactivate small portions of the SNr. Before and after these injections, parkinsonian motor signs were evaluated with a battery of behavioral observation methods. Injections into the centrolateral SNr reduced contralateral limb akinesia and bradykinesia in two animals. By contrast, medial injections induced generalized activation, contralateral turning, and saccadic eye movements in all animals. Injections in the most lateral and posterior portions of the nucleus had no effects. Two of the animals also received ibotenic acid lesions of the SNr, followed by a series of similar observations. These injections induced improvements in limb akinesia, postural improvements, and turning. The experiments suggest that the anterolateral "motor" territory of the SNr is involved in the development of appendicular parkinsonian motor signs.

Cerebral perfusion in chronic fatigue syndrome and depression

BACKGROUND

Patients with chronic fatigue syndrome (CFS) and depressive illness share many, but not all, features.

AIMS

To test the hypothesis that patients with CFS have abnormal cerebral perfusion, that differs from that in patients with depressive illness.

METHOD

We recruited 30 patients with CFS who were not depressed, 12 depressed patients and 15 healthy volunteers. Regional cerebral perfusion at rest was assessed using region of interest (ROI) and voxel-based statistical parametric mapping (SPM) techniques.

RESULTS

On SPM analysis there was increased perfusion in the right thalamus, pallidum and putamen in patients with CFS and in those with depressive illness. CFS patients also had increased perfusion in the left thalamus. Depressed patients differed from those with CFS in having relatively less perfusion of the left prefrontal cortex. The results were similar on ROI analysis.

CONCLUSIONS

Abnormal cerebral perfusion patterns in CFS subjects who are not depressed are similar but not identical to those in patients with depressive illness. Thalamic overactivity may be a correlate of increased attention to activity in CFS and depression; reduced prefrontal perfusion in depression may be associated with the greater neuropsychological deficits in that disorder.

Electrical stimulation of motor cortex for pain control: a combined PET-scan and electrophysiological study

Although electrical stimulation of the precentral gyrus (MCS) is emerging as a promising technique for pain control, its mechanisms of action remain obscure, and its application largely empirical. Using positron emission tomography (PET) we studied regional changes in cerebral flood flow (rCBF) in 10 patients undergoing motor cortex stimulation for pain control, seven of whom also underwent somatosensory evoked potentials and nociceptive spinal reflex recordings. The most significant MCS-related increase in rCBF concerned the ventral-lateral thalamus, probably reflecting cortico-thalamic connections from motor areas. CBF increases were also observed in medial thalamus, anterior cingulate/orbitofrontal cortex, anterior insula and upper brainstem; conversely, no significant CBF changes appeared in motor areas beneath the stimulating electrode. Somatosensory evoked potentials from SI remained stable during MCS, and no rCBF changes were observed in somatosensory cortex during the procedure. Our results suggest that descending axons, rather than apical dendrites, are primarily activated by MCS, and highlight the thalamus as the key structure mediating functional MCS effects. A model of MCS action is proposed, whereby activation of thalamic nuclei directly connected with motor and premotor cortices would entail a cascade of synaptic events in pain-related structures receiving afferents from these nuclei, including the medial thalamus, anterior cingulate and upper brainstem. MCS could influence the affective-emotional component of chronic pain by way of cingulate/orbitofrontal activation, and lead to descending inhibition of pain impulses by activation of the brainstem, also suggested by attenuation of spinal flexion reflexes. In contrast, the hypothesis of somatosensory cortex activation by MCS could not be confirmed by our results.

A review of otolith pathways to brainstem and cerebellum

Our knowledge of otolith pathways is developing rapidly, but is still far from complete. Primary afferents from the sacculus and utricle terminate mainly in the lateral, inferior and caudal superior vestibular nuclei, and the ventral cerebellum, in particular the nodulus. Otolith signals descend via reticulo- and vestibulospinal pathways in the spinal cord to influence neck motoneurons and ascending proprioceptive afferents. Utricular information can reach the extraocular eye muscles via mono-, di-, and multisynaptic pathways, but saccular afferents probably only by multisynaptic pathways. The otolith signals are relayed from the vestibular nuclei, medullary reticular formation, inferior olive, and lateral reticular nucleus to sagittal zones in the caudal cerebellar vermis (nodulus and uvula), and influence the deep cerebellar nuclei. The graviceptive information could be channeled by the cerebellar efferents back to the vestibular and inferior olive complex, or fed into ascending pathways that would innervate the mescencephalon, the thalamus, and cerebral cortex.

Thalamic interaction between the input and the output systems of the basal ganglia

The striatal return through the thalamus is largely neglected in current studies dealing with basal ganglia function, and its role within this circuitry remains obscure. In this contribution the thalamus is regarded as an important place of interaction between the input and the output organization of the basal ganglia. In support of this idea, a brief overview is provided of some of the most recent findings concerning the thalamus in relation to the basal ganglia circuitry. In particular, we have focused on the thalamostriatal projections themselves, on the output of the basal ganglia to the thalamus and also on the overlapping territories between the thalamic projection of the output nuclei and the thalamostriatal neurons. These data support the existence of several thalamic feedback circuits within the basal ganglia neural system. Finally, some considerations are provided upon the functional significance of these thalamic feedback circuits in the overall organization of the basal ganglia.

Double retrograde tracer study of the thalamostriatal projections to the cat caudate nucleus

The distribution of thalamostriatal neurons projecting to the cat caudate nucleus was examined by retrograde fluorescent tracers. Thus, Fast Blue and Diamidino Yellow were concomitantly injected in different rostrocaudal, dorsoventral, or mediolateral sectors of the caudate nucleus. The main findings of this study are as follows: (1) few double-labeled cells were found after two injections in different sectors of the caudate nucleus; (2) double-labeled neurons were more abundant after adjacent injections and they were mainly located in the caudal intralaminar nuclei, in the rhomboid nucleus and in the dorsal mediodorsal nucleus; and (3) there were variations in the spatial organization of the thalamostriatal neurons projecting to various sectors of the caudate nucleus in the different thalamic nuclei known to project to this part of the striatum.

Characterization and distribution of Hxt1, a Na(+)/Cl(-)-dependent orphan transporter, in the human brain

Rxt1, a transporter-like protein structurally related to the large family of Na(+)/Cl(-)-dependent carriers, was isolated from the rat brain. In the present study, Hxt1, the homologue of Rxt1, was isolated from human cortex cDNA. Comparison of their respective nucleotidic sequences revealed a 96% conservation between Hxt1 and Rxt1. Genetic mapping with human genome radiation hybrids allowed the location of the gene coding for Hxt1 between 323ya5 and 084xb3 AFM markers, on a portion of chromosome 1p which spans over 7 cM or 118 cRay. Northern blot analyses demonstrated that Hxt1 mRNA ( approximately 7.5 Kb) is expressed in the human brain but not in peripheral tissues. The immunodistribution of Hxt1 was determined with antibodies raised against the C-terminus of Rxt1. Hxt1 is concentrated in the cerebral cortex, caudate-putamen, substantia nigra, hippocampus, and cerebellum, appearing as a diffuse or a punctate labeling at the light microscope level. This regional and cellular distribution suggests that Hxt1, as its rat homologue, could be present in axon terminals of glutamatergic neurons. The high pressure of selection exerted upon this protein, its strategic anatomical and subcellular distributions suggest that this orphan transporter could be involved in critical functions in the central nervous system.

Pathology of symptomatic tremors

Symptomatic tremors are labeled in the literature under different names including rubral tremor, midbrain tremor, thalamic tremor, myorhythmia, Holmes' tremor, cerebellar tremor, and goal-directed tremor. The most common tremor is a delayed-onset postural and action tremor with a low frequency of 3 Hz and a proximal distribution. Resting irregular tremor is sometimes present. Mild cerebellar dysmetria is often detected. The lesions are mainly located in the thalamus, the brain stem, and the cerebellum, with secondary interruption and degeneration of various pathways and olivary hypertrophy. The more consistent lesions are found in the cerebello-thalamo-cortical and dentato-rubro-olivary pathways. The role of superimposed dysfunction of the nigrostriatal system may account for the rest component. The role of the basal ganglia in the emergence and control of tremor is poorly understood.

Calretinin-immunoreactive neurons in the human thalamus

The distribution of the calcium-binding protein calretinin in the thalamus of normal human individuals was studied with immunohistochemistry. Calretinin immunoreactivity was weak in the geniculate bodies and in nuclei of the ventral and posterior groups, moderate in the reticular nucleus and in nuclei of the anterior, medial, and lateral groups, and strong in nuclei of the midline group and anterior intralaminar nuclei. The mediodorsal nucleus was unique among thalamic nuclei because it contained a wide variety of intensely immunostained perikarya embedded in a moderately-labelled neuropil. The reticular nucleus displayed several small and uniformly distributed neuronal clusters composed of immunostained perikarya lying in a moderately-labelled neuropil. Intense and uniform immunostaining was observed in all midline nuclei and in the anterior intralaminar nuclei, including the paracentral and central lateral nuclei. These nuclei, which harboured numerous intensely-stained perikarya lying in a dense immunoreactive neuropil, were the most strongly-immunoreactive structures of the entire human thalamus. At the level of the posterior intralaminar nuclei, the central median nucleus was virtually free of immunostaining whereas the parafascicular nucleus was moderately labelled. The nucleus submedius located just beneath the central median/parafascicular complex displayed a very intense calretinin immunostaining. This study has provided evidence for the presence of the protein calretinin in the human thalamus. The pattern of distribution of calretinin, as delineated in the present study, suggests that this calcium-binding protein may participate in various subcortical and cortical thalamic systems involved in the modulation of emotional and motivational states.

Nigrothalamic projections and nigrothalamocortical pathway to the medial agranular cortex in the rat: single- and double-labeling light and electron microscopic studies

Although the rat medial agranular cortex (AGm) has been implicated in a variety of motor functions, the source of the afferents impinging upon thalamic neurons projecting to the AGm is not directly known. The main purpose of this study was to determine whether the AGm is a major recipient of the nigrothalamocortical pathway. This issue was addressed by two sets of experiments. First, the organization of the nigrothalamic projections was studied by light and electron microscopy following injections of the anterograde tracer Phaseolus vulgaris-leucoagglutinin (PHA-L) into the pars reticulata of the substantia nigra (SNR). The major finding of this study was the disclosure of a heretofore unknown projection to the rostromedial part of the ventral anterior-ventral lateral complex (VAL). This projection originates exclusively from the ventral portion of the SNR and is comparable in strength to the well-known nigrothalamic projection to the ventromedial nucleus (VM). Electron microscopic examination revealed differences in the synaptic organization of nigral terminals in the VAL and the VM. A large proportion of the labeled terminals in the VAL was involved in axosomatic synapses, whereas, in the VM, the axosomatic synapses were rare, and 67% of nigral terminals were found in contact with thin dendrites. To assess a possible disynaptic nigrothalamocortical pathway to the AGm, a double-labeling strategy combining PHA-L injections in the SNR and pressure injections of the retrograde tracer, cholera toxin subunit B (CTB) in the AGm was employed. The greatest density of CTB-labeled neurons was found in the rostral and central portion of the VAL, coincident with the nigrothalamic labeling originating from the ventral SNR. Electron microscopic analysis confirmed that some of the PHA-L-labeled terminals established synaptic contacts with the CTB-labeled cell bodies and large dendrites. In conclusion, our findings indicate that there exist two different nigrothalamocortical pathways through the motor thalamus in the rat. The SNR-VAL-AGm cortical projection may play a role in oculomotor functions, whereas the SNR-VM-cortical pathway has been implicated in arousal mechanisms.

High-frequency stimulation of the globus pallidus internalis in Parkinson's disease: a study of seven cases

The effectiveness of ventroposterolateral pallidotomy in the treatment of akinesia and rigidity is not a new discovery and agrees with recent investigations into the pathogenesis of Parkinson's disease, which highlight the role played by the unbridled activity of the subthalamic nucleus (STN) and the consequent overactivity of the globus pallidus internalis (GPi). Because high-frequency stimulation can reversibly incapacitate a nerve structure, we applied stimulation to the same target. Seven patients suffering from severe Parkinson's disease (Stages III-V on the Hoehn and Yahr scale) and, particularly, bradykinesia, rigidity, and levodopa-induced dyskinesias underwent unilateral electrode implantation in the posteroventral GPi. Follow-up evaluation using the regular Unified Parkinson's Disease Rating Scale has been conducted for 1 year in all seven patients, 2 years in five of them, and 3 years in one. In all cases high-frequency stimulation has alleviated akinesia and rigidity and has generally improved gait and speech disturbances. In some cases tremor was attenuated. In a similar manner, the authors observed a marked diminution in levodopa-induced dyskinesias. This could be an excellent primary therapy for younger patients exhibiting severe bradykinesia, rigidity, and levodopa-induced dyskinesias, which would allow therapists to keep ventroposterolateral pallidotomy in reserve as a second weapon.

A reliable system of ventricular coordinates for the cartography and stereotaxy of the amygdala (and anterior hippocampus) in macaques

The most reliable stereotactic methods in primates resort to ventricular as opposed to bony landmarks. The usual CA-CP system did not appear satisfactory for stereotaxy of the amygdala and anterior hippocampus. Variation studies on ventriculograms and reconstructions from histological material were done to find more reliable systems. The most precise system of coordinates for cartography and stereotaxy of the amygdala is based on the 'amygdalar notch', a ventricular diverticulum forming the angle between the inferior and anterior borders of the temporal horn of the lateral ventricle, located beneath the inferior border of the amygdala. The AN point, the vertex of the notch in the sagittal direction, is a reliable ventricular landmark in the antero-posterior and infero-superior directions. The medial vertex of the diverticulum, ANm, is a ventricular landmark in the medio-lateral direction. The 'AN system of coordinates' is a reliable system for amygdalar stereotaxy. Stereotaxy of the anterior hippocampus would benefit from other ventricular landmarks. The most lateral point of the main body (HiL) is a landmark for the medio-lateral dimension. Ventriculography and ventricular landmarks should always be used for stereotaxy in primate species.

Three-dimensional morphology and distribution of pallidal axons projecting to both the lateral region of the thalamus and the central complex in primates

This study presents a three-dimensional analysis of pallido-thalamic axons and axonal endings in the monkey (Macaca mulatta and M. irus). Injections of the anterograde tracer biocytin were made in the dorsal, associative region of the medial pallidum. Numerous axonal endings were observed within the pallidal territory of the lateral region of the thalamus and the central complex. Individual axons were reconstructed from serial sections and traced in three dimensions. Two axons made a collateral branch in the ventral part of the lateral region and ended in the central complex. In the pallidal territory of the lateral region, axons divided several times before ending in different parts of the territory in a 'bunch', a characteristic dense terminal arborization. Axonal endings in the central complex were differently organized. Our data show that associative medial pallidal information is distributed throughout the pallidal territory of the lateral region and the pars media of the central complex by means of individual axons with numerous branches and axonal endings specific to each of the two targets.