Comparison of basal ganglia regions across murine brain atlases using metadata models and the Waxholm Space.

The murine basal ganglia regions are targets for research into complex brain functions such as motor control and habit formation. However, there are several ways to name and annotate these regions, posing challenges for interpretation and comparison of data across studies. Here, we give an overview of basal ganglia terms and boundaries in the literature and reference atlases, and describe the criteria used for annotating these regions in the Waxholm Space rat brain atlas. We go on to compare basal ganglia annotations in stereotaxic rat brain atlases and the Allen Mouse brain Common Coordinate Framework to those in the Waxholm Space rat brain atlas. We demonstrate and describe considerable differences in the terms and boundaries of most basal ganglia regions across atlases and their versions. We also register information about atlases and regions in the openMINDS metadata framework, facilitating integration of data in neuroscience databases. The comparisons of terms and boundaries across rat and mouse atlases support analysis and interpretation of existing and new data from the basal ganglia.

Common limbic and frontal-striatal disturbances in patients with obsessive compulsive disorder, panic disorder and hypochondriasis.

BACKGROUND: Direct comparisons of brain function between obsessive compulsive disorder (OCD) and other anxiety or OCD spectrum disorders are rare. This study aimed to investigate the specificity of altered frontal-striatal and limbic activations during planning in OCD, a prototypical anxiety disorder (panic disorder) and a putative OCD spectrum disorder (hypochondriasis). METHOD: The Tower of London task, a 'frontal-striatal' task, was used during functional magnetic resonance imaging measurements in 50 unmedicated patients, diagnosed with OCD (n=22), panic disorder (n=14) or hypochondriasis (n=14), and in 22 healthy subjects. Blood oxygen level-dependent (BOLD) signal changes were calculated for contrasts of interest (planning versus baseline and task load effects). Moreover, correlations between BOLD responses and both task performance and state anxiety were analysed. RESULTS: Overall, patients showed a decreased recruitment of the precuneus, caudate nucleus, globus pallidus and thalamus, compared with healthy controls. There were no statistically significant differences in brain activation between the three patient groups. State anxiety was negatively correlated with dorsal frontal-striatal activation. Task performance was positively correlated with dorsal frontal-striatal recruitment and negatively correlated with limbic and ventral frontal-striatal recruitment. Multiple regression models showed that adequate task performance was best explained by independent contributions from dorsolateral prefrontal cortex (positive correlation) and amygdala (negative correlation), even after controlling for state anxiety. CONCLUSIONS: Patients with OCD, panic disorder and hypochondriasis share similar alterations in frontal-striatal brain regions during a planning task, presumably partly related to increased limbic activation.

[Impulse control disorders in Parkinson's disease]. / Impulscontrolestoornissen bij de ziekte van Parkinson en de relatie tot andere stoornissen binnen het impulsieve-compulsieve spectrum.

BACKGROUND: Parkinson's disease is characterised not only by the classic triad of bradykinesia, rigidity and tremor, but also by the frequent occurrence of various non-motor symptoms such as the impulse control disorders (pathological gambling, hypersexuality, compulsive buying, binge eating, punding and dopamine dependency). AIM: To increase insight into the clinical presentation, risk factors, treatment and the underlying pathophysiological mechanisms of impulse control disorders in Parkinson's disease. METHOD: Relevant literature was reviewed. RESULTS: Impulse control disorders belong to an important group of neuropsychiatric disorders that occur at some point in 5-10% of patients with Parkinson's disease. They generally occur in conjunction with dopaminergic medication and can have a marked social, relational and/ or financial impact. CONCLUSION: Early recognition of impulse control disorders in Parkinson's disease is important and a close collaboration between the neurologist and the psychiatrist is essential in order to ensure correct diagnosis and the best possible treatment. Impulse control disorders in Parkinson's disease show considerable phenomenological overlap with other repetitive behaviours within the impulsive-compulsive spectrum of disorders to which the obsessive-compulsive disorders and addiction disorders belong. The overlap can possibly be explained by a shared pathophysiological mechanism involving an imbalance between the direct and indirect pathways of the dorsal and ventral frontal-striatal circuits.

Anatomical evidence for direct connections between the shell and core subregions of the rat nucleus accumbens.

The nucleus accumbens is thought to subserve different aspects of adaptive and emotional behaviors. The anatomical substrates for such actions are multiple, parallel ventral striatopallidal output circuits originating in the nucleus accumbens shell and core subregions. Several indirect ways of interaction between the two subregions and their associated circuitry have been proposed, in particular through striato-pallido-thalamic and dopaminergic pathways. In this study, using anterograde neuroanatomical tracing with Phaseolus vulgaris-leucoagglutinin and biotinylated dextran amine as well as single-cell juxtacellular filling with neurobiotin, we investigated the intra-accumbens distribution of local axon collaterals for the identification of possible direct connections between the shell and core subregions. Our results show widespread intra-accumbens projection patterns, including reciprocal projections between specific parts of the shell and core. However, fibers originating in the core reach more distant areas of the shell, including the rostral pole (i.e. the calbindin-poor part of the shell anterior to the core) and striatal parts of the olfactory tubercle, than those arising in the shell and projecting to the core. The latter projections are more restricted to the border region between the shell and core. The density of the fiber labeling within both the shell and core was very similar. Moreover, specific intrinsic projections within shell and core were identified, including a relatively strong projection from the rostral pole to the rostral shell, reciprocal projections between the rostral and caudal shell, as well as projections within the core that have a caudal-to-rostral predominance. The results of the juxtacellular filling experiments show that medium-sized spiny projection neurons and medium-sized aspiny neurons (most likely fast-spiking) contribute to these intra-accumbens projections. While such neurons are GABAergic, the intrastriatal projection patterns indicate the existence of lateral inhibitory interactions within, as well as between, shell and core subregions of the nucleus accumbens.

Bilateral control of brain activity by dopamine D1 receptors: evidence from induction patterns of regulator of G protein signaling 2 and c-fos mRNA in D1-challenged hemiparkinsonian rats.

Recent reports show that striatal dopamine D1-type receptors from one side of the normal rat brain can control brain activity (as measured by c-fos induction) on both sides of the brain. However, this phenomenon has not yet been studied in the presence of sensitized dopamine D1-type receptors. Here we address this issue by investigating the extent to which dopamine D1-type receptors control brain activation in rats with unilaterally sensitized dopamine D1-type receptors. Gene induction assays were used to identify activated regions from midbrain to forebrain in unilaterally 6-hydroxydopamine lesioned (hemiparkinsonian) rats challenged with the full dopamine D1-type agonist SKF82958 (3 mg/kg, 0.5 and 2 h). The genes used are c-fos, the proven neuronal activity marker, and Regulator of G protein Signaling 2, a gene we propose as a marker of signaling homeostasis. SKF82958-mediated induction of both genes is greatly enhanced in hemiparkinsonian rats compared with shams, in both the lesioned and the intact hemisphere. For example, in the denervated caudate-putamen at 2 h postinjection, this enhancement is more than 80-fold for c-fos and up to 20-fold for Regulator of G protein Signaling 2; for the intact side this is 35-fold for c-fos and 27-fold for Regulator of G protein Signaling 2. Cortical induction of c-fos and Regulator of G protein Signaling 2 was generalized to most neocortical regions and was essentially equivalent in both the denervated and intact hemispheres. Interestingly, hippocampal structures also showed strong bilateral induction of both genes. This overall pattern of brain activation can be accounted for by the basal-ganglia thalamocortical and hippocampal circuits which both contain hemisphere-crossing connections and which can be initially activated in the lesioned hemisphere. Some regions, such as the intact striatum or the CA1 region, showed relatively low c-fos induction and relatively high Regulator of G protein Signaling 2 induction, possibly indicating that these regions are engaged in unusually strong signaling regulation activities. Our results show that, besides basal ganglia-thalamocortical circuits, dopamine D1-type-mediated brain activation in hemiparkinsonian rats also involves hippocampal circuits.

Intrinsic connections of the cingulate cortex in the rat suggest the existence of multiple functionally segregated networks.

The cingulate cortex is a functionally and morphologically heterogeneous cortical area comprising a number of interconnected subregions. To date, the exact anatomy of intracingulate connections has not been studied in detail. In the present study we aimed to determine the topographical and laminar characteristics of intrinsic cingulate connections in the rat, using the anterograde tracers Phaseolus vulgaris-leucoagglutinin and biotinylated dextran amine. For assessment of these data we further refined and compared the existing cytoarchitectonic descriptions of the two major cingulate constituents, the anterior cingulate and retrosplenial cortices. The results of this study demonstrate that rostral areas, i.e. the infralimbic and prelimbic cortices and the rostral one third of the dorsal anterior cingulate cortex are primarily interconnected with each other and not with other cingulate areas. The caudal two thirds of the dorsal anterior cingulate cortex project to the caudal part of the ventral anterior cingulate cortex, whereas the entire ventral anterior cingulate cortex projects to only the mid-rostro-caudal part of the dorsal anterior cingulate cortex. Dense reciprocal connections exist between the remaining, i.e. the supracallosal parts of the anterior cingulate and retrosplenial cortices with a general rostro-caudal topography, in the sense that the rostral part of the anterior cingulate cortex and caudal part of the retrosplenial cortex are interconnected and the same holds true for the caudal part of the anterior cingulate cortex and rostral part of the retrosplenial cortex. This topographical pattern of intracingulate connections relates to the results of several functional studies, suggesting that specific cingulate functions depend on a number of interconnected cingulate subregions. Through their intricate associational connections, these subregions form functionally segregated networks.

C-fos activation patterns in rat prefrontal cortex during acquisition of a cued classical conditioning task.

The prefrontal cortex (PFC) is known to be involved in associative learning; however, its specific role in acquisition of cued classical conditioning has not yet been determined. Furthermore, the role of regional differences within the PFC in the acquisition of cued conditioning is not well described. These issues were addressed by exposing rats to either one or four sessions of a cued classical conditioning task, and subsequently examining c-fos immunoreactivity in various areas of the PFC. Differences in patterns of c-fos immunopositive nuclei were found when comparing the PFC areas examined. No significant differences were found between rats presented with a temporally contingent conditioned stimulus (CS) light and food (paired groups) and those presented with the same stimuli temporally non-contingently (unpaired groups). In lateral and orbital PFC, both the paired and unpaired groups showed more c-fos immunopositive nuclei than control groups exposed only to the behavioral setup (context exposed groups), and all groups showed a drop in c-fos immunopositive nuclei from session 1 to session 4. In dorsal medial PFC, no differences were seen between the paired, unpaired and context exposed groups. These groups did, however, differ from naive animals, an effect that was not seen in the ventral medial PFC. The results of this study do not support a role for the PFC in the acquisition of a cued classical conditioning task. The differences seen between paired, unpaired and context exposed groups in orbital and lateral PFC could be due to contextual conditioning or reward-related effects.

Detailed localization of regulator of G protein signaling 2 messenger ribonucleic acid and protein in the rat brain.

Regulator of G protein signaling (RGS) proteins are a recently identified family of proteins which dampen G protein-coupled receptor-mediated signaling by accelerating the intrinsic GTPase activity of Galpha subunits of heterotrimeric G proteins. More than 20 different RGSs have been identified and at least 10 are expressed in the CNS. The present study describes in detail the localization in the rat brain of one member of this family, RGS2. The distribution of RGS2 mRNA and protein has been studied in parallel by performing in situ hybridization and immunoautoradiography on adjacent rat brain sections. Our localization study reveals that RGS2 mRNA and protein are widely expressed in the brain. Protein and mRNA are mostly colocalized such as in neocortex, piriform cortex, caudate-putamen, septum, hippocampus, locus coeruleus. Some mismatches were also observed such as presence of mRNA but not protein in the paraventricular nucleus, the substantia nigra pars compacta and the red nucleus, suggesting that RGS2 protein is present in neuronal projections. Previous reports describing an induction of RGS2 mRNA in the rat striatum after psychostimulants (amphetamine, cocaine) led us to focus on the distribution of RGS2 in the basal ganglia circuitry. The absence of RGS2 mRNA and protein in the globus pallidus suggests that RGS2 would play its regulatory role more in the direct (striatonigral) than in the indirect (striatopallidal) striatal output pathway. In addition, to delineate the implication of RGS2 in pre- and/or postsynaptic functions in the basal ganglia, we performed lesions of the nigrostriatal pathway by 6-hydroxydopamine (6-OHDA) and striatal quinolinic acid lesions. The 6-OHDA lesion did not modify RGS2 mRNA or protein levels in the caudate-putamen whereas the intrastriatal quinolinic acid infusion caused a marked reduction of RGS2 mRNA and protein in the lesioned zone. These data indicate that RGS2 is predominantly expressed in intrinsic striatal neurons. Moreover, the absence of detectable change in RGS2 expression after injections of 6-OHDA suggests also that RGS2 is not primarily involved in the hypersensitization of postsynaptic dopamine receptors observed after lesion of the nigrostriatal pathway.

Dynorphin modulates dopamine D1-receptor mediated turning behavior in 6-hydroxydopamine-lesioned rats.

We investigated if the potentiated turning response to a challenge with the partial dopamine D1 receptor agonist SKF-38393, as seen after priming with L-dihydroxyphenylalanine (DOPA) of unilaterally 6-hydroxydopamine-lesioned rats, can be modulated by infusion of dynorphin A (1-17) in the striatum. Seventeen days after the 6-hydroxydopamine lesion, rats received intrastriatal dynorphin (0. 08 or 3.85 microg) followed by L-DOPA (50 mg/kg i.p.) and were challenged 3 days later with SKF-38393 (3.0 mg/kg s.c.). Compared to controls, the lower dose of dynorphin caused an earlier onset of turning, while the higher dose decreased the response to SKF-38393. These findings suggest a dose-dependent modulatory role for striatal dynorphin in L-DOPA-priming of a D1-mediated behavioral response.

Leads for the development of neuroprotective treatment in Parkinson's disease and brain imaging methods for estimating treatment efficacy.

Patients suffering from Parkinson's disease display severe and progressive deficits in motor behavior, predominantly as a consequence of the degeneration of dopaminergic neurons, located in the mesencephalon and projecting to striatal regions. The cause of Parkinson's disease is still an enigma. Consequently, the pharmacotherapy of Parkinson's disease consists of symptomatic treatment, with in particular L-dihydroxyphenylalanine (L-DOPA) and/or dopamine receptor agonists. These induce a dramatic initial improvement. However, serious problems gradually develop during long-term treatment. Therefore, a more rational, c.q. causal treatment is needed which requires the introduction of compounds ameliorating the disease process itself. The development of such compounds necessitates (1) more information on the etiopathogenesis, i.e., the cascade of events that ultimately leads to degeneration of the dopaminergic neurons, and (2) brain imaging methods, to estimate the extent of the degeneration of the dopaminergic neurons in the living patient. This is not only important for the early diagnosis, but will also allow to monitor the effectiveness of alleged neuroprotective compounds on a longitudinal base. In this paper, etiopathogenic mechanisms are highlighted along the line of the oxidative stress hypothesis and within this framework, attention is mainly focused on the putative role of glutathione, dopamine auto-oxidation and phase II biotransformation enzymes. Especially, drugs able to increase the activity of phase II biotransformation enzymes seem to elicit a broad-spectrum (neuro)protective response and look very promising leads for the development of neuroprotective treatment strategies in Parkinson's disease. New developments in brain imaging methods (single photon emission computed tomography (SPECT) and positron emission tomography (PET)) to visualize the integrity of the striatal dopaminergic neurons in humans are highlighted as well. Especially, the introduction of radioligands that bind selectively to the dopamine transporter seems to be a significant step forward for the early diagnosis of Parkinson's disease. Performing these brain imaging studies with fixed time intervals does not only create the possibility to follow the degeneration rate of the dopaminergic neurons in Parkinson's disease but also provides the opportunity to estimate therapeutic effects of putative neuroprotective agents in the individual patient.

Convergence and segregation of ventral striatal inputs and outputs.

The ventral striatum, which prominently includes the nucleus accumbens (Acb), is a heterogeneous area. Within the Acb of rats, a peripherally located shell and a centrally situated core can be recognized that have different connectional, neurochemical, and functional identities. Although the Acb core resembles in many respects the dorsally adjacent caudate-putamen complex in its striatal character, the Acb shell has, in addition to striatal features, a more diverse array of neurochemical characteristics, and afferent and efferent connections. Inputs and outputs of the Acb, in particular of the shell, are inhomogeneously distributed, resulting in a mosaical arrangement of concentrations of afferent fibers and terminals and clusters of output neurons. To determine the precise relationships between the distributional patterns of various afferents (e.g., from the prefrontal cortex, the basal amygdaloid complex, the hippocampal formation, and the midline/intralaminar thalamic nuclei) and efferents to the ventral pallidum and mesencephalon, neuroanatomical anterograde and retrograde tracing experiments were carried out. The results of the double anterograde, double retrograde, and anterograde/retrograde tracing experiments indicate that various parts of the shell (dorsomedial, ventromedial, ventral, and lateral) and the core (medial and lateral) have different input-output characteristics. Furthermore, within these Acb regions, various populations of neurons can be identified, arranged in a cluster-like fashion, onto which specific sets of afferents converge and that project to particular output stations, distinct from the input-output relationships of neighboring, cluster-like neuronal populations. These results support the idea that the nucleus accumbens may consist of a collection of neuronal ensembles with different input-output relationships and, presumably, different functional characteristics.

Integration and segregation of limbic cortico-striatal loops at the thalamic level: an experimental tracing study in rats.

The frontal lobe and the basal ganglia are involved in a number of parallel, functionally segregated circuits. Information is thought to pass from distinct parts of the (pre)frontal cortex, via the striatum, the pallidum/substantia nigra and the thalamus, back to the premotor/prefrontal cortices. Currently, different views exist as to whether these circuits are to be considered as open or closed loops, as well as to the degree of interconnection between different circuits. The main goal of the present study is to answer some of these questions for the limbic corticostriatal circuits. The latter circuits involve the nucleus accumbens, the ventral pallidum/dorsomedial substantia nigra pars reticulata, the medial parts of the mediodorsal and ventromedial thalamic nuclei and the prefrontal cortex. Within the nucleus accumbens, a core and a shell region are recognized on the basis of anatomical and functional criteria. The shell of the nucleus accumbens projects predominantly to the mediodorsal, the midline and the reticular thalamic nuclei via the ventral pallidum, whereas the core reaches primarily the medial part of the ventromedial thalamic nucleus, the intralaminar and mediodorsal thalamic nuclei via a relay in the dorsomedial substantia nigra pars reticulata. By means of double labeling experiments with injections of anterograde tracers in both the ventral pallidum and the substantia nigra of rats, we were able to demonstrate that circuits involving the shell and the core of the nucleus accumbens remain largely segregated at the level of the thalamus. Only restricted areas of overlap of ventral pallidal and reticular nigral projections occur in the mediodorsal and ventromedial thalamic nuclei, which allows for a limited degree of integration, at the thalamic level, of information passing through the two circuits.

Regional and cellular distribution of serotonin 5-hydroxytryptamine2a receptor mRNA in the nucleus accumbens, olfactory tubercle, and caudate putamen of the rat.

This paper describes the regional and cellular distribution of serotonin 5-hydroxytryptamine2a (5-HT2a) receptor mRNA in (sub)regions of the rat striatum by using in situ hybridization. Our results indicate that 5-HT2a mRNA is distributed heterogeneously in this brain region. Regional densitometry of autoradiograms from striatal sections hybridized with isotope-labeled cRNA probes showed that mRNA levels were highest in the olfactory tubercle, lower in the nucleus accumbens, and lowest in the caudate-putamen. In the nucleus accumbens, the average mRNA levels in the shell were higher than those in the core. These data suggest a particular relevance for the 5-HT2a receptor for olfactory tubercle- and shell-related functions. Therefore, in the nucleus accumbens and the olfactory tubercle, the cellular localization of 5-HT2a mRNA was investigated by determining the colocalization of 5-HT2a mRNA with enkephalin mRNA or dynorphin mRNA. 5-HT2a mRNA was found in enkephalinergic as well as dynorphinergic neurons. Thus, there does not seem to be a differential distribution of this receptor in the output routes of the ventral striatum. In all of the subregions investigated (core, medial shell, and lateral shell of the nucleus accumbens and the olfactory tubercle), only subpopulations of the total enkephalinergic and dynorphinergic populations were found to contain 5-HT2a mRNA. For enkephalin, the percentage colocalization was higher in the lateral shell (61%) compared with the other subregions (38-45%). For dynorphin, the percentage colocalization was higher in the olfactory tubercle (68%) than in the other subregions (34-43%). The differences in (sub)regional mRNA levels and in colocalization with opioids suggest a considerable regional differentiation in the effects of 5-HT2a-mediated neurotransmission in the striatum.

Morphology of the human internal vertebral venous plexus: a cadaver study after intravenous Araldite CY 221 injection.

Reviewing the literature on the vascular anatomy of the spinal epidural space, it appeared that the knowledge of the internal vertebral venous plexus is limited. Injection studies of the entire internal vertebral venous plexus after application of modern techniques, to the best of our knowledge, have never been performed. Based on the clinical importance of these structures, it was decided to study the human vertebral venous system after Araldite CY 221 injection, in order to update the morphological characteristics of the internal vertebral venous system. The vertebral venous systems of ten fresh human cadavers, between 64 and 93 years of age, were injected with Araldite CY 221 mixture. All cadavers were dissected and the posterior and anterior internal vertebral venous plexuses were studied in detail. The anterior part of the internal vertebral venous plexus is fairly constant. On the contrary, the posterior internal vertebral venous plexus showed a striking segmental and interindividual variability. In the thoracic area, two types of traversing veins are observed. Both types show a somewhat symmetrical "inversed V" configuration. No anatomical valves were observed. Nevertheless, anterograde flushing (via the femoral veins) of the vertebral venous system appeared to proceed much faster than retrograde flushing (via the superior vena cava). The classical picture of the internal vertebral venous plexus appears a simplification of the actual situation. Especially in the posterior part, segmental and interindividual differences are prominent. The preferential direction of the flow during flushing suggests the presence of functional valves, which are probably located in the thoracic part of the posterior internal vertebral venous plexus, resulting from the typical shape of the veins in this area. This might explain the difficulties with imaging of the posterior part of the internal vertebral venous plexus in vitro as well as in vivo. Further study is needed to determine whether the configuration of the posterior internal vertebral venous plexus in younger individuals is different, compared with the presently studied aged subjects.

Injection of dye into neurones in rat and human post-mortem brain, in combination with acetylcholinesterase histochemistry: permanent preparations.

We describe a protocol for the intracellular injection of dye into neurones in thick sections of fixed, post-mortem rat and human brain tissue. To render the sections with the intracellularly injected neurones permanent, they are sectioned again, and the resulting subsections are either immunocytochemically treated or stained histochemically for acetylcholinesterase (AChE) activity. The resultant preparations can be stored at room temperature for prolonged periods. Background staining produced by accumulation of erythrocytes in blood vessels is greatly reduced or virtually eliminated by exposure of the sections to ultraviolet radiation prior to the intracellular injection. The pattern of AChE staining is not affected by this procedure. This ability to stain sections according to a histochemical AChE procedure after the intracellular injection of dyes into striatal neurons opens the possibility to study the relationship of neuronal dendritic trees with the striosome/matrix compartmental boundaries in post-mortem (human) brain tissue of Huntington's disease patients.

The accumbens: beyond the core-shell dichotomy.

This article highlights recent discoveries related to the accumbens and closely associated structures, with special reference to their importance in neuropsychiatry. The development of "striatal patches" in the accumbens is reviewed in a series of pictures. Neuronal ensembles are discussed as potentially important functional-anatomical units. Attention is also drawn to recent discoveries related to the neuronal circuits that the primate accumbens establishes with the mesencephalic dopamine system. On the basis of histological and neurochemical differences, the accumbens has been divided into core and shell compartments. In the context of this article, the shell, which is an especially diversified part of the accumbens, is the subject of special attention because of its close relation to the extended amygdala and distinctive response to antipsychotic and psychoactive drugs.

The anatomical relationships of the prefrontal cortex with limbic structures and the basal ganglia.

This paper briefly discusses the anatomical criteria that have been used to delineate the prefrontal cortex (PFC) from the (pre)motor cortical areas in the frontal lobe. Single anatomical criteria, such as cytoarchitecture, connectivity with the mediodorsal thalamic nucleus or a dopaminergic innervation, are insufficient to unequivocally define the PFC. It is argued that, with respect to a number of structural aspects, the prefrontal and the (pre)motor cortical areas must be viewed as a continuum, whereas a (functional) differentiation is based on the type of information that is being processed in different parts of the frontal lobe. The involvement of the PFC, like the premotor cortex, in a number of basal ganglia-thalamocortical circuits may be interpreted in the same way. The paper also summarizes the organization of the inputs from midline/intralaminar thalamic nuclei, the basal amygdaloid complex and the hippocampus into the PFC-ventral striatal system. The results of tracing studies in rats indicate that these thalamic and limbic inputs both at the level of the PFC and the ventral striatum show various patterns of convergence and segregation. This leads to the conclusion that the PFC-ventral striatal system consists of a number of smaller modules.

Patterns of overlap and segregation between insular cortical, intermediodorsal thalamic and basal amygdaloid afferents in the nucleus accumbens of the rat.

Regions of the prefrontal cortex that project to the nucleus accumbens in the rat receive input from midline thalamic and basal amygdaloid nuclei which also project to the same striatal region as their prefrontal cortical target. For example, the prelimbic cortex projects to the medial nucleus accumbens, and receives input from the paraventricular thalamic nucleus and the parvicellular basal amygdala. These latter two areas also project to the medial nucleus accumbens. It has been shown that afferents from the prelimbic cortex, the paraventricular thalamic nucleus and the parvicellular basal amygdala to the nucleus accumbens overlap or are separated in the nucleus accumbens, depending upon their position in the shell and core. The dorsal agranular insular cortex, the intermediodorsal thalamic nucleus and the magnocellular basal amygdaloid nucleus terminate in the lateral part of the nucleus accumbens and adjacent ventral part of the caudate-putamen. The intermediodorsal thalamic nucleus and the magnocellular basal amygdaloid nucleus reach both the dorsal agranular insular cortex and the lateral nucleus accumbens, and thus appear positioned to influence the prefrontal corticostriatal system at cortical and striatal levels. However, all three afferent systems have a heterogeneous distribution within this striatal region, and whether these projections actually reach the same areas is unknown. We investigated the patterns of separation and overlap in the nucleus accumbens between dorsal agranular insular cortical, magnocellular basal amygdaloid and intermediodorsal thalamic afferents with respect to the histochemical features of the nucleus. Techniques allowing the detection of two different anterograde tracers, or a single anterograde tracer and Calbindin-D28k immunoreactivity, in the same tissue sections were used. The results demonstrate that the afferents from the dorsal agranular insular area and the intermediodorsal thalamic nucleus avoid the shell of the lateral nucleus accumbens, which receives strong inputs from the magnocellular basal amygdala. In the matrix of the core and the ventral part of the caudate-putamen, fibers from the superficial layers of the dorsal agranular insular area overlap precisely with afferents from the intermediodorsal nucleus. In the patches, projections from the deep layers of the dorsal agranular insular cortex coincide with those from the magnocellular basal amygdala. The present findings have implications for the compartmental structure of the nucleus accumbens and provide novel insights into the organizational principles of prefrontal corticostriatal circuits.