Dopamine D1 receptors have subcellular distributions conducive to interactions with prodynorphin in the rat nucleus accumbens shell

Activation of dopamine (DA) D1 receptors (D1Rs) in the nucleus accumbens (Acb) markedly affects the levels of prodynorphin, the precursor of aversion-associated dynorphin peptides. The location of prodynorphin, specifically as related to the dopaminergic inputs and D1Rs in the Acb, is fundamental for establishing the physiologically relevant sites. To determine these sites, we examined the electron microscopic dual-immunolabeling of prodynorphin and D1R or tyrosine hydroxylase (TH), a marker of catecholamine terminals in the rat Acb shell. This subregion is targeted by mesolimbic dopaminergic inputs affecting reward-aversion responses and locomotor activity. Prodynorphin was prominently localized to large (100-200 nm) granular aggregates in somatodendritic and axonal profiles, some of which expressed dynorphin A/B. In somata and dendrites, prodynorphin was often found in punctate clusters in the cytoplasm. Of the total prodynorphin-labeled dendrites, approximately 63% expressed D1Rs, which were largely located on the plasma membranes. In comparison with dendrites, many more axon terminals contained prodynorphin, although only 15% of these terminals contained D1R-labeling. Prodynorphin terminals formed symmetric synapses with D1R-labeled or unlabeled dendrites, and also apposed TH-containing axon terminals. Our results provide ultrastructural evidence that in the Acb shell, the prodynorphin is available for cleavage to physiologically active peptides in both dendrites and terminals of neurons that express D1Rs. They also indicate that dynorphin peptides have distributions that would enable their participation in modulation of DA release or D1R-mediated postsynaptic responses in Acb shell neurons.     

Major coexpression of kappa-opioid receptors and the dopamine transporter in nucleus accumbens axonal profiles.

The behavioral effects of psychostimulants, which are produced at least in part through inhibition of the dopamine transporter (DAT), are modulated by kappa-opioid receptors (KOR) in the nucleus accumbens (Acb). Using electron microscopic immunocytochemistry, we reveal that in the Acb KOR labeling is mainly, and DAT immunoreactivity is exclusively, presynaptic. From 400 KOR-labeled presynaptic structures, including axon terminals, intervaricosities, and small axons, 51% expressed DAT and 29% contacted another population of terminals exclusively labeled for DAT. Within axonal profiles that contained both antigens, DAT and KOR were prominently localized to plasma membrane segments that showed overlapping distributions of the respective immunogold-silver and immunoperoxidase markers. KOR labeling was also localized to membranes of small synaptic vesicles in terminals with or without DAT immunoreactivity. In addition, from 24 KOR-immunoreactive dendritic spines 42% received convergent input from DAT-containing varicosities and unlabeled terminals forming asymmetric, excitatory-type synapses. Our results provide the first ultrastructural evidence that in the Acb, KOR is localized to strategic sites for involvement in the direct presynaptic release and/or reuptake of dopamine. These data also suggest a role for KOR in the presynaptic modulation of other neurotransmitters and in the postsynaptic excitatory responses of single spiny neurons in the Acb. Dual actions on dopamine terminals and their targets in the Acb may account for KOR-mediated attenuation of drug reinforcement and sensitization.     

High-affinity neurotensin receptors in the rat nucleus accumbens: subcellular targeting and relation to endogenous ligand

Neurotensin is present in selective mesolimbic dopaminergic projections to the nucleus accumbens (NAc) shell but also is synthesized locally in this region and in the motor-associated NAc core. We examined the electron microscopic immunolabeling of the high-affinity neurotensin receptor (NTR) and neurotensin in these subdivisions of rat NAc to determine the sites for receptor activation and potential regional differences in distribution. Throughout the NAc, NTR immunoreactivity was localized discretely within both neurons and glia. NTR-labeled neuronal profiles were mainly axons and axon terminals with diverse synaptic structures, which resembled dopaminergic and glutamatergic afferents, as well as collaterals of inhibitory projection neurons. These terminals had a significantly higher numerical density in the NAc core than in the shell but were prevalent in both regions, suggesting involvement in both motor and limbic functions. In each region, neurotensin was detected in a few NTR-immunoreactive axon terminals and in terminals that formed symmetric, inhibitory type synapses with NTR-labeled somata and dendrites. The NTR labeling, however, was not seen within these synapses and, instead, was localized to segments of dendritic and glial plasma membranes often near excitatory type synapses. Neuronal NTR immunoreactivity also was associated with cytoplasmic tubulovesicles and nuclear membranes. Our results suggests that, in the NAc shell and core, NTR is targeted mainly to presynaptic sites, playing a role in the regulated secretion and/or retrograde signaling in diverse, neurotransmitter-specific neurons. The findings also support a volume mode of neurotensin actions, specifically affecting excitatory transmission through activation of not only axonal but also dendritic and glial NTR.     

Y1 receptors in the nucleus accumbens: Ultrastructural localization and association with neuropeptide Y

Neuropeptide Y (NPY) is present in aspiny neurons in the nucleus accumbens (NAc), which also contains moderate levels of ligand binding and mRNA for the Y1 receptor. To determine the potential functional sites for receptor activation, we examined the electron microscopic immunocytochemical localization of antipeptide antisera against the Y1 receptor in the rat NAc. We also combined immunogold and immunoperoxidase labeling to show that, in this region, Y1 receptors are present in certain somatodendritic and axonal profiles that contain NPY or that appose NPY containing neurons. The Y1-like immunoreactivity (Y1-LI) was seen occasionally along plasma membranes but was associated more commonly with smooth endoplasmic reticulum (SER) and tubulovesicular organelles in somata and dendrites of spiny and aspiny neurons. The mean density of immunoreactive dendrites and spines per unit volume was greater in the “motor-associated” core than in the shell of the NAc. Y1-LI was also seen in morphologically heterogenous axon terminals, including those forming asymmetric excitatory-type synapses, and in selective astrocytic processes near this type of junction. We conclude that Y1 receptors play a role in autoregulation of NPY-containing neurons but are also likely to be internalized along with endogenous NPY in NAc. Our results also implicate Y1 receptors in the NAc in post- and presynaptic effects of NPY and in glial functions involving excitatory neurotransmission. In addition, they suggest involvement of Y1 receptors in determining the output of a select population of neurons associated with motor control in the NAc core. J. Neurosci. Res. 52:54–68, 1998. © 1998 Wiley-Liss, Inc.     

Effect of lead exposure on dopaminergic receptors in rat striatum and nucleus accumbens

Haloperidol- and sulpiride-displaceable [3H]spiroperidol binding and the dopamine-inhibited adenylate cyclase were measured in rats chronically exposed to lead acetate. Haloperidol-displaceable [3H]spiroperidol binding was unmodified while sulpiride-displaceable binding was increased in striatum and decreased in nucleus accumbens. In addition, the decrease of sulpiride-displaceable binding in nucleus accumbens was paralleled by a reduced ability of bromocriptine to inhibit cAMP formation in presence of the D1 receptor antagonist SCH 23390. The results support the concept that in vivo lead treatment affects dopaminergic receptors and that the binding sites labelled by [3H]spiroperidol displaced by haloperidol may be different from those which recognize sulpiride.     

Projections from the paraventricular nucleus of the thalamus to the rat prefrontal cortex and nucleus accumbens shell: ultrastructural characteristics and spatial relationships with dopamine afferents

The paraventricular nucleus of the thalamus (PVT) participates in the functional integration of limbic cortical and striatal circuitry. In the rat, the PVT projects to the deep layers of the medial prefrontal cortex (PFC) and to the shell of the nucleus accumbens (NAc). However, the synaptic organization of PVT afferents within these regions remains undescribed. Furthermore, although dopamine (DA) modulates excitatory glutamate transmission in both areas, possible anatomic substrates for specific DA modulation of PVT inputs have not yet been investigated. To address these issues, immunoperoxidase labeling for tyrosine hydroxylase (TH) in DA axons was combined with anterograde tract-tracing, either by biotinylated dextran amine (BDA) labeled with immunogold-silver or by degeneration after lesions of the PVT. In both regions, and with either tracing method, PVT terminals formed primarily asymmetric axospinous synapses; in the NAc, a proportion of PVT terminals also synapsed onto dendrites. PVT profiles in both regions were often seen in direct apposition to TH-immunoreactive axons; this association was more evident in the NAc where the DA innervation is denser. Within the PFC, PVT profiles and TH-labeled axons were occasionally apposed to the same dendrites, but synaptic specializations were not typically seen at these seeming points of convergence. Within the NAc, PVT profiles occasionally made synapses onto spines and distal dendrites that received convergent synapses from TH-immunoreactive varicosities. These findings represent the first demonstration of postsynaptic convergence between DA and thalamic afferents to a striatal region and are consistent with direct synaptic modulation of PVT transmission by DA in the NAc but not the PFC.     

Specificity in the efferent projections of the nucleus accumbens in the rat: Comparison of the rostral pole projection patterns with those of the core and shell

The efferent connections of the rostral pole of the rat accumbens, where distinct core and shell subterritories can not be identified, were examined with the aid of the anterogradely transported plant lectin, Phaseolus vulgaris-leucoagglutinin (PHA-L), for comparison with the previously reported projection patterns of the accumbal core and shell. Injection sites and transported PHA-L were evaluated with the aid of reference to adjacent sections processed to display substance P or calbindin 28 kD immunoreactivities, i.e., markers that demonstrate the core and shell. Lateral parts of the rostral pole gave rise to a “core-like” projection system that involved the rostroventral globus pallidus, subcommissural ventral pallidum, entopeduncular nucleus and an adjacent part of the lateral hypothalamus, lateral ventral tegmental area, dorsal pars compacta, and structures in the lateral mesencephalic tegmentum and central grey. The medial part of the rostral pole gave rise to a “shell-like” innervation of the subcommissural ventral pallidum, lateral preoptic region, lateral hypothalamus, ventral tegmental area, dorsalmost pars compacta, retrorubral field, lateral midbrain tegmentum, and central grey. In contrast to the large numbers of axon varicosities observed through the entire length of lateral hypothalamus following shell injections, dense accumulations of axon collaterals and varicosities in hypothalamus were limited to the levels of origin of the stria medullaris bundle and entopeduncular nucleus and to the posterlateral region following medial injections. The medial part of the rostral pole contributed some projections to preoptic and sublenticular regions, but not to the bed nucleus of the stria terminalis. Noteworthy concentrations of calbindin immunoreactive cells observed in the lateral rostral pole correlate with the origin of the “basal ganglia-like” projection system, provoking the speculation that ventral striatal calbindin immunoreactive cells contribute principally to basal ganglia-like projections while cells lacking calbindin immunoreactivity contribute to the innervation of hypothalamus and midbrain tegmentum. © 1993 Wiley-Liss, Inc.     

Ultrastructure of serotonin-immunoreactive terminals in the core and shell of the rat nucleus accumbens: Cellular substrates for interactions with catecholamine afferents

The nucleus accumbens is composed of a core region involved in motor functions and a shell region implicated in emotional and motivational processes. Both of these regions receive serotonin- and dopamine-containing afferents. We examined whether the serotonin innervation or relation to catecholamine (mainly dopamine) axons in the nucleus accumbens shows common features or specializations corresponding to the noted functional differences in core and shell subregions. To address this question, we examined the ultrastructure of serotonin-containing axons and their relation to catecholamine-containing afferents in either the core or shell of the nucleus accumbens. Single coronal sections through the rat forebrain were processed for immunoperoxidase labeling of serotonin and immunogold silver labeling of tyrosine hydroxylase, the catecholamine-synthesizing enzyme. Varicose processes showing peroxidase product for serotonin by light microscopy were confirmed to be axons and terminals by electron microscopy. In a quantitative analysis of serotonin-immunoreactive terminals forming one or more contacts in single sections, some common features were observed. For the core (n = 120) and the shell (n = 82), 41% formed synaptic junctions with unlabeled dendrites, 75% were in apposition with unlabeled terminals, which often formed asymmetric junctions, and 20% were in apposition with axons or terminals containing tyrosine hydroxylase. Thus, in both the core and shell of the nucleus accumbens, serotonin terminals synapse on postsynaptic neurons and are likely to modulate or be modulated by presynaptic interactions with excitatory axons forming asymmetric junctions and by catecholaminergic afferents. Marked differences in the morphology of serotonin axons were also seen in the core versus shell of the nucleus accumbens. By light microscopy, serotonin-immunoreactive axons were thicker and more varicose than those found in the core. Ultrastructural analysis confirmed that, in contrast to the core, serotonin-immunoreactive axons and terminals in the shell were larger in cross-sectional diameter size (0.7 μm vs. 0.3 μm). Additionally, serotonin axon terminals in the shell contained more numerous immunoreactive large dense core vesicles and more frequently formed symmetric as opposed to asymmetric contacts with dendrites. The larger size and more numerous dense core vesicles in serotonin-immunoreactive terminals in the shell support the concept that serotonin or co-existing neurotransmitter may be more tonically released in the shell versus core of the nucleus accumbens. © 1993 Wiley-Liss, Inc.     

The nucleus accumbens shell in reinstatement and extinction of drug seeking

The contexts where drugs are self‐administered have important control over relapse and extinction of drug‐seeking behavior. The nucleus accumbens shell (AcbSh) is essential to this contextual control over drug‐seeking behavior. It has been consistently implicated in both the expression of context‐induced reinstatement and the expression of extinction, across a variety of drug classes and other rewards. Here, we review the evidence linking AcbSh to the extinction and reinstatement of drug seeking. We consider whether this dual role can be linked to known heterogeneities in AcbSh cell types, their major afferents, and their major efferents. We show that although these heterogeneities are each important and can determine extinction vs. reinstatement, they do not seem adequate to explain the body of findings from the behavioral literature. Rather, we suggest that this functional specialization of AcbSh may be more profitably viewed in terms of the segregation and compartmentalization of AcbSh channels.     

Origin of noradrenergic afferents to the shell subregion of the nucleus accumbens: anterograde and retrograde tract-tracing studies in the rat

The nucleus accumbens (NAcc) can be subdivided into `core' and `shell' based on anatomical connections and histochemical markers. Previous studies have demonstrated dopamine-β-hydroxylase immunoreactive (DBH-ir) fibers in the NAcc shell, but the source of these noradrenergic (NE) afferents has not been determined. Therefore, we have investigated in detail the anatomy of NE afferents to this subregion. Dual immunohistochemistry for DBH and substance P demonstrated numerous DBH-ir fibers in the caudal NAcc shell. Neurons projecting to the NAcc were identified with Fluoro-Gold (FG) or cholera toxin B (CTb) retrograde tracing and tyrosine hydroxylase (TH) immunohistochemistry. Single- and double-labeled neurons were observed in the A2 and A1 NE cell groups following FG injections into the caudal NAcc shell. Numerous FG and CTb single-labeled neurons were found in the rostral locus coeruleus (LC), subcoeruleus and pericoerulear dendritic region, with an occasional double-labeled neuron in the LC. Few labeled neurons were seen in the brainstem after FG injections into the NAcc core, consistent with the lack of DBH-ir in this subterritory. To confirm these results, injections of Phaseolus vulgaris leucoagglutinin or biotinylated dextran amine were made into the LC or nucleus tractus solitarius (NTS). Virtually no labeled fibers were observed in the NAcc following injections into central LC. However, fibers were observed in the NAcc shell after injections in the NTS. These results indicate that the primary source(s) of NE afferents to the NAcc shell is the A2 region of the NTS, with lesser contributions from A1 and LC.     

Ultrastructural characteristics of enkephalin-immunoreactive boutons and their postsynaptic targets in the shell and core of the nucleus accumbens of the rat

The present study compared the ultrastructural morphology of enkephalin-immunoreactive boutons and their postsynaptic targets in different territories of the nucleus accumbens in the rat. The synaptic bouton profiles were identified by antibodies directed against [leu⁵]enkephalin. Ninety-five percent of the synaptic contacts were symmetric in configuration and the remaining 5% were asymmetric. Axosomatic contacts comprised 6% of all enkephalin-immunoreactive junctions and were distributed equally in all parts of the nucleus. Most (76%) synaptic terminals contacted dendrites but they contacted proportionally fewer dendrites in the shell (71%) than in the core (78%). Moreover, enkephalin-immunoreactive synaptic boutons in the shell (19%) and caudal enkephalin-rich areas (17%) of the core contacted twice as many spines than in the remaining parts of the core (8.5%). In the core, long pallidum-like dendrites were occasionally found ensheathed in enkephalin-immunoreactive terminal boutons. We conclude that the differential arrangement of enkephalinergic contacts in the shell and core could have important functional consequences, especially when considered in relation to other known morphological and neurochemical differences between these regions. © 1993 Wiley-Liss, Inc.     

The distribution and origin of the calretinin-containing innervation of the nucleus accumbens of the rat

The nucleus accumbens of the rat consists of several subregions that can be distinguished on the basis of histochemical markers. For example, the calcium-binding protein calbindin D28k is a useful marker of the core compartment of the nucleus accumbens. Calretinin, another calcium-binding protein, is found in a dense fibre plexus in the accumbal shell and septal pole regions. The source of the accumbal calretinin innervation is not known. We examined the distribution of calretinin in the nucleus accumbens and used tract-tracing and lesion methods to determine the source of this calretinin innervation. Intense calretinin immunoreactivity was present in the medial shell, but the density of calretinin axons diminished sharply in the ventrolateral shell. Regions of dense calretinin immunostaining and those areas with calbindin-like immunoreactive cell bodies were generally segregated in the nucleus accumbens, although some overlap in the transition region between the core and shell was seen. Small clusters of calretinin-immunoreactive fibres were seen in the core, where they were restricted to calbindin-negative patches. Injections of the anterograde tracer biotinylated dextran amine into the paraventricular thalamic nucleus labelled fibres in calretinin-rich regions of the accumbens. Conversely, injections of Fluoro-gold into the accumbal shell retrogradely labelled numerous cells in the paraventricular thalamic nucleus that were calretinin-immunoreactive. Electrolytic lesions of the paraventricular thalamic nucleus reduced calretinin levels in the shell by approximately 80%. These data indicate that the calretinin innervation of the nucleus accumbens is derived primarily from the thalamic paraventricular nucleus, and marks accumbal territories that are largely complementary to those defined by calbindin.     

Morphine withdrawal-induced morphological changes in the nucleus accumbens

Morphine withdrawal produces a hypofunction of mesencephalic dopamine neurons that impinge upon medium spiny neurons (MSN) of the forebrain. After chronic treatment (from 20 to 140 mg/kg of morphine twice a day over 14 days at escalating doses) rats were withdrawn from chronic morphine spontaneously and pharmacologically. In these two distinct conditions we studied the effects of withdrawal on the morphology of MSN of the core and shell of the nucleus accumbens (Nacc). MSN were stained with the Golgi-Cox procedure and analysed by a confocal laser-scanning microscope (CLSM). Our analysis shows that, shell and core MSN differed significantly for perikarya size and spine density, and the various morphine treatments did not affect the perikarya morphometry. Both spontaneous and naloxone-induced withdrawal produced a similar reduction in spine density in MS shell neurons, as compared with MS core neurons. This effect is selectively localized at the level of second order dendritic trunks where afferents converge. By contrast, spine density counts of accumbens MSN from rats chronically treated with morphine, did not reveal any change. Collectively, the results of the present study are twofold: (i) spontaneous and pharmacologically precipitated withdrawal, but not chronic morphine per se, affects spine density of target structures of a reduced mesolimbic dopamine transmission, and (ii) the reduction of spine density in second order dendritic trunks is selectively segregated in the MSN of the shell of the Nacc. In conclusion, morphine withdrawal dramatically alters spine density, selectively in second order dendritic trunks of Nacc shell MSN, thereby further impoverishing the already abated dopamine (DA) transmission. This is in line with recent views suggesting the hypodopaminergic state as a cardinal feature of opioid dependence.     

Pre- and postsynaptic sites for serotonin modulation of GABA-containing neurons in the shell region of the rat nucleus accumbens

The shell of the nucleus accumbens receives a dense serotonergic innervation and contains abundant gamma-aminobutyric acid (GABA)-immunoreactive neurons. Moreover, serotonin (5-hydroxytryptamine: 5-HT) and GABA have been implicated in a variety of common motivational and motor-related functions partially ascribed to this brain area. We used immunoelectron microscopy of antisera directed against 5-HT and GABA in the same section of tissue to examine whether there were cellular substrates that might indicate more specific sites for functional interactions involving these transmitters in the shell region of the rat nucleus accumbens. Immunogold-silver labeling for GABA was localized to perikarya, dendrites, axons and axon terminals, whereas immunoperoxidase labeling for 5-HT was restricted to axons and axon terminals. Approximately half (187/366) of the 5-HT-immunoreactive axon terminals apposed or formed synaptic junctions with postsynaptic neurons. These junctions were mainly of the symmetric-type (83/187) characteristic of inhibitory transmitters, and were equally prevalent on dendrites with and without detectable gold-silver labeling for GABA. Of the 187 5-HT-labeled axon terminals with recognized synaptic contacts, 36% also showed convergence on a common dendrite with a GABA-labeled axon terminal. In addition, 5-HT- and GABA-immunoreactive axon terminals were commonly (83/366) identified in direct apposition to one another. Within a single plane of section, 41% of the apposed GABA-immunoreactive axon terminals formed symmetric-type junctions with dendrites or somata, whereas, the apposed 5-HT-labeled axon terminals rarely showed postsynaptic contacts. These results indicate that 5-HT-containing axon terminals may postsynaptically inhibit GABAergic neurons and their targets within the shell of the rat nucleus accumbens. Additionally, our results strongly suggest that, in this brain region, appositions between 5-HT and GABA axons and axon terminals may facilitate presynaptic interactions between these transmitter systems. © 1996 Wiley-Liss, Inc.     

Ultrastructural localization of delta-opioid receptors in the rat caudate-putamen nucleus during postnatal development: relation to synaptogenesis

During development, delta-opioid receptors (DORs) in the rat caudate-putamen nucleus (CPN) appear later than mu-opioid receptors (MORs), whose developmental pattern specifically relates to synaptogenesis. We used electron microscopic immunocytochemistry to determine whether there are also age-related changes in subcellular localization of DORs in the rat CPN. Sections from postnatal day (P) 0-P30 and adult dorsomedial CPN were immunogold-silver labeled to examine the plasmalemmal and cytoplasmic distribution of these receptors. In addition, immunoperoxidase labeling was used to determine the numerical density of synapses relative to DOR-labeled profiles. Immunolabeling for DOR was undetectable at P0, light at P5, and dense from P10 onward. The labeling during P5-P10 was mainly localized in somatodendritic profiles but also was readily seen in axon terminals, most of which formed asymmetric synapses with dendrites. From P15, a few immunogold particles were seen in contact with postsynaptic densities in spines, and the proportion of these particles significantly increased in P30 and adult CPN. Other particles were localized in the cytoplasm of dendrites and terminals without significant age-related changes. Stereological analysis showed that compared with labeled dendritic shafts and spines, labeled axon terminals have a closer correlation with synapse formation. These results are in marked contrast with MORs, which show an age-related increase in association with dendritic plasma membrane and a good correlation in the developmental pattern of MOR-labeled spines with synapse formation (Wang et al. [2003] Neuroscience 118:695-708). Together, our results suggest receptor-type specific roles for endogenous opioids acting at both pre- and postsynaptic sides in the developing CPN.     

Direct comparison of projections from the central amygdaloid region and nucleus accumbens shell

Certain neurochemical and connectional characteristics common to extended amygdala and the nucleus accumbens shell suggest that the two represent a single functional-anatomical continuum. If this is so, it follows that the outputs of the two structures should be substantially similar. To address this, projections from the caudomedial shell and central nucleus of the amygdala, a key extended amygdala structure, were demonstrated in Sprague-Dawley rats with different anterograde axonal tracers processed separately to exhibit distinguishable brown and blue-black precipitates. The caudomedial shell projection is strong in the ventral pallidum and along the medial forebrain bundle through the lateral preopticohypothalamic continuum into the ventral tegmental area, distal to which it thins abruptly. The central nucleus projects strongly to the bed nucleus of the stria terminalis and the sublenticular extended amygdala, but substantially to the lateral hypothalamus only at levels behind the rostral part of the entopeduncular nucleus. Innervation of the ventral tegmental area by the central amygdala is minimal, but the lateral one-third of the substantia nigra, pars compacta and an adjacent lateral part of the retrorubral field receive substantial central amygdala input. Central amygdaloid projections are robust in caudal brainstem sites, such as the reticular formation, parabrachial nucleus, nucleus of the solitary tract and dorsal vagal complex, all of which receive little input from the accumbens. The substantial differences in the output systems of the caudomedial shell of accumbens and central amygdala suggest that the two represent distinct functional-anatomical systems.     

Ventral mesopontine projections of the caudomedial shell of the nucleus accumbens and extended amygdala in the rat: double dissociation by organization and development.

The shell of the nucleus accumbens and central division of the extended amygdala are telencephalic structures that influence motor activity and lately have been regarded by some as components of a single functional-anatomic continuum. Each has a highly differentiated internal organization and output system and distinct pharmacologic responses however, and it is thus likely that each subserves distinct contributions to behavior. In this investigation, nucleus accumbens and extended amygdala outputs were compared by using retrograde tracing in adult and postnatal rats. Fluoro-Gold, when injected into the ventral tegmental area, produced substantial retrograde labeling in the adult nucleus accumbens shell, but only trivial amounts in the central division of the extended amygdala. Injection sites in the lateral mesopontine tegmentum produced robust labeling in the central extended amygdala but little in the nucleus accumbens. The projections of extended amygdala were substantially developed by postnatal day 1, whereas those of the caudomedial shell of the nucleus accumbens only reached the ventral tegmental area by approximately postnatal day 6. Few neurons projecting from the caudomedial shell of the accumbens to the ventral tegmental area were observed even at postnatal day 21. In consideration of the reported importance of the nucleus accumbens, particularly the caudomedial shell, in neural processing related to reward and motivation and the central nervous system response to antipsychotic drugs, it may be important to determine whether processes occurring during the protracted postnatal development of the caudomedial shell are vulnerable to destructive circumstances, such as drug intoxication, maternal separation, or social isolation.     

Cellular sites for dynorphin activation of kappa-opioid receptors in the rat nucleus accumbens shell.

The nucleus accumbens (Acb) is prominently involved in the aversive behavioral aspects of kappa-opioid receptor (KOR) agonists, including its endogenous ligand dynorphin (Dyn). We examined the ultrastructural immunoperoxidase localization of KOR and immunogold labeling of Dyn to determine the major cellular sites for KOR activation in this region. Of 851 KOR-labeled structures sampled from a total area of 10,457 microm2, 63% were small axons and morphologically heterogenous axon terminals, 31% of which apposed Dyn-labeled terminals or also contained Dyn. Sixty-eight percent of the KOR-containing axon terminals formed punctate-symmetric or appositional contacts with unlabeled dendrites and spines, many of which received convergent input from terminals that formed asymmetric synapses. Excitatory-type terminals that formed asymmetric synapses with dendritic spines comprised 21% of the KOR-immunoreactive profiles. Dendritic spines within the neuropil were the major nonaxonal structures that contained KOR immunoreactivity. These spines also received excitatory-type synapses from unlabeled terminals and were apposed by Dyn-containing terminals. These results provide ultrastructural evidence that in the Acb shell (AcbSh), KOR agonists play a primary role in regulating the presynaptic release of Dyn and other neuromodulators that influence the output of spiny neurons via changes in the presynaptic release of or the postsynaptic responses to excitatory amino acids. The cellular distribution of KOR complements those described previously for the reward-associated mu- and delta-opioid receptors in the Acb shell.