Amygdala efferents form inhibitory-type synapses with a subpopulation of catecholaminergic neurons in the rat Nucleus tractus solitarius

Amygdalo-nigral inputs target dopaminergic and GABAergic neurons in the primate: a view from dendrites and soma

The central nucleus (CeN) of the amygdala is an important afferent to the DA system that mediates motivated learning. We previously found that CeN terminals in nonhuman primates primarily overlap the elongated lateral VTA (parabrachial pigmented nucleus, PBP, A10), and retrorubral field(A8) subregion. Here, we examined CeN afferent contacts on cell somata and proximal dendrites of DA and GABA neurons, and distal dendrites of each, using confocal and electron microscopy (EM) methods, respectively. At the soma/proximal dendrites, the proportion of TH+ and GAD1+ cells receiving at least one CeN afferent contact was surprisingly similar (TH = 0.55: GAD1=0.55 in PBP; TH = 0.56; GAD1 =0.51 in A8), with the vast majority of contacted TH+ and GAD1+ soma/proximal dendrites received 1-2 contacts. Similar numbers of tracer-labeled terminals also contacted TH-positive and GAD1-positive small dendrites and/or spines (39% of all contacted dendrites were either TH- or GAD1-labeled). Overall, axon terminals had more symmetric (putative inhibitory) axonal contacts with no difference in the relative distribution in the PBP versus A8, or onto TH+ versus GAD1+ dendrites/spines in either region. The striking uniformity in the amygdalonigral projection across the PBP-A8 terminal field suggests that neither neurotransmitter phenotype nor midbrain location dictates likelihood of a terminal contact. We discuss how this afferent uniformity can play out in recently discovered differences in DA:GABA cell densities between the PBP and A8, and affect specific outputs.

Significance statement

The amygdala's central nucleus (CeN) channels salient cues to influence both appetitive and aversive responses via DA outputs. In higher species, the broad CeN terminal field overlaps the parabrachial pigmented nucleus ('lateral A10') and the retrorubral field (A8). We quantified terminal contacts in each region on DA and GABAergic soma/proximal dendrites and small distal dendrites. There was striking uniformity in contacts on DA and GABAergic cells, regardless of soma and dendritic compartment, in both regions. Most contacts were symmetric (putative inhibitory) with little change in the ratio of inhibitory to excitatory contacts by region.We conclude that post-synaptic shifts in DA-GABA ratios are key to understanding how these relatively uniform inputs can produce diverse effects on outputs.

Effects of high-fat diet and gastric bypass on neurons in the caudal solitary nucleus

Bariatric surgery is an effective treatment for obesity that involves both peripheral and central mechanisms. To elucidate central pathways by which oral and visceral signals are influenced by high-fat diet (HFD) and Roux-en-Y gastric bypass (RYGB) surgery, we recorded from neurons in the caudal visceral nucleus of the solitary tract (cNST, N=287) and rostral gustatory NST (rNST,N=106) in rats maintained on a HFD and lab chow (CHOW) or CHOW alone, and subjected to either RYGB or sham surgery. Animals on the HFD weighed significantly more than CHOW rats and RYGB reversed and then blunted weight gain regardless of diet. Using whole-cell patch clamp recording in a brainstem slice, we determined the membrane properties of cNST and rNST neurons associated with diet and surgery. We could not detect differences in rNST neurons associated with these manipulations. In cNST neurons, neither the threshold for solitary tract stimulation nor the amplitude of evoked EPSCs at threshold varied by condition; however suprathreshold EPSCs were larger in HFD compared to chow-fed animals. In addition, a transient outward current, most likely an IA current, was increased with HFD and RYGB reduced this current as well as a sustained outward current. Interestingly, hypothalamic projecting cNST neurons preferentially express IA and modulate transmission of afferent signals (Bailey, '07). Thus, diet and RYGB have multiple effects on the cellular properties of neurons in the visceral regions of NST, with potential to influence inputs to forebrain feeding circuits.

Comparisons of terminal densities of cardiovascular function-related projections from the amygdala subnuclei

The amygdala is important in higher-level control of cardiovascular functions. In this study, we compared cardiovascular-related projections among the subnuclei of the amygdala. Biotinylated dextran amine was injected into the central, medial, and basolateral nuclei of the amygdala, and the distributions and densities of anterograde-labeled terminal boutons were analyzed. We found that the medial, basolateral, and central nuclei all had projections into the cardiovascular-related areas of the hypothalamus. However, only the central nucleus had a significant direct projection into the medulla. By contrast, the medial nucleus had limited projections, and the basolateral nucleus had no terminals extending into the medulla. We concluded that the medial, central, and basolateral nuclei of the amygdala may influence cardiovascular-related nuclei through monosynaptic connections with cardiovascular-related nuclei in the hypothalamus and medulla.

Central neural responses to restraint stress are altered in rats with an early life history of repeated brief maternal separation

Repeated brief maternal separation (i.e. 15 min daily, MS15) of rat pups during the first one to two postnatal weeks enhances active maternal care received by the pups and attenuates their later behavioral and neuroendocrine responses to stress. In previous work, we found that MS15 also alters the developmental assembly and later structure of central neural circuits that control autonomic outflow to the viscera, suggesting that MS15 may alter central visceral circuit responses to stress. To examine this, juvenile rats with a developmental history of either MS15 or no separation (NS) received microinjection of retrograde neural tracer, FluoroGold (FG), into the hindbrain dorsal vagal complex (DVC). After 1 week, FG-injected rats and surgically intact littermates were exposed to either a 15-min restraint stress or an unrestrained control condition, and then perfused 1 h later. Brain tissue sections from surgically intact littermates were processed for Fos alone or in combination with phenotypic markers to examine stress-induced activation of neurons within the paraventricular nucleus of the hypothalamus (PVN), bed nucleus of the stria terminalis (BNST), and hindbrain DVC. Compared to NS controls, MS15 rats displayed less restraint-induced Fos activation within the dorsolateral BNST (dBNST), the caudal PVN, and noradrenergic neurons within the caudal DVC. To examine whether these differences corresponded with altered neural inputs to the DVC, sections from tracer-injected rats were double-labeled for FG and Fos to quantify retrogradely labeled neurons within hypothalamic and limbic forebrain regions of interest, and the proportion of these neurons activated after restraint. Only the dBNST displayed a significant effect of postnatal experience on restraint-induced Fos activation of DVC-projecting neurons. The distinct regional effects of MS15 on stress-induced recruitment of neurons within hypothalamic, limbic forebrain, and hindbrain regions has interesting implications for understanding how early life experience shapes the functional organization of stress-responsive circuits.

Amygdala connections with jaw, tongue and laryngo-pharyngeal premotor neurons

As the central nucleus (CE) is the only amygdaloid nucleus to send axons to the pons and medulla, it is thought to be involved in the expression of conditioned responses by accessing hindbrain circuitry generating stereotypic responses to aversive stimuli. Responses to aversive oral stimuli include gaping and tongue protrusion generated by central pattern generators and other premotor neurons in the ponto-medullary reticular formation. We investigated central nucleus connections with the reticular formation by identifying premotor reticular formation neurons through the retrograde trans-synaptic transport of pseudorabies virus (PRV) inoculated into masseter, genioglossus, thyroarytenoid or inferior constrictor muscles in combination with anterograde labeling of CE axons with biotinylated dextran amine. Three dimensional mapping of PRV infected premotor neurons revealed specific clusters of these neurons associated with different oro-laryngo-pharyngeal muscles, particularly in the parvicellular reticular formation. CE axon terminals were concentrated in certain parvicellular clusters but overall putative contacts were identified with premotor neurons associated with all four oro-laryngo-pharyngeal muscles investigated. We also mapped the retrograde trans-synaptic spread of PRV through the various nuclei of the amygdaloid complex. Medial CE was the first amygdala structure infected (4 days post-inoculation) with trans-synaptic spread to the lateral CE and the caudomedial parvicellular basolateral nucleus by day 5 post-inoculation. Infected neurons were only very rarely found in the lateral capsular CE and the lateral nucleus and then at only the latest time points. The data demonstrate that the CE is directly connected with clusters of reticular premotor neurons that may represent complex pattern generators and/or switching elements for the generation of stereotypic oral and laryngo-pharyngeal movements during aversive oral stimulation. Serial connections through the amygdaloid complex linked with the oro-laryngo-pharyngeal musculature appear quite distinct from those believed to sub-serve fear responses, suggesting there are distinct "channels" for the acquisition and expression of particular conditioned behaviors.

Hindbrain noradrenergic A2 neurons: diverse roles in autonomic, endocrine, cognitive, and behavioral functions

Central noradrenergic (NA) signaling is broadly implicated in behavioral and physiological processes related to attention, arousal, motivation, learning and memory, and homeostasis. This review focuses on the A2 cell group of NA neurons, located within the hindbrain dorsal vagal complex (DVC). The intra-DVC location of A2 neurons supports their role in vagal sensory-motor reflex arcs and visceral motor outflow. A2 neurons also are reciprocally connected with multiple brain stem, hypothalamic, and limbic forebrain regions. The extra-DVC connections of A2 neurons provide a route through which emotional and cognitive events can modulate visceral motor outflow and also a route through which interoceptive feedback from the body can impact hypothalamic functions as well as emotional and cognitive processing. This review considers some of the hallmark anatomical and chemical features of A2 neurons, followed by presentation of evidence supporting a role for A2 neurons in modulating food intake, affective behavior, behavioral and physiological stress responses, emotional learning, and drug dependence. Increased knowledge about the organization and function of the A2 cell group and the neural circuits in which A2 neurons participate should contribute to a better understanding of how the brain orchestrates adaptive responses to the various threats and opportunities of life and should further reveal the central underpinnings of stress-related physiological and emotional dysregulation.

Central amygdaloid axon terminals are in contact with retrorubral field neurons that project to the parvicellular reticular formation of the medulla oblongata in the rat

The retrorubral field (RRF) contains numerous dopaminergic neurons and projects to the parvicellular reticular formation (RFp) of the medullary and pontomedullary brainstem, where many premotor neurons project to the orofacial motor nuclei. To know how the amygdala affects the RRF-RFp pathway in the rat, we first examined the synaptic organization between the central amygdaloid nucleus (CeA) fibers and the RFp-projecting RRF neurons by using combined anterograde and retrograde tracing techniques. After ipsilateral injections of biotinylated dextran amine (BDA) into the CeA and Fluoro-gold (FG) into the RFp, the prominent overlapping distribution of BDA-labeled axon terminals and FG-labeled neurons was found in the lateral part of the RRF ipsilateral to the injection sites, where the BDA-labeled axon terminals made symmetrical synapses with somata and dendrites of the FG-labeled neurons. Using a combination of retrograde tracing and immunohistochemistry for tyrosine hydroxylase (TH), we secondly demonstrated that the RFp-projecting RRF neurons were immunonegative for TH. Using a combination of anterograde tracing and immunohistochemistry for glutamic acid decarboxylase (GAD), we finally revealed that the CeA axon terminals in the RRF were immunoreactive for GAD. The present results suggest that GABAergic CeA neurons may exert inhibitory influences on non-dopaminergic RRF neurons that project to the RFp in the control of orofacial movements closely related to emotional behavior.

Nicotine modulation of stress-related peptide neurons

Nicotine has been shown to activate stress-related brain nuclei, including the paraventricular nucleus of the hypothalamus (PVN) and the central nucleus of the amygdala (CEA), through complex mechanisms involving direct and indirect pathways. To determine the neurochemical identities of rat brain neurons which are activated by a low dose (0.175 mg/kg) of nicotine given 30 minutes before sacrifice, we have used single- and double-label in situ hybridization. Neuronal activation was quantified by localization of (35)S-labeled probe for the immediate early gene, c-fos. Corticotrophin releasing factor (CRF), enkephalin (ENK), and dynorphin (DYN) mRNAs were colocalized using a colorimetric, digoxigenin-labeled probe. Film autoradiographic studies showed that nicotine significantly increased c-fos mRNA expression in both PVN and CEA. Pretreatment with the centrally acting nicotinic antagonist, mecamylamine (1 mg/kg), blocked nicotine's effects, whereas pretreatment with the peripherally acting antagonist, hexamethonium (5 mg/kg), did not, indicating that c-fos induction was mediated by a central nicotinic receptor. Double labeling studies showed that nicotine induced c-fos expression within CRF cells in the PVN, as well as in a small population of ENK cells, but not in PVN DYN cells. In contrast, there was no significant nicotine-induced increase in c-fos expression in CEA CRF or DYN cells, whereas nicotine treatment did increase c-fos expression within CEA ENK cells.

A light and electron microscopic analysis of the convergent insular cortical and amygdaloid projections to the posterior lateral hypothalamus in the rat, with special reference to cardiovascular function

The synaptic organization between and among the insular cortex (IC) axons, central amygdaloid nucleus (ACe) axons and posterolateral hypothalamus (PLH) neurons was investigated in the rat using double anterograde tracing and anterograde tracing combined with postembedding immunogold analysis. After ipsilateral injections of biotinylated dextran amine (BDA) into the IC and Phaseolus vulgaris-leucoagglutinin (PHA-L) into the ACe, the conspicuous overlapping distribution of BDA-labeled axon terminals and PHA-L-labeled axon terminals was found in the PLH region just medial to the subthalamic nucleus ipsilateral to the injection sites. At the electron microscopic level, approximately two-thirds of the IC terminals made synapses with small-sized dendrites and the rest did with dendritic spines of the PLH neurons, whereas about 79%, 16% and 5% of the ACe terminals established synapses with small- to medium-sized dendrites, somata, and dendritic spines, respectively, of the PLH neurons. In addition, the IC axon terminals contained densely packed round clear vesicles and their synapses were of asymmetrical type. On the other hand, most of the ACe terminals contained not only pleomorphic clear vesicles but also dense-cored vesicles and their synapses were of symmetrical type although some ACe terminals contained densely packed round clear vesicles and formed asymmetrical synapses. Most of the postsynaptic elements received synaptic inputs from the IC or ACe terminals, and some of single postsynaptic elements received convergent synaptic inputs from both sets of terminals. Furthermore, almost all the ACe terminals were revealed to be immunoreactive for gamma-aminobutyric acid (GABA), by using the anterograde BDA tracing technique combined with immunohistochemistry for GABA. The present data suggest that single PLH neurons are under the excitatory influence of the IC and/or inhibitory influence of the ACe in the circuitry involved in the regulation of cardiovascular functions.

Increased AMPA GluR1 receptor subunit labeling on the plasma membrane of dendrites in the basolateral amygdala of rats self-administering morphine

Glutamate-dependent synaptic plasticity is emerging as an important neural substrate of addiction. These drug-dependent neural adaptations may occur within brain systems that mediate reward, emotion, and cognitive function such as the amygdala complex. Modification of glutamate receptor targeting may be a key mechanism mediating neural plasticity; however, evidence for alteration of amygdala AMPA receptor localization in response to drug self-administration is lacking. High-resolution immunogold electron microscopic immunocytochemistry was used to compare surface and intracellular labeling of the calcium sensitive AMPA GluR1 receptor subunit in the basolateral (BLA) and central (CeA) nuclei of the amygdala in rats self-administering escalating doses of morphine or saline. Morphine self-administration was associated with regionally diverse effects on dendritic GluR1 targeting in the BLA and CeA. In the BLA of morphine self-administering animals, there was a significant increase in the proportion of immunogold particles for GluR1 on the plasma membrane of dendrites, particularly in association with extrasynaptic sites, which was most prominent in large (2-4 microm) profiles. In contrast, there were no significant differences in surface or intracellular immunogold labeling in the CeA between morphine self-administering and control animals. In both amygdala regions, GluR1 and the micro-opioid receptor, the major cellular target of morphine, were only infrequently colocalized. These results indicate that GluR1 targeting is a dynamic process that can be differentially affected in distinct amygdala regions in response to chronic self-administration of morphine. Homeostatic adaptations in the subcellular localization of calcium sensitive AMPA receptors within the BLA may be an important neural substrate for alterations in reward, autonomic function, and behavioral processes associated with opiate addiction.

Opposing influence of basolateral amygdala and footshock stimulation on neurons of the central amygdala

BACKGROUND

The basolateral complex (BLA) and the central nucleus of the amygdala (CeA) are believed to mediate the expression of affective responses. After affective learning, conditioned stimulus-related information is thought to be conveyed from the BLA to the CeA; the medial CeA (Cem), in turn, projects to hypothalamic and brainstem structures involved with induction of affective responses. Although the conditioned stimulus and unconditioned stimulus both evoke affective responses, the precise response often differs. It is unknown whether this difference is represented by distinct activity patterns of single Cem neurons. Furthermore, the nature of the interaction between the BLA and Cem is unknown.

METHODS

Using in vivo extracellular and intracellular recordings, we examined how the BLA affects the Cem and compared this with effects induced by footshock (unconditioned stimulus) in the same neurons.

RESULTS

Our results demonstrate that, contrary to conventional views, BLA stimulation primarily inhibits Cem neurons by a polysynaptic circuit, and show that single Cem neurons respond to both BLA input and footshock in an opposite manner.

CONCLUSIONS

These results demonstrate the predominantly inhibitory nature of the BLA-Cem interaction. These data further demonstrate the distinct cellular events that might lead to differential modulation of conditioned and unconditioned affective responses.

A GABAergic projection from the central nucleus of the amygdala to the parabrachial nucleus: an ultrastructural study of anterograde tracing in combination with post-embedding immunocytochemistry in the rat

To determine whether axonal terminals emanating from the central nucleus of amygdala (Ce) to the parabrachial nucleus (PBN) contain gamma-aminobutyric acid (GABA) as their neurotransmitter, an electron microscopic study was performed employing the combined techniques of WGA-HRP anterograde tracing and post-embedding immunocytochemistry for GABA. Our analysis distinguished a large population of GABA immunopositive axonal terminals from the Ce that exhibited symmetrical synaptic contacts with neurons in the lateral parabrachial nucleus. Additionally, most retrogradely labeled dendrites and perikarya received synaptic contacts from GABA immunoreactive terminals, with some of them originating from the Ce. The present study provides the first direct ultrastructural evidence for a monosynaptic, GABAergic link between Ce axons and neurons of the parabrachial nucleus via classical symmetrical synapses.

Possible mechanisms of involvement of the amygdaloid complex in the control of gastric motor function

Electrophysiological and neuroanatomical experiments on Wistar rats were performed to study the mechanisms of the modulatory influences of the amygdaloid nuclei on reflex motor activity in the stomach. Electrical stimulation of the central nucleus was accompanied by reproducible changes in the ongoing motor activity of the stomach in activity evoked by activation of the vagovagal reflex arc. The most marked, and predominantly inhibitory, effects were seen in response to stimulation of the medial part of the nucleus. Microinjections of the anterograde neuron marker Phaseolus vulgaris leucoagglutinin (PHA-L) into the central nucleus of the amygdala revealed the existence of direct descending projections from its dorsomedial part to the area containing cells of the vagosolitary complex, associated with performance of the vagovagal reflexes of the stomach. Electrical stimulation of this part of the central nucleus led to changes in neuron responses in the bulbar "gastric" center evoked by stimulation of the vagus nerve. These features may underlie one of the mechanisms of the amygdalar modulation of the reflex activity of the stomach.

Gene expression in autonomic areas of the medulla and the central nucleus of the amygdala in rats during and after space flight

During space flight astronauts show vestibular-related changes in balance, eye movements, and spontaneous and reflex control of cardiovascular, respiratory and gastrointestinal function, sometimes associated with space motion sickness. These symptoms undergo compensation over time. Here we used changes in the expression of two immediate-early gene (IEG) products to identify cellular and molecular changes occurring in autonomic brainstem regions of adult male albino rats killed at different times during the Neurolab Space Mission (STS-90). Both direct effects of gravitational changes, as well as indirect effects of gravitational changes on responses to light exposure were examined. Regions under the direct control of vestibular afferents such as the area postrema and the caudal part of the nucleus of the tractus solitarius (NTSC) were both directly and indirectly affected by gravity changes. These areas showed no changes in the expression of IEG products during exposure to microgravity with respect to ground controls, but did show a significant increase 24 h after return to 1 G (gravity). Exposure to microgravity significantly inhibited gene responses to light exposure seen after return to 1 G. A similar direct and indirect response pattern was also shown by the central nucleus of the amygdala, a basal forebrain structure anatomically and functionally related to the NTS. The rostral part of the NTS (NTSR) receives different afferent projections than the NTSC. This region did not show any direct gravity-related changes in IEG expression, but showed an indirect effect of gravity on IEG responses to light. A similar pattern was also obtained in the intermediate reticular nucleus and the parvocellular reticular nucleus. Two other medullary reticular structures, the dorsal and the ventral medullary reticular nuclei showed a less well defined pattern of responses that differed from those seen in the NTSC and NTSR. The short- and long-lasting molecular changes in medullary and basal forebrain gene expression described here are thought to play an important role in the integration of autonomic and vestibular signals that ultimately regulate neural adaptations to space flight.

The central nucleus of the amygdala modulates gut-related neurons in the dorsal vagal complex in rats

Using retrograde tract-tracing and electrophysiological methods, we characterized the anatomical and functional relationship between the central nucleus of the amygdala and the dorsal vagal complex. Retrograde tract-tracing techniques revealed that the central nucleus of the amygdala projects to the dorsal vagal complex with a topographic distribution. Following injection of retrograde tracer into the vagal complex, retrogradely labelled neurons in the central nucleus of the amygdala were clustered in the central portion at the rostral level and in the medial part at the middle level of the nucleus. Few labelled neurons were seen at the caudal level. Electrical stimulation of the central nucleus of the amygdala altered the basal firing rates of 65 % of gut-related neurons in the nucleus of the solitary tract and in the dorsal motor nucleus of the vagus. Eighty-one percent of the neurons in the nucleus of the solitary tract and 47 % of the neurons in the dorsal motor nucleus were inhibited. Electrical stimulation of the central nucleus of the amygdala also modulated the response of neurons in the dorsal vagal complex to gastrointestinal stimuli. The predominant effect on the neurons of the nucleus of the solitary tract was inhibition. These results suggest that the central nucleus of the amygdala influences gut-related neurons in the dorsal vagal complex and provides a neuronal circuitry that explains the regulation of gastrointestinal activity by the amygdala.

Alpha-2A-adrenergic receptors are present on neurons in the central nucleus of the amygdala that project to the dorsal vagal complex in the rat

The descending pathway between the central nucleus of the amygdala (CeA) and the dorsal vagal complex (DVC) is an important substrate for autonomic functions associated with emotion. Activity in this circuit is crucially modulated by catecholamines and agonists of the alpha-2A-adrenergic receptor (alpha(2A)-AR), which relieve cardiovascular and gastrointestinal symptoms associated with experience of aversive stimuli. The subcellular distribution of alpha(2A)-AR within the CeA, however, has not been characterized. It is also not known if any alpha(2A)-AR-expressing neurons in the CeA project to the dorsal vagal complex. In order to address these questions, we examined the immunocytochemical labeling of alpha(2A)-AR in the CeA of rats receiving microinjection of the retrograde tracer fluorogold (FG) into the dorsal vagal complex at the level of the area postrema, an area involved in cardiorespiratory and gastrointestinal functions. Of all alpha(2A)-AR-labeled profiles in the CeA, the majority were either dendrites (42%) or somata (24%). alpha(2A)-AR labeling was often present on the plasmalemma in dendrites and was mainly found in endosome-like organelles in somata. Of all alpha(2A)-AR immunoreactive somata, 62% also contained immunolabeling for FG and 23% of all dendrites also showed labeling for the retrograde tracer. The intracellular distribution of alpha(2A)-AR did not differ in somata or dendrites with or without detectable FG. The remaining singly labeled alpha(2A)-AR profiles consisted of axons (11%), axon terminals (12%), and glial processes (13%). In numerous instances, alpha(2A)-AR-labeled glia or axon terminals were apposed to DVC projecting neurons. Together, this evidence suggests that the principal site for alpha(2A)-AR activation is at extrasynaptic sites on dendrites of CeA neurons, many of which project to the DVC and also show endosomal receptor labeling. In addition, these results indicate that activation of alpha(2A)-AR in the CeA may influence the activity of DVC projecting neurons through indirect mechanisms, including changes in presynaptic transmitter release or glial function. These results suggest that alpha(2A)-AR agonists in the CeA may modulate numerous processes including stress-evoked autonomic reactions and feeding behavior.

Somatostatin immunoreactivity in axon terminals in rat nucleus tractus solitarii arising from central nucleus of amygdala: coexistence with GABA and postsynaptic expression of sst2A receptor

Axon terminals synapsing on neurones in the nucleus tractus solitarii (NTS) that originate from the central nucleus of the amygdala (CeA) have been shown to contain gamma-aminobutyric acid (GABA) immunoreactivity. Here we investigated whether such terminals also contain somatostatin (SOM), a neuropeptide found in axons distributed throughout the NTS and in somata in the CeA, and known to modulate cardiovascular reflexes when microinjected into the NTS. With fluorescence microscopy, SOM immunoreactivity was seen in the varicosities of some axons throughout the NTS that were anterogradely labelled with biotin dextran amine injected into the CeA. Such varicosities were frequently observed in close proximity to dendrites of NTS neurones that were immunoreactive for the SOM receptor sst(2A) subtype, and in many cases also for catecholamine synthesising enzymes. In the caudal, cardioregulatory zone of NTS, SOM immunoreactivity was localised by electron microscopic pre-embedding gold labelling to boutons containing dense-cored and clear pleomorphic vesicles and forming symmetrical synapses, mostly onto dendrites. Additional post-embedding gold labelling for GABA suggested that a subpopulation (29%) of GABAergic terminals sampled in this area of NTS contained SOM. Almost all boutons anterogradely labelled from the amygdala were GABA-immunoreactive (-IR) and 21% of these were SOM-IR. A similar proportion of these boutons (22%) formed synapses onto dendrites containing immunoreactivity for the SOM receptor sst(2A) subtype. These observations provide evidence that some of the GABAergic projection neurones in the CeA that inhibit baroreceptor reflex responses in the NTS in response to fear or emotional stimuli could release SOM, which might modulate the activity of NTS neurones via an action on sst(2A) receptors.

Opposing roles for medial and central amygdala in the initiation of noradrenergic cell responses to a psychological stressor

Psychological stressors trigger the activation of medullary noradrenergic cells, an effect that has been shown to depend upon yet-to-be-identified structures located higher in the brain. To test whether the amygdala is important in this regard, we examined the effects of amygdala lesions on noradrenergic cell responses to restraint, and also looked at whether any amygdala cells that respond to restraint project directly to the medulla. Ibotenic acid lesions of the medial amygdala completely abolished restraint-induced Fos expression in A1 and A2 noradrenergic cells. In contrast, lesions of the central amygdala actually facilitated noradrenergic cell responses to restraint. Tracer deposits in the dorsomedial (but not ventrolateral) medulla retrogradely labelled many cells in the central nucleus of the amygdala, but none of these cells expressed Fos in response to restraint. These data suggest for the first time that the medial amygdala is critical to the activation of medullary noradrenergic cells by a psychological stressor whereas the central nucleus exerts an opposing, inhibitory influence upon noradrenergic cell recruitment. The initiation of noradrenergic cell responses by the medial amygdala does not involve a direct projection to the medulla. Accordingly, a relay through some other structure, such as the hypothalamic paraventricular nucleus, warrants careful consideration.

Stress-induced relapse to drug seeking in the rat: role of the bed nucleus of the stria terminalis and amygdala

There is growing interest in the role that the bed nucleus of the stria terminalis (BNST) and central nucleus of the amygdala (CeA), components of the extended amygdala, play in drug addiction. Within the BNST and CeA, there is an extensive system of intrinsic, primarily GABAergic, interconnections known to synthesize a variety of neuropeptides, including corticotrophin-releasing factor (CRF). The actions of CRF at extrahypothalamic sites,including the BNST and CeA, have been implicated in stress responses and in the aversive effects of withdrawal from drugs of abuse. Most recently, we have shown a critical role for extrahypothalamic CRF in stress-induced reinstatement of drug seeking in rats. In attempting to determine which brain circuitry mediates the effect of stress on relapse and, more specifically, where in the brain CRF acts to initiate the behaviours involved in relapse, we focused on the BNST and CeA. In the present paper, we summarize studies we have conducted that explore the role of these brain sites in stress-induced relapse to heroin and cocaine seeking, and then consider how our findings can be understood within the more general context of what is known about the role of the BNST and CeA in stress-related and general approach behaviours, such as drug seeking.

The central nucleus of the amygdala; a conduit for modulation of HPA axis responses to an immune challenge?

Physical stressors such as infection, inflammation and tissue injury elicit activation of the hypothalamic-pituitary-adrenal (HPA) axis. This response has significant implications for both immune and central nervous system function. Investigations in rats into the neural substrates responsible for HPA axis activation to an immune challenge have predominantly utilized an experimental paradigm involving the acute administration of the pro-inflammatory cytokine interleukin- 1β (IL-1β). It is well recognized that medial parvocellular corticotrophin-releasing factor cells of the paraventricular nucleus (mPVN CRF) are critical in generating HPA axis responses to an immune challenge but little is known about how peripheral immune signals can activate and/or modulate the mPVN CRF cells. Studies that have examined the afferent control of the mPVN CRF cell response to systemic IL-1β have centred largely on the inputs from brainstem catecholamine cells. However, other regulatory neuronal populations also merit attention and one such region is a component of the limbic system, the central nucleus of the amygdala (CeA). A large number of CeA cells are recruited following systemic IL-lβ administration and there is a significant body of work indicating that the CeA can influence HPA axis function. However, the contribution of the CeA to HPA axis responses to an immune challenge is only just beginning to be addressed. This review examines three aspects of HPA axis control by systemic IL-1β: (i) whether the CeA has a role in generating HPA axis responses to systemic IL-1β, (ii) the identity of the neural connections between the CeA and mPVN CRF cells that might be important to HPA axis responses and(iii) the mechanisms by which systemic IL-Iβ triggers the recruitment of CeA cells.

A GABAergic projection from the central nucleus of the amygdala to the nucleus of the solitary tract: a combined anterograde tracing and electron microscopic immunohistochemical study

The central nucleus of the amygdala is involved in the modulation of autonomic, somatic and endocrine functions, as well as behavioural responses to stressful stimuli. Anatomical and physiological studies have suggested that this nucleus sends projections to the nucleus of the solitary tract, the primary site of termination of vagal and glossopharyngeal afferent fibres in the brain stem. To determine the neurochemical nature of the amygdaloid input to the nucleus of the solitary tract, anterograde tracing with biotinylated dextran amine was combined with post-embedding immunogold labelling for GABA and glutamate immunoreactivities and with pre-embedding labelling for the vesicular GABA transporter. Following injection of biotin dextran amine into the central nucleus of the amygdala, anterogradely labelled axons and varicosities were found throughout the rostrocaudal extent of the nucleus of the solitary tract, particularly in the medial, ventral and ventrolateral subnuclei. The anterogradely labelled terminals were found to make predominantly symmetrical synaptic contacts with dendrites, and occasionally onto cell bodies and dendritic spines, and to contain immunoreactivity for GABA and for the vesicular GABA transporter. Immunolabelling of serial sections with antibodies to glutamate showed that none of these axon terminals contained high enough densities of gold particle labelling to suggest that they contained other than low metabolic levels of glutamate immunoreactivity. These results provide conclusive evidence for a GABAergic pathway from the central nucleus of the amygdala to the nucleus of the solitary tract. This GABAergic projection may provide a substrate for inhibition of lower brain stem visceral reflexes, including baroreflex inhibition, through which the central nucleus of the amygdala could participate in cardiovascular regulation related to emotional behaviour and the defence reaction.

Distribution of parvalbumin, calretinin, and calbindin-D(28k) immunoreactivity in the rat amygdaloid complex and colocalization with gamma-aminobutyric acid

To understand the organization of inhibitory circuitries in the rat amygdala, the distribution of parvalbumin, calretinin, and calbindin immunoreactivity was investigated in the rat amygdaloid complex. Colocalization of various calcium-binding proteins with the inhibitory transmitter gamma-aminobutyric acid (GABA) was studied by using the mirror technique. Parvalbumin-immunoreactive (-ir) elements were located mostly in the deep amygdaloid nuclei, whereas the calretinin-ir and calbindin-ir staining were most intense in the cortical nuclei as well as in the central nucleus and the amygdalohippocampal area. Second, the distribution of immunopositive neurons largely parallelled the distribution of terminal and neuropil labeling. Third, immunostained neurons could be divided into four major morphologic types (types 1-4) based on the characteristics of the somata and the dendritic trees. The fourth lightly stained neuronal type that had a pyramidal GABA-negative soma was observed only in calretinin and calbindin preparations. Fourth, parvalbumin-ir terminals formed basket-like plexus and cartridges, which suggests that parvalbumin labels GABAergic inhibitory basket cells and axo-axonic chandelier cells, respectively. Colocalization studies indicated that 521 of 553 (94%) of parvalbumin-ir, 419 of 557 (75%) of calbindin-ir, and 158 of 657 (24%) of calretinin-ir neurons were GABA-positive in the deep amygdaloid nuclei. A high density of large GABA-negative calbindin-ir neurons was observed caudally in the medial division of the lateral nucleus and GABA-negative calretinin-ir neurons were observed in the magnocellular division of the accessory basal nucleus as well as in the intermediate and parvicellular divisions of the basal nucleus. These data suggest that in various amygdaloid areas, neuronal excitability is controlled by GABAergic neurons that contain different calcium-binding proteins. The appearance of basket-like plexus and cartridges in the parvalbumin preparations, but not in calretinin preparations, suggests that like in the hippocampus, the distribution of inhibitory terminals in the dendritic and perisomatic regions of postsynaptic neurons in the rat amygdala is organized in a topographic manner.

Catecholamine systems in the brain of vertebrates: new perspectives through a comparative approach

A comparative analysis of catecholaminergic systems in the brain and spinal cord of vertebrates forces to reconsider several aspects of the organization of catecholamine systems. Evidence has been provided for the existence of extensive, putatively catecholaminergic cell groups in the spinal cord, the pretectum, the habenular region, and cortical and subcortical telencephalic areas. Moreover, putatively dopamine- and noradrenaline-accumulating cells have been demonstrated in the hypothalamic periventricular organ of almost every non-mammalian vertebrate studied. In contrast with the classical idea that the evolution of catecholamine systems is marked by an increase in complexity going from anamniotes to amniotes, it is now evident that the brains of anamniotes contain catecholaminergic cell groups, of which the counterparts in amniotes have lost the capacity to produce catecholamines. Moreover, a segmental approach in studying the organization of catecholaminergic systems is advocated. Such an approach has recently led to the conclusion that the chemoarchitecture and connections of the basal ganglia of anamniote and amniote tetrapods are largely comparable. This review has also brought together data about the distribution of receptors and catecholaminergic fibers as well as data about developmental aspects. From these data it has become clear that there is a good match between catecholaminergic fibers and receptors, but, at many places, volume transmission seems to play an important role. Finally, although the available data are still limited, striking differences are observed in the spatiotemporal sequence of appearance of catecholaminergic cell groups, in particular those in the retina and olfactory bulb.

Presynaptic NMDA receptor subunit immunoreactivity in GABAergic terminals in rat brain

N-methyl-D-aspartate (NMDA) receptors are commonly found post-synaptically; they mediate fast excitatory neurotransmission in the central nervous system. In this study, we provide immunocytochemical data supporting the existence of presynaptic NMDA receptors in GABAergic terminals using polyclonal antisera raised against the C-terminus of the NMDAR1 subunit. At the light microscope level, rich plexuses of NMDAR1-positive varicose fibers were found in various nuclei in the basal forebrain (bed nucleus of stria terminalis, septum, parastrial nucleus, vascular organ of the lamina terminalis), thalamus (paraventricular nucleus, midline nuclei), and hypothalamus (parvocellular paraventricular nucleus, arcuate nucleus, preoptic nucleus, suprachiasmatic nucleus). In the brainstem, labeled fibers were much less abundant and were confined to the ventral tegmental area, periaqueductal gray, parabrachial nucleus, and locus coeruleus. At the electron microscope level, NMDAR1-immunoreactive terminals examined in the bed nucleus of stria terminalis, parvocellular paraventricular hypothalamic nucleus, and arcuate nucleus formed symmetric synapses, contained darkly stained large dense-core vesicles, and displayed gamma-aminobutyric acid (GABA) immunoreactivity. Terminals with similar ultrastructural features were found in the paraventricular thalamic nucleus. These findings demonstrate the existence of NMDAR1 subunit immunoreactivity in subsets of GABAergic terminals, which raises questions about the potential roles and mechanisms of activation of presynaptic NMDA heteroreceptors in the rat central nervous system. The pattern of distribution and ultrastructural features of these boutons suggest that they may arise from local GABAergic projections interconnecting a group of brain structures mediating stress responses and/or other endocrine, autonomic, and limbic functions.

Flurothyl-induced seizures in rats activate Fos in brainstem catecholaminergic neurons

Autonomic changes accompany seizures in both animals and humans. While ictal autonomic dysfunction can be life-threatening, the participating neural networks involved are poorly understood. In this study we examined the activation of Fos following generalized seizures in brainstem structures known to mediate autonomic function. Adult female rats were sacrificed 2 h after flurothyl-induced seizures. Double-immunostaining for c-Fos and dopamine-beta-hydroxylase (DBH), and c-Fos and phenylethanol-N-methyl-transferase (PNMT) were performed in brainstem slices. Numbers of DBH-labeled neurons expressing Fos-like immunoreactivity (FLI) (DBH/Fos) and PNMT labeled neurons expressing FLI (PNMT/Fos) were counted in the noradrenergic (A1, A2, A5, A7) and adrenergic (C1, C2) cell groups localized in pons and medulla oblongata. Among the experimental animals, the highest degree of co-localization of DBH/Fos neurons was observed in the locus coeruleus (A6; 87.7%), and in the A1(72.8%) cell group located in the caudal ventrolateral medulla (VLM). No co-localization of DBH/Fos neurons was observed in control animals. The highest degree of co-localization of PNMT/Fos neurons was observed in the C1 adrenergic cell group (84.2%) located in the rostral VLM. Control animals showed very few (5.5%) PNMT/Fos co-localized neurons in the C1 adrenergic cell group. Our results indicate that flurothyl-induced generalized seizures in rats activate catecholaminergic neurons in the pons and medulla oblongata. Further studies are necessary to determine whether activation of brainstem catecholaminergic neurons contribute to the autonomic manifestations that frequently accompany epileptic seizures.

The central amygdala modulates hypothalamic-pituitary-adrenal axis responses to systemic interleukin-1beta administration

In the present study we examined the role of the central nucleus of the amygdala in hypothalamic-pituitary-adrenal axis responses to an immune challenge in the form of systemic administration of the proinflammatory cytokine interleukin-1beta (1 microg/kg). We found that bilateral ibotenic acid lesions of the central amygdala substantially reduced adrenocorticotropin hormone release and hypothalamic corticotropin-releasing factor and oxytocin cell c-fos expression responses to interleukin-1,8 suggesting a facilitatory role for this structure in the generation of hypothalamic-pituitary-adrenal axis responses to an immune challenge. Since only a small number of central amygdala cells project directly to the paraventricular nucleus, we then examined the effect of central amygdala lesions on the activity of other brain nuclei that might act as relay sites in the control of the hypothalamic-pituitary-adrenal axis function. We found that bilateral central amygdala lesions significantly reduced interleukin-1beta-induced c-fos expression in cells of the ventromedial and ventrolateral subdivisions of the bed nucleus of the stria terminalis and brainstem catecholamine cell groups of the nucleus tractus solitarius (A2 noradrenergic cells) and ventrolateral medulla (A1 noradrenergic and C1 adrenergic cells). These findings, in conjunction with previous evidence of bed nucleus of the stria terminalis and catecholamine cell group involvement in hypothalamic-pituitary-adrenal axis regulation, suggest that ventromedial and ventrolateral bed nucleus of the stria terminalis cells and medullary catecholamine cells might mediate the influence of the central amygdala on hypothalamic-pituitary-adrenal axis responses to an immune challenge. Thus these data establish that the central amygdala influences hypothalamic-pituitary-adrenal axis responses to a systemic immune challenge but indicate that it primarily acts by modulating the activity of other control mechanisms.

Presence of mu-opioid receptors in targets of efferent projections from the central nucleus of the amygdala to the nucleus of the solitary tract

Opioids acting at mu-opioid receptors (MORs) within the nucleus of the solitary tract (NTS) potently modulate autonomic functions that are also known to be influenced by inputs from the central nucleus of the amygdala (CEA). In addition, many of the physiological effects of MOR agonists have been attributed to interactions with neurons that contain gamma-aminobutyric acid (GABA), one of the neurotransmitters present in CEA-derived terminals and their targets in the medial NTS. Together, these observations suggest that MORs are present at pre- or postsynaptic sites within the CEA to NTS circuitry. To test this hypothesis, we combined anterograde transport of biotinylated dextran amine (BDA) with immunogold-silver localization of an antipeptide antiserum against the MOR in the NTS of adult rats. In animals receiving bilateral CEA injections of BDA, anterogradely labeled axons were seen throughout the rostrocaudal NTS. Electron microscopy of the medial NTS at rostral and intermediate levels showed anterograde BDA-labeling in many small unmyelinated axons and axon terminals, none of which contained detectable MOR. The BDA-labeled axon terminals formed mainly symmetric, inhibitory-type synapses with somata and dendrites. Over half of the somatic and approximately 10% of the dendritic targets showed nonsynaptic plasmalemmal immunogold labeling for MOR. The BDA-labeled axon terminals were also frequently apposed by other small axons that contained MORs. These results suggest that within the medial NTS, MOR agonists modulate the postsynaptic inhibition produced by CEA afferents and also play a role in the presynaptic release of other neurotransmitters.

Chemical stimulation of the laryngopharynx increases Fos-like immunoreactivity in the rat hypothalamus and amygdala

Using immunohistochemical detection of the Fos protein as a cellular marker of neuronal activation, we examined forebrain areas that may be activated upon chemical stimulation of the laryngeal opening. Anesthetized rats were subject to multiple infusions of a chemical solution into the laryngopharynx. These animals were compared to two control groups: a surgical control group in which the animals were subject to the surgical procedure but received no stimulus infusions and a flow control group in which physiological saline replaced the chemical stimulus. Comparing the numbers of Fos-like-immunoreactive neurons in regions of the forebrain across groups revealed that infusing the chemical stimulus solution into the laryngopharyngeal opening selectively increased the number of Fos-like-immunoreactive nuclei in the paraventricular nucleus of the hypothalamus and the central nucleus of the amygdala, two autonomic-visceral related forebrain regions. Within the paraventricular nucleus of the hypothalamus, Fos-like-immunoreactive nuclei were significantly increased in the parvocellular subdivision while in the central nucleus of the amygdala, significant increases in Fos-like-immunoreactive nuclei were limited to the lateral capsular subdivision. These data suggest that in the rat laryngopharyngeal chemosensory stimulation activates forebrain regions that receive oral sensory information and are involved in visceral and autonomic functions.

Ultrastructure of glutamate and GABA immunoreactive axon terminals of the rat nucleus tractus solitarius, with a note on infralimbic cortex afferents

The principal fast neurotransmitters in the CNS are glutamate and GABA. Our aim was to provide a baseline account on the ultrastructure of the axon terminals immunoreactive to glutamate or GABA present in the nucleus tractus solitarius (NTS) of the rat. In addition, we wanted to complete our study of cortico-solitary afferents at the electron microscopic level, by analyzing the inputs from the infralimbic cortex. Using post-embedding immunogold, we found that nearly 61% of the axon terminals were glutamatergic, and 36% were GABAergic in the rat visceral NTS. In general, axons making asymmetric synaptic contacts were enriched in glutamate, compared to axons involved in symmetric synapses. In contrast, the vast majority of the GABAergic axon terminals made symmetric synaptic contacts. We could discern five types of glutamatergic and two types of GABAergic axon terminals that differed in their fine structure. Afferents from the infralimbic cortex were small, with clear synaptic vesicles and no dense core vesicles; they made asymmetric contacts with fine dendrites, and were glutamatergic. We conclude that most axon terminals in the NTS use glutamate or GABA as fast transmitters, in addition to being a heterogeneous population of morphological types.

Evidence for a GABAergic projection from the central nucleus of the amygdala to the brainstem of the macaque monkey: a combined retrograde tracing and in situ hybridization study

The central nucleus of the amygdala is interconnected with a variety of visceral and autonomic nuclei of the brainstem. These include the parabrachial nucleus, the nucleus of the solitary tract, the nucleus ambiguus and the dorsal motor nucleus of the vagus. Despite repeated attempts, neurochemical characterization of the major subcortical connections of the central nucleus has not yet been accomplished. Based on earlier immunohistochemical and in situ hybridization evidence indicating the presence of numerous GABAergic neurons in the macaque monkey central nucleus, we predicted that a sizeable portion of the descending projections may be GABAergic. We tested this hypothesis using a novel double labelling method with gold conjugated WGA-apoHRP as a retrograde tracer and in situ hybridization for detecting the mRNA that encodes the enzyme glutamic acid decarboxylase (GAD67) as a marker for GABAergic cells. Following WGA-apoHRP-gold injections into the brainstem, a large number of retrogradely labelled cells was observed in the medial and lateral divisions of the central nucleus. Of the retrogradely labelled cells observed in the medial division of the central nucleus, approximately half were double-labelled for GAD67 mRNA; about 30% double labelling was observed in the lateral division. These data support the view that a sizeable component of the central nucleus projection to the brainstem is GABAergic.

Interaction of the hypothalamic paraventricular nucleus and central nucleus of the amygdala in naloxone blockade of neuropeptide Y-induced feeding revealed by c-fos expression

Neuropeptide Y (NPY) is a powerful inducer of food intake with a key site of action in the paraventricular nucleus (PVN) of the hypothalamus. An effective method for inhibiting the effects of NPY is pretreatment with the opioid antagonists naloxone or naltrexone. In the present study, we used immunohistochemistry for cFos as a marker of neuronal activity to map the effects of PVN-injected NPY and blockade of these effects by peripheral injection of naloxone. Injection of NPY into the PVN resulted in an increase in food intake that was blocked by peripheral administration of naloxone. PVN NPY also resulted in increased cFos immunoreactivity (cFos-IR) in the PVN independent of food intake, and although peripheral naloxone inhibited NPY-induced feeding, it did not alter cFos-IR in the PVN. cFos-IR in the central nucleus of the amygdala (CNA) increased in response to both NPY and naloxone. Furthermore, the response to NPY and naloxone was additive, suggesting that peripheral naloxone and PVN NPY activate different neuronal populations in the CNA. Three other brain regions, the nucleus of the solitary tract, the ventrolateral medulla, and the supraoptic nucleus, all showed increases in cFos-IR in this study, but these changes came only as a result of increased food intake after PVN-injected NPY. The current data suggest that the CNA is a site important for the integration of the NPY and opioid systems.

Input from central nucleus of the amygdala efferents to pericoerulear dendrites, some of which contain tyrosine hydroxylase immunoreactivity

Light microscopic anterograde tracing studies indicate that neurons in the central nucleus of the amygdala (CNA) project to a region of the dorsal pontine tegmentum ventral to the superior cerebellar peduncle which contains noradrenergic dendrites of the nucleus locus coeruleus (LC). However, it has not been established whether the efferent terminals from the CNA target catecholamine-containing dendrites of the LC or dendrites of neurons from neighboring nuclei which may extend into this region. To examine this question, we combined immunoperoxidase labeling of the anterograde tracer biotinylated dextran amine (BDA) from the CNA with immunogold-silver labeling of the catecholamine-synthesizing enzyme tryrosine hydroxylase (TH) in the rostrolateral LC region of adult rats. By light microscopy, BDA-labeled processes were dense in the dorsal pons within the parabrachial nuclei as well as in the pericoerulear region immediately ventral to the superior cerebellar peduncle. Higher magnification revealed that BDA-labeled varicose fibers overlapped TH-labeled processes in this pericoerulear region. By electron microscopy, anterogradely labeled axon terminals contained small, clear as well as some large dense core vesicles and were commonly apposed to astrocytic processes along some portion of their plasmalemma. BDA-labeled terminals mainly formed symmetric type synaptic contacts characteristic of inhibitory transmitters. Of 250 BDA-labeled axon terminals examined where TH immunoreactivity was present in the neuropil, 81% contacted unlabeled and 19% contacted TH-labeled dendrites. Additionally, amygdala efferents were often apposed to unlabeled axon terminals forming asymmetric (excitatory type) synapses. These results demonstrate that amygdaloid efferents may directly alter the activity of catecholaminergic and non-catecholaminergic neurons in this pericoerulear region of the rat brain. Furthermore, our study suggests that CNA efferents may indirectly affect the activity of pericoerulear neurons through modulation of excitatory afferents. Amygdaloid projections to noradrenergic neurons may help integrate behavioral and visceral responses to threatening stimuli by influencing the widespread noradrenergic projections from the LC.

GABAergic neurons in rat nuclei of solitary tracts receive inhibitory-type synapses from amygdaloid efferents lacking detectable GABA-immunoreactivity

Gamma-aminobutyric acid (GABA) is a prominent inhibitory transmitter in both the central nucleus of the amygdala (Ce) and the medial nuclei of the solitary tracts (mNTS). These regions are reciprocally connected by anatomical pathways mediating the coordinated visceral responses to emotional stress. To further determine whether GABA is present in the amygdaloid efferents or their targets in the mNTS, we combined peroxidase labeling of Phaseolus vulgaris leucoagglutinin (PHA-L) or biotinylated dextran amine (BDA) anterogradely transported from the Ce with immunogold-silver detection of antibodies against GABA in the rat mNTS. By light microscopy, peroxidase labeling for either PHA-L or BDA was seen in varicose processes, whereas immunogold-silver labeling for GABA was detected in perikarya and processes throughout the rostrocaudal mNTS. The intermediate mNTS at the level of the area postrema, a region receiving mainly cardiorespiratory and gastric visceral afferents, were examined by electron microscopy. In this region, anterograde labeling was observed exclusively in unmyelinated axons and axon terminals. These terminals lacked detectable GABA-immunoreactivity, but formed symmetric synapses that are associated with inhibition. The targets of the anterogradely labeled terminals were medium-sized dendrites both with and without GABA-labeling. These dendrites often also received convergent input from terminals that were intensely GABA-immunoreactive. We conclude that visceral activation accompanying emotional response to stress is likely to involve inhibition of GABAergic neurons in the mNTS by non-GABA-containing amygdaloid efferents. Furthermore, our results indicate that the inhibition of these GABAergic neurons may be further augmented by release of GABA from other converging terminals in the mNTS.