The generation of emotional responses by the basolateral amygdala is largely determined by the balance of excitatory and inhibitory inputs to its principal neurons, the pyramidal cells. The activity of these neurons is tightly controlled by GABAergic interneurons, especially a parvalbumin-positive (PV+) subpopulation that constitutes almost half of all interneurons in the basolateral amygdala. In the present semi-quantitative investigation we studied the incidence of synaptic inputs of PV+ axon terminals onto pyramidal neurons in the rat basolateral nucleus (BLa). Pyramidal cells were identified using calcium/calmodulin-dependent protein kinase II (CaMK) immunoreactivity as a marker. In order to appreciate the relative abundance of PV+ inputs compared to excitatory inputs and other non-PV+ inhibitory inputs, we also analyzed the proportions of asymmetrical (presumed excitatory) synapses and symmetrical (presumed inhibitory) synapses formed by unlabeled axon terminals targeting pyramidal neurons. The results indicate that the perisomatic region of pyramidal cells is innervated almost entirely by symmetrical synapses, whereas the density of asymmetrical synapses increases as one proceeds from thicker proximal dendritic shafts to thinner distal dendritic shafts. The great majority of synapses with dendritic spines are asymmetrical. PV+ axon terminals mainly form symmetrical synapses. These PV+ synapses constitute slightly more than half of the symmetrical synapses formed with each postsynaptic compartment of BLa pyramidal cells. These data indicate that the synaptology of basolateral amygdalar pyramidal cells is remarkably similar to that of cortical pyramidal cells, and that PV+ interneurons provide a robust inhibition of both the perisomatic and distal dendritic domains of these principal neurons. Keywordsimmunocytochemistry; electron microscopy; inhibition, calcium/calmodulin-protein kinase II The basolateral amygdala (ABL), which consists of the lateral, basolateral, and basomedial amygdalar nuclei, is one of the most important brain regions for the generation of emotional behavior and the formation of emotional memories (Aggleton, 1992;Aggleton, 2000;Shinnick-Gallagher et al., 2003). It receives sensory information from the thalamus and cerebral cortex (McDonald, 1998) and produces appropriate emotional responses by activating a variety of subcortical regions including the central amygdalar nucleus, bed nucleus of the stria terminalis, and striatum. The outputs of the ABL arise from pyramidal cells (McDonald, 1992b), which resemble their counterparts in the cerebral cortex. These neurons, which constitute about 85% of the neurons in the ABL, are characterized by a pyramidal or piriform cell body, and spiny dendrites (Hall, 1972;McDonald, 1982McDonald, , 1984 1992a,b;Millhouse and DeOlmos, 1983). Some ABL pyramidal cells have a marked pyramidal morphology, with a clear differentiation of thicker "apical" dendrites from thinner "basal" dendrites, whereas others have a semipyramidal or even stellate appe...

Summary Activation of corticotrophin releasing factor (CRF) neurons in the paraventricular nucleus of the hypothalamus (PVN) is necessary for establishing the classic endocrine response to stress, while activation of forebrain CRF neurons mediates affective components of the stress response. Previous studies have reported that mRNA for CRF2 receptor (CRFR2) is expressed in the bed nucleus of the stria terminalis (BNST) as well as hypothalamic nuclei, but little is known about the localization and cellular distribution of CRFR2 in these regions. Using immunofluorescence with confocal microscopy, as well as electron microscopy, we demonstrate that in the BNST CRFR2-immunoreactive fibers represent moderate to strong labeling on axons terminals. Dual-immunofluorescence demonstrated that CRFR2-fibers co-localize oxytocin (OT), but not arginine-vasopressin (AVP), and make perisomatic contacts with CRF neurons. Dual-immunofluorescence and single cell RT-PCR demonstrate that in the hypothalamus, CRFR2 immunoreactivity and mRNA are found in OT, but not in CRF or AVP-neurons. Furthermore, CRF neurons of the PVN and BNST express mRNA for the oxytocin receptor, while the majority of OT/CRFR2 neurons in the hypothalamus do not. Finally, using adenoviral-based anterograde tracing of PVN neurons, we show that OT/CRFR2-immunoreactive fibers observed in the BNST originate in the PVN. Our results strongly suggest that CRFR2 located on oxytocinergic neurons and axon terminals might regulate the release of this neuropeptide and hence might be a crucial part of potential feedback loop between the hypothalamic oxytocin system and the forebrain CRF system that could significantly impact affective and social behaviors, in particular during times of stress.

The basolateral amygdala contains several subpopulations of inhibitory interneurons that can be distinguished on the basis of their content of calcium-binding proteins or peptides. Although previous studies have shown that interneuronal subpopulations containing parvalbumin (PV) or vasoactive intestinal peptide (VIP) innervate distinct postsynaptic domains of pyramidal cells as well as other interneurons, very little is known about the synaptic outputs of the interneuronal subpopulation that expresses somatostatin (SOM). The present study utilized dual-labeling immunocytochemical techniques at the light and electron microscopic levels to analyze the innervation of pyramidal cells, PV+ interneurons, and VIP+ interneurons in the anterior basolateral amygdalar nucleus (BLa) by SOM+ axon terminals. Pyramidal cell somata and dendrites were selectively labeled with antibodies to calcium/calmodulin-dependent protein kinase II (CaMK); previous studies have shown that the vast majority of dendritic spines, whether CAMK+ or not, arise from pyramidal cells. Almost all SOM+ axon terminals formed symmetrical synapses. The main postsynaptic targets of SOM+ terminals were small-caliber CaMK+ dendrites and dendritic spines, some of which were CaMK+. These SOM+ synapses with dendrites were often in close proximity to asymmetrical (excitatory) synapses to these same structures formed by unlabeled terminals. Few SOM+ terminals formed synapses with CaMK+ pyramidal cell somata or large-caliber (proximal) dendrites. Likewise, only 15% of SOM+ terminals formed synapses with PV+, VIP+, or SOM+ interneurons. These findings suggest that inhibitory inputs from SOM+ interneurons may interact with excitatory inputs to pyramidal cell distal dendrites in the BLa. These interactions might affect synaptic plasticity related to emotional learning.

Although calcium/calmodulin-dependent protein kinase II (CaMK) has been shown to play a critical role in long-term potentiation (LTP) and emotional learning mediated by the basolateral amygdala, little is known about its cellular localization in this region. We have utilized immunohistochemical methods to study the neuronal localization of CaMK, and its relationship to gamma-aminobutyric acid (GABA)-ergic structures, in the rat basolateral amygdala (ABL). Light microscopic observations revealed dense CaMK staining in the ABL. Although the cell bodies and proximal dendrites of virtually every pyramidal cell appeared to be CaMK(+), the cell bodies of small nonpyramidal neurons were always unstained. Dual localization of CaMK and GABA immunoreactivity with confocal immunofluorescence microscopy revealed that CaMK and GABA were found in different neuronal populations in the ABL. CaMK was contained only in pyramidal neurons; GABA was contained only in nonpyramidal cells. At the ultrastructural level, it was found that CaMK was localized to pyramidal cell bodies, thick proximal dendrites, thin distal dendrites, most dendritic spines, axon initial segments, and axon terminals forming asymmetrical synapses. These findings suggest that all portions of labeled pyramidal cells, with the exception of some dendritic spines, can exhibit CaMK immunoreactivity. By using a dual immunoperoxidase/immunogold-silver procedure at the ultrastructural level, GABA(+) axon terminals were seen to innervate all CaMK(+) postsynaptic domains, including cell bodies (22%), thick (>1 microm) dendrites (34%), thin (<1 microm) dendrites (22%), dendritic spines (17%), and axon initial segments (5%). These findings indicate that CaMK is a useful marker for pyramidal neurons in ultrastructural studies of ABL synaptology and that the activity of pyramidal neurons in the ABL is tightly controlled by a high density of GABAergic terminals that target all postsynaptic domains of pyramidal neurons.

Although it is well established that the activity of pyramidal projection neurons in the basolateral amygdala (ABL) is controlled by gamma-aminobutyric acid (GABA)ergic inhibitory interneurons, very little is known about the connections of specific interneuronal subpopulations in this region. In the present study, immunohistochemical techniques were used at the light and electron microscopic levels to identify specific populations of interneurons and to analyze their connections with each other and with unlabeled presumptive pyramidal neurons. Double-labeling immunofluorescence experiments revealed that antibodies to vasoactive intestinal peptide (VIP) and calbindin-D28K (CB) labeled two separate interneuronal subpopulations in the ABL. Light microscopic double-labeling immunoperoxidase experiments demonstrated that many VIP-positive (VIP+) axon terminals formed intimate synaptic-like contacts with the CB-positive (CB+) neurons and that both CB+ and VIP+ terminals often contributed to the formation of pericellular baskets that surrounded unlabeled perikarya of pyramidal neurons. By using a dual immunoperoxidase/immunogold-silver procedure at the ultrastructural level, it was found that 30% of VIP+ terminals in the anterior subdivision of the basolateral nucleus innervated interneurons that were either CB+ (25%) or VIP+ (5%). A smaller percentage (15%) of CB+ terminals formed synapses with labeled interneurons. Both VIP+ and CB+ terminals also innervated unlabeled perikarya, dendrites, and spines, most of which probably belonged to pyramidal neurons. The interconnections between interneurons may be important for disinhibitory mechanisms and the mediation of rhythmic oscillations in the ABL.

Recent studies indicate that the basolateral amygdala exhibits fast rhythmic oscillations during emotional arousal, but the neuronal mechanisms underlying this activity are not known. Similar oscillations in the cerebral cortex are generated by a network of parvalbumin (PV)-immunoreactive interneurons interconnected by chemical synapses and dendritic gap junctions. The present immunoelectron microscopic study revealed that the basolateral amygdalar nucleus (BLa) contains a network of parvalbumin-immunoreactive (PVϩ) interneurons interconnected by chemical synapses, dendritic gap junctions, and axonal gap junctions. Twenty percent of synapses onto PVϩ neurons were formed by PVϩ axon terminals. All of these PVϩ synapses were symmetrical. PVϩ perikarya exhibited the greatest incidence of PVϩ synapses (30%), with lower percentages associated with PVϩ dendrites (15%) and spines (25%). These synapses comprised half of all symmetrical synapses formed with PVϩ cells. A total of 18 dendrodendritic gap junctions between PVϩ neurons were observed, mostly involving secondary and more distal dendrites (0.5-1.0 m thick). Dendritic gap junctions were often in close proximity to PVϩ chemical synapses. Six gap junctions were observed between PVϩ axon terminals. In most cases, one or both of these terminals formed synapses with the perikarya of principal neurons. This is the first study to describe dendritic gap junctions interconnecting PVϩ interneurons in the basolateral amygdala. It also provides the first documentation of gap junctions between interneuronal axon terminals in the mammalian forebrain. These data provide the anatomical basis for a PVϩ network that may play a role in the generation of rhythmic oscillations in the BLa during emotional arousal.

The basolateral nuclear complex of the amygdala (BLC) receives a dense serotonergic innervation that appears to play a critical role in the regulation of mood and anxiety. However, little is known about how serotonergic inputs interface with different neuronal subpopulations in this region. To address this question, dual-labeling immunohistochemical techniques were used at the light and electron microscopic levels to examine inputs from serotonin-immunoreactive (5-HT+) terminals to different neuronal subpopulations in the rat BLC. Pyramidal cells were labeled by using antibodies to calcium/calmodulin-dependent protein kinase II, whereas different interneuronal subpopulations were labeled by using antibodies to a variety of interneuronal markers including parvalbumin (PV), vasoactive intestinal peptide (VIP), calretinin, calbindin, cholecystokinin, and somatostatin. The BLC exhibited a dense innervation by thin 5-HT+ axons. Electron microscopic examination of the anterior basolateral nucleus (BLa) revealed that 5-HT+ axon terminals contained clusters of small synaptic vesicles and a smaller number of larger dense-core vesicles. Serial section reconstruction of 5-HT+ terminals demonstrated that 76% of these terminals formed synaptic junctions. The great majority of these synapses were symmetrical. The main targets of 5-HT+ terminals were spines and distal dendrites of pyramidal cells. However, in light microscopic preparations it was common to observe apparent contacts between 5-HT+ terminals and all subpopulations of BLC interneurons. Electron microscopic analysis of the BLa in sections dual-labeled for 5-HT/PV and 5-HT/VIP revealed that many of these contacts were synapses. These findings suggest that serotonergic axon terminals differentially innervate several neuronal subpopulations in the BLC.

Three morphologically distinct classes of receptor-neurons are proposed: (1) type I ciliar cells, (2) microvillar cells and (3) type II ciliar cells. Retrograde transport of horseradish peroxidase by axons in the olfactory nerve to the olfactory organs of goldfish (Carassius auratus) and channel catfish (Ictalurus punctatus) provided evidence that these axon-bearing cells are present in the organs of both species. Goldfish olfactory organs were also studied with scanning electron microscopy, dissociated with papain for isolated cell preparations, and processed for ultrastructural localization of acid phosphatase activity. Type I ciliar cells are similar to ciliar olfactory receptors found in all vertebrate classes. Microvillar cells are present in the olfactory organs of most fishes and in the tetrapod vomeronasal organ. In goldfish and catfish, type I ciliar and microvillar cells are concentrated on the inner third of each lamella, nearest to the median raphe. Type II ciliar cells have often been described as respiratory-type or ciliated nonsensory cells. They are structurally similar to respiratory epithelial cells in the nasal cavities of tetrapods and have motile cilia that beat synchronously, indicative of their role in mediating fluid flow over the olfactory epithelium. In goldfish they occur singly and in aggregates throughout the organ. In catfish they are segregated from type I ciliar and microvillar cells on the outer two-thirds of each lamella. In goldfish and catfish they have axons that pass through the olfactory nerve to the olfactory bulb; hence, they are receptor-neurons as well as analogous to respiratory epithelium. In addition to the three receptor types described above, cells resembling receptors with rodlike distal processes were seen filled with horseradish peroxidase and observed with scanning and transmission electron microscopy. Cells of similar structure have been documented elsewhere, often called "rod cells," and sometimes considered a separate receptor type in fishes. In this study, a number of rodlike processes were found with their ciliar or microvillar components partially fused. High levels of acid phosphatase activity were localized to these processes, and examples were found that corresponded to each of the three receptor types. Olfactory receptor turnover is believed to persist through life. The evidence presented supports the hypothesis that fusion of their dendritic apical processes marks an early stage of receptor cell senescence.