Role of the neural cell adhesion molecule (NCAM) in amygdalo-hippocampal interactions and salience determination of contextual fear memory

Dopamine D1-like receptors modulate synchronized oscillations in the hippocampal-prefrontal-amygdala circuit in contextual fear

Contextual fear conditioning (CFC) is mediated by a neural circuit that includes the hippocampus, prefrontal cortex, and amygdala, but the neurophysiological mechanisms underlying the regulation of CFC by neuromodulators remain unclear. Dopamine D1-like receptors (D1Rs) in this circuit regulate CFC and local synaptic plasticity, which is facilitated by synchronized oscillations between these areas. In rats, we determined the effects of systemic D1R blockade on CFC and oscillatory synchrony between dorsal hippocampus (DH), prelimbic (PL) cortex, basolateral amygdala (BLA), and ventral hippocampus (VH), which sends hippocampal projections to PL and BLA. D1R blockade altered DH-VH and reduced VH-PL and VH-BLA synchrony during CFC, as inferred from theta and gamma coherence and theta-gamma coupling. D1R blockade also impaired CFC, as indicated by decreased freezing at retrieval, which was characterized by altered DH-VH and reduced VH-PL, VH-BLA, and PL-BLA synchrony. This reduction in VH-PL-BLA synchrony was not fully accounted for by non-specific locomotor effects, as revealed by comparing between epochs of movement and freezing in the controls. These results suggest that D1Rs regulate CFC by modulating synchronized oscillations within the hippocampus-prefrontal-amygdala circuit. They also add to growing evidence indicating that this circuit synchrony at retrieval reflects a neural signature of learned fear.

Aversive memory reactivation: A possible role for delta oscillations in the hippocampus-amygdala circuit

Memory labilization, the process by which memories become susceptible to update, is essential for memory reconsolidation and has been a target for novel therapies for traumatic memory-associated disorders. Maternal separation (MS) in male rats produced memories resistant to labilization in adulthood. Based on previous results, we hypothesized that temporal desynchronization between the dorsal hippocampus (DHc) and the basolateral amygdala (BLA), during memory retrieval, could be responsible for this impairment. Our goal was to investigate possible differences in oscillatory activity and synchrony between the DHc and BLA during fear memory reactivation, between MS and non-handled (NH) rats. We used male adult Wistar rats, NH or MS, with electrodes for local field potential (LFP) recordings implanted in the DHc and BLA. Animals were submitted to aversive memory reactivation by exposure to the conditioned context (Reat) or to pseudo-reactivation in a neutral context (pReat), and LFP was recorded. Plasticity markers linked to reconsolidation were evaluated one hour after reactivation. The power of delta oscillations and DHc-BLA synchrony in Reat animals was increased, during freezing. Besides, delta modulation of gamma oscillations amplitude in the BLA was associated with the increase in DHc Zif268 levels, an immediate early gene specifically associated with reconsolidation. Concerning early life stress, we found lower power of delta and strength of delta-gamma oscillations coupling in MS rats, compared to NH, which could explain the low Zif268 levels in a subgroup of MS animals. These results suggest a role for delta oscillations in memory reactivation that should be further investigated.

Hippocampal network oscillations at the interplay between innate anxiety and learned fear

The hippocampus plays a central role as a hub for episodic memory and as an integrator of multimodal sensory information in time and space. Thereby, it critically determines contextual setting and specificity of episodic memories. It is also a key site for the control of innate anxiety states and involved in psychiatric diseases with heightened anxiety and generalized fear memory such as post-traumatic stress disorder (PTSD). Expression of both innate "unlearned" anxiety and "learned" fear requires contextual processing and engagement of a brain-wide network including the hippocampus together with the amygdala and medial prefrontal cortex. Strikingly, the hippocampus is also the site of emergence of oscillatory rhythms that coordinate information processing and filtering in this network. Here, we review data on how the hippocampal network oscillations and their coordination with amygdalar and prefrontal oscillations are engaged in innate threat evaluation. We further explore how such innate oscillatory communication might have an impact on contextualization and specificity of "learned" fear. We illustrate the partial overlap of fear and anxiety networks that are built by the hippocampus in conjunction with amygdala and prefrontal cortex. We further propose that (mal)-adaptive interplay via (dis)-balanced oscillatory communication between the anxiety network and the fear network may determine the strength of fear memories and their resistance to extinction.

Hippocampal network oscillations as mediators of behavioural metaplasticity: Insights from emotional learning

Behavioural metaplasticity is evident in experience-dependent changes of network activity patterns in neuronal circuits that connect the hippocampus, amygdala and medial prefrontal cortex. These limbic regions are key structures of a brain-wide neural network that translates emotionally salient events into persistent and vivid memories. Communication in this network by-and-large depends on behavioural state-dependent rhythmic network activity patterns that are typically generated and/or relayed via the hippocampus. In fact, specific hippocampal network oscillations have been implicated to the acquisition, consolidation and retrieval, as well as the reconsolidation and extinction of emotional memories. The hippocampal circuits that contribute to these network activities, at the same time, are subject to both Hebbian and non-Hebbian forms of plasticity during memory formation. Further, it has become evident that adaptive changes in the hippocampus-dependent network activity patterns provide an important means of adjusting synaptic plasticity. We here summarise our current knowledge on how these processes in the hippocampus in interaction with amygdala and medial prefrontal cortex mediate the formation and persistence of emotional memories.

HIPP neurons in the dentate gyrus mediate the cholinergic modulation of background context memory salience

Cholinergic neuromodulation in the hippocampus controls the salience of background context memory acquired in the presence of elemental stimuli predicting an aversive reinforcement. With pharmacogenetic inhibition we here demonstrate that hilar perforant path-associated (HIPP) cells of the dentate gyrus mediate the devaluation of background context memory during Pavlovian fear conditioning. The salience adjustment is sensitive to reduction of hilar neuropeptide Y (NPY) expression via dominant negative CREB expression in HIPP cells and to acute blockage of NPY-Y1 receptors in the dentate gyrus during conditioning. We show that NPY transmission and HIPP cell activity contribute to inhibitory effects of acetylcholine in the dentate gyrus and that M1 muscarinic receptors mediate the cholinergic activation of HIPP cells as well as their control of background context salience. Our data provide evidence for a peptidergic local circuit in the dentate gyrus that mediates the cholinergic encoding of background context salience during fear memory acquisition.

Intra-hippocampal circuits are essential for associating a background context with behaviorally salient stimuli and involve cholinergic modulation at SST⁺ interneurons. Here the authors show that the salience of the background context memory is modulated through muscarinic activation of NPY⁺ hilar perforant path associated interneurons and NPY signaling in the dentate gyrus.

Identification of Parvalbumin Interneurons as Cellular Substrate of Fear Memory Persistence

Parvalbumin-positive (PV) basket cells provide perisomatic inhibition in the cortex and hippocampus and control generation of memory-related network activity patterns, such as sharp wave ripples (SPW-R). Deterioration of this class of fast-spiking interneurons has been observed in neuropsychiatric disorders and evidence from animal models suggests their involvement in the acquisition and extinction of fear memories. Here, we used mice with neuron type-targeted expression of the presynaptic gain-of-function glycine receptor RNA variant GlyR α3L^185L to genetically enhance the network activity of PV interneurons. These mice showed reduced extinction of contextual fear memory but normal auditory cued fear memory. They furthermore displayed increase of SPW-R activity in area CA3 and CA1 and facilitated propagation of this particular network activity pattern, as determined in ventral hippocampal slice preparations. Individual freezing levels during extinction and SPW-R propagation were correlated across genotypes. The same was true for parvalbumin immunoreactivity in the ventral hippocampus, which was generally augmented in the GlyR mutant mice and correlated with individual freezing levels. Together, these results identify PV interneurons as critical cellular substrate of fear memory persistence and associated SPW-R activity in the hippocampus. Our findings may be relevant for the identification and characterization of physiological correlates for posttraumatic stress and anxiety disorders.

Trait Anxiety Modulates Brain Activity during Performance of Verbal Fluency Tasks

Trait anxiety is thought to be associated with pathological anxiety, and a risk factor for psychiatric disorders. The present study examines the brain mechanisms associated with trait anxiety during the performing of verbal fluency tasks. The aim is to show how trait anxiety modulates executive functions as measured by verbal fluency, and to explore the link between verbal fluency and anxiety due to the putative negative biases in high-anxious individuals. Seven tasks of verbal fluency were used: letter "k," "f," verbs, "animals," "vehicles," "joy," and "fear." The results of 35 subjects (whole sample), and 17 subjects (nine men, eight women) selected from the whole sample for the low/high-anxious groups on the basis of Trait Anxiety scores were analyzed. The subjects were healthy, Polish speaking, right-handed and aged from 20 to 35 years old. fMRI (whole-brain analysis with FWE corrections) was used to show the neural signals under active participation in verbal fluency tasks. The results confirm that trait anxiety slightly modulates neural activation during the performance of verbal fluency tasks, especially in the more difficult tasks. Significant differences were found in brain activation during the performance of more complex tasks between individuals with low anxiety and those with high anxiety. Greater activation in the right hemisphere, frontal gyri, and cerebellum was found in people with low anxiety. The results reflect better integration of cognitive and affective capacities in individuals with low anxiety.

Promoter variant rs2301228 on the neural cell adhesion molecule 1 gene confers risk of schizophrenia in Han Chinese

BACKGROUND

Schizophrenia is recognized as a disorder of the brain and neuronal connectivity. The neural cell adhesion molecule 1 (NCAM1) gene plays a crucial role in regulating neuronal connectivity.

METHODS

We conducted a two-stage association analysis on 17 NCAM1 SNPs in two independent Han Chinese schizophrenia case-control cohorts (discovery sample from Hunan Province: 986 patients and 1040 normal controls; replication sample from Yunnan Province: 564 cases and 547 healthy controls). Allele, genotype and haplotype frequencies were compared between case and control samples. Transcription factor binding site prediction and luciferase reporter assays were employed to assess the potential function of promoter SNPs. We detected developmental changes at the transcriptional level of NCAM1 during neuron differentiation in Macaca mulatta neural progenitor cells (NPC). Serum levels of NCAM1 were measured in 72 cases and 88 controls.

RESULTS

A promoter variant, rs2301228, was found to be associated with schizophrenia at the allelic level and was validated in a replication cohort. Luciferase reporter assays demonstrated that risk allele rs2301228-A significantly down-regulated NCAM1 gene transcription compared to the G-allele. Concordantly, schizophrenia patients had a significantly lower level of serum NCAM1 compared to healthy donors. During the NPC neuronal differentiation, NCAM1 mRNA was significantly increased, suggesting a critical role of this gene in neural development.

CONCLUSIONS

Our results provide direct evidence for NCAM1 as a susceptibility gene for schizophrenia, which offers support to a neurodevelopmental model and neuronal connectivity hypothesis in the onset of schizophrenia.

Networks of protein kinases and phosphatases in the individual phases of contextual fear conditioning in the C57BL/6J mouse

Although protein kinases and phosphatases have been reported to be involved in fear memory, information about these signalling molecules in the individual phases of contextual fear conditioning (cFC) is limited. C57BL/6J mice were tested in cFC, sacrificed and hippocampi were used for screening of approximately 800 protein kinases and phosphatases by protein microarrays with subsequent Western blot confirmation of threefold higher or lower hippocampal levels as compared to foot shock controls. Immunoblotting of the protein kinases and phosphatases screened out was carried out by Western blotting. A network of protein kinases and phosphatases was generated (STRING 9.1). Animals learned the task in the paradigm and protein kinase and phosphatase levels were determined in the individual phases acquisition, consolidation and retrieval and compared to foot shock controls. Protein kinases discoidin containing receptor 2 (DDR2), mitogen activated protein kinase kinase kinase 7 (TAK1), protein phosphatases dual specificity protein phosphatase (PTEN) and protein phosphatase 2a (PP2A) were modulated in the individual phases of cFC. Phosphatidyl-inositol-3,4,5-triphosphate 3-phosphatase and phosphatidylinositol-3 kinase (PI3K) that is interacting with PTEN were modulated as well. Freezing time was correlating with PP2A levels in the retrieval phase of cFC. The abovementioned protein kinases, phosphatases and inositol-signalling enzymes were not reported so far in cFC and the results are relevant for interpretation of previous and design of future studies in cFC or fear memory. Protein phosphatase PP2A was, however, the only signalling compound tested that was directly linked to retrieval in the cFC.

Contextual fear conditioning modulates hippocampal AMPA-, GluN1- and serotonin receptor 5-HT1A-containing receptor complexes

Although the roles of AMPAR (α-amino-3-hydroxyl-5-methyl-4-isoxazole propionate receptor), NMDAR (N-methyl-D-aspartate receptor) and 5HT1AR (5-hydroxytryptamine sub type 1A) in contextual fear conditioning (cFC) have been studied, information about receptor-containing complexes (RC) is not available. Moreover, there are no data on membrane or endosomal NMDA-, 5HT1A- or AMPA-RC levels, which would likely be indicative of the trafficking of these receptors. cFC was carried out in C57BL/6j mice and animals were sacrificed in the individual phases and hippocampi were taken for the determination of receptor complex and subunit levels using BN- and SDS-PAGE with subsequent Western blotting. GluA1-4, GluN1 (NMDAR subunit NR1)- and 5HT1A-RC were differentially regulated during the individual phases and differentially regulated in the membrane and endosomal fractions. GluA1-RC levels in the membrane were increased in acquisition, consolidation and retrieval phases; GluA2-RC and GluA3-RC membrane levels were reduced and modulated in early endosomes during these phases. GluA4-RC and GluN1-RC levels as well as their subunits showed the same pattern in the membrane during consolidation while 5HT1A-RC membrane and endosome levels were mainly increased during consolidation and retrieval. Taken together, the results suggest that levels of 5-HT1A-RC, NMDA-RC and AMPA-RC and subunits in membrane and endosomal preparations are paralleling individual phases of cFC. The findings from the current study suggest phase-specific receptor complex and subunit formation and propose that receptor complexes should be examined in parallel with receptor subunits to aid the interpretation of previous work and to design future work on neurotransmitter receptors in memory paradigms.

Glutamic acid decarboxylase 65: a link between GABAergic synaptic plasticity in the lateral amygdala and conditioned fear generalization

An imbalance of the gamma-aminobutyric acid (GABA) system is considered a major neurobiological pathomechanism of anxiety, and the amygdala is a key brain region involved. Reduced GABA levels have been found in anxiety patients, and genetic variations of glutamic acid decarboxylase (GAD), the rate-limiting enzyme of GABA synthesis, have been associated with anxiety phenotypes in both humans and mice. These findings prompted us to hypothesize that a deficiency of GAD65, the GAD isoform controlling the availability of GABA as a transmitter, affects synaptic transmission and plasticity in the lateral amygdala (LA), and thereby interferes with fear responsiveness. Results indicate that genetically determined GAD65 deficiency in mice is associated with (1) increased synaptic length and release at GABAergic connections, (2) impaired efficacy of GABAergic synaptic transmission and plasticity, and (3) reduced spillover of GABA to presynaptic GABAB receptors, resulting in a loss of the associative nature of long-term synaptic plasticity at cortical inputs to LA principal neurons. (4) In addition, training with high shock intensities in wild-type mice mimicked the phenotype of GAD65 deficiency at both the behavioral and synaptic level, indicated by generalization of conditioned fear and a loss of the associative nature of synaptic plasticity in the LA. In conclusion, GAD65 is required for efficient GABAergic synaptic transmission and plasticity, and for maintaining extracellular GABA at a level needed for associative plasticity at cortical inputs in the LA, which, if disturbed, results in an impairment of the cue specificity of conditioned fear responses typifying anxiety disorders.

NCAM deficiency in the mouse forebrain impairs innate and learned avoidance behaviours

The neural cell adhesion molecule (NCAM) has been implicated in the development and plasticity of neural circuits and the control of hippocampus- and amygdala-dependent learning and behaviour. Previous studies in constitutive NCAM null mutants identified emotional behaviour deficits related to disturbances of hippocampal and amygdala functions. Here, we studied these behaviours in mice conditionally deficient in NCAM in the postmigratory forebrain neurons. We report deficits in both innate and learned avoidance behaviours, as observed in elevated plus maze and passive avoidance tasks. In contrast, general locomotor activity, trait anxiety or neophobia were unaffected by the mutation. Altered avoidance behaviour of the conditional NCAM mutants was associated with a deficit in serotonergic signalling, as indicated by their reduced responsiveness to (±)-8-hydroxy-2-(dipropylamino)-tetralin-induced hypothermia. Another serotonin-dependent behaviour, namely intermale aggression that is massively increased in constitutively NCAM-deficient mice, was not affected in the forebrain-specific mutants. Our data suggest that genetically or environmentally induced changes of NCAM expression in the late postnatal and mature forebrain determine avoidance behaviour and serotonin (5-HT)1A receptor signalling.

Circadian modulation of anxiety: a role for somatostatin in the amygdala

Pharmacological evidence suggests that the neuropeptide somatostatin (SST) exerts anxiolytic action via the amygdala, but findings concerning the putative role of endogenous SST in the regulation of emotional responses are contradictory. We hypothesized that an endogenous regulation of SST expression over the course of the day may determine its function and tested both SST gene expression and the behavior of SST knock out (SST^-/-) mice in different aversive tests in relation to circadian rhythm. In an open field and a light/dark avoidance test, SST^-/- mice showed significant hyperactivity and anxiety-like behavior during the second, but not during the first half of the active phase, failing to show the circadian modulation of behavior that was evident in their wild type littermates. Behavioral differences occurred independently of changes of intrinsically motivated activity in the home cage. A circadian regulation of SST mRNA and protein expression that was evident in the basolateral complex of the amygdala of wild type mice may provide a neuronal substrate for the observed behavior. However, fear memory towards auditory cue or the conditioning context displayed neither a time- nor genotype-dependent modulation. Together this indicates that SST, in a circadian manner and putatively via its regulation of expression in the amygdala, modulates behavior responding to mildly aversive conditions in mice.

Role of stress system disturbance and enhanced novelty response in spatial learning of NCAM-deficient mice

The neural cell adhesion molecule (NCAM) plays a crucial role in stress-related brain function, emotional behavior and memory formation. In this study, we investigated the functions of the glucocorticoid and serotonergic systems in mice constitutively deficient for NCAM (NCAM-/- mice). Our data provide evidence for a hyperfunction of the hypothalamic-pituitary-adrenal axis, with enlarged adrenal glands and increased stress-induced corticosterone release, but reduced hippocampal glucocorticoid receptor expression in NCAM-/- mice when compared to NCAM+/+ mice. We also obtained evidence for a hypofunction of 5-HT1A autoreceptors as indicated by increased 8-0H-DPAT-induced hypothermia. These findings suggest a disturbance of both humoral and neural stress systems in NCAM-/- mice. Accordingly, we not only confirmed previously observed hyperarousal of NCAM-/- mice in various anxiety tests, but also observed an increased response to novelty exposure in these animals. Spatial learning deficits of the NCAM-/- mice in a Morris Water maze persisted, even when mice were pretrained to prevent effects of novelty or stress. We suggest that NCAM-mediated processes are involved in both novelty/stress-related emotional behavior and in cognitive function during spatial learning.

A common genetic mechanism underlying susceptibility to posttraumatic stress disorder

We hypothesize that susceptibility to post-traumatic stress disorder (PTSD) may be determined in part by aberrant microtubule-associated protein tau expression in neurons of critical brain structures. The following lines of evidence support this hypothesis. First, epidemiologic data suggest the involvement of genetic factors in the susceptibility to PTSD. Second, the common features of both abnormal tau expression and PTSD include amygdalar and hippocampal atrophy, upregulation of norepinephrine biosynthetic capacity in the surviving locus coeruleus neurons and dysfunction of N-methyl-D-aspartate-receptors. Finally, our experiments using rTg4510 mice, a model that over-expresses human mutant tau and develops age-dependent tauopathy, demonstrate that these animals display circling behavior thought to be related to states of anxiety. To detect the potential molecular mechanisms underlying PTSD episodes, laser-assisted/capture microdissection can be used with microarray analysis as an alternative approach to identify changes in gene expression in excitatory and/or inhibitory neurons in critical brain structures (i.e., hippocampus and amygdala) in response to the onset of PTSD.

Pain and analgesia: the value of salience circuits

Evaluating external and internal stimuli is critical to survival. Potentially tissue-damaging conditions generate sensory experiences that the organism must respond to in an appropriate, adaptive manner (e.g., withdrawal from the noxious stimulus, if possible, or seeking relief from pain and discomfort). The importance we assign to a signal generated by a noxious state, its salience, reflects our belief as to how likely the underlying situation is to impact our chance of survival. Importantly, it has been hypothesized that aberrant functioning of the brain circuits which assign salience values to stimuli may contribute to chronic pain. We describe examples of this phenomenon, including 'feeling pain' in the absence of a painful stimulus, reporting minimal pain in the setting of major trauma, having an 'analgesic' response in the absence of an active treatment, or reporting no pain relief after administration of a potent analgesic medication, which may provide critical insights into the role that salience circuits play in contributing to numerous conditions characterized by persistent pain. Collectively, a refined understanding of abnormal activity or connectivity of elements within the salience network may allow us to more effectively target interventions to relevant components of this network in patients with chronic pain.

A key role for nectin-1 in the ventral hippocampus in contextual fear memory

Nectins are cell adhesion molecules that are widely expressed in the brain. Nectin expression shows a dynamic spatiotemporal regulation, playing a role in neural migratory processes during development. Nectin-1 and nectin-3 and their heterophilic trans-interactions are important for the proper formation of synapses. In the hippocampus, nectin-1 and nectin-3 localize at puncta adherentia junctions and may play a role in synaptic plasticity, a mechanism essential for memory and learning. We evaluated the potential involvement of nectin-1 and nectin-3 in memory consolidation using an emotional learning paradigm. Rats trained for contextual fear conditioning showed transient nectin-1—but not nectin-3—protein upregulation in synapse-enriched hippocampal fractions at about 2 h posttraining. The upregulation of nectin-1 was found exclusively in the ventral hippocampus and was apparent in the synaptoneurosomal fraction. This upregulation was induced by contextual fear conditioning but not by exposure to context or shock alone. When an antibody against nectin-1, R165, was infused in the ventral-hippocampus immediately after training, contextual fear memory was impaired. However, treatment with the antibody in the dorsal hippocampus had no effect in contextual fear memory formation. Similarly, treatment with the antibody in the ventral hippocampus did not interfere with acoustic memory formation. Further control experiments indicated that the effects of ventral hippocampal infusion of the nectin-1 antibody in contextual fear memory cannot be ascribed to memory non-specific effects such as changes in anxiety-like behavior or locomotor behavior. Therefore, we conclude that nectin-1 recruitment to the perisynaptic environment in the ventral hippocampus plays an important role in the formation of contextual fear memories. Our results suggest that these mechanisms could be involved in the connection of emotional and contextual information processed in the amygdala and dorsal hippocampus, respectively, thus opening new venues for the development of treatments to psychopathological alterations linked to impaired contextualization of emotions.

Are NCAM deficient mice an animal model for schizophrenia?

Genetic and biomarker studies in patients have identified the Neural Cell Adhesion Molecule (NCAM) and its associated polysialic acid (PSA) as a susceptibility factors for schizophrenia. NCAM and polysialtransferase mutant mice have been generated that may serve as animal models for this disorder and allow to investigate underlying neurodevelopmental alterations. Indeed, various schizophrenia-relevant morphological, cognitive and emotional deficits have been observed in these mutants. Here we studied social interaction and attention of NCAM null mutant (NCAM^−/−) mice as further hallmarks of schizophrenia. Nest building, which is generally associated with social behavior in rodents, was severely impaired, as NCAM^−/− mice continuously collected smaller amounts of nest building material than their wild type littermates and built nests of poorer quality. However, social approach tested in a three—compartment—box was not affected and latent inhibition of Pavlovian fear memory was not disturbed in NCAM^−/− mice. Although NCAM deficient mice do not display a typical schizophrenia-like phenotype, they may be useful for studying specific endophenotypes with relevance to the disease.

Synaptology of ventral CA1 and subiculum projections to the basomedial nucleus of the amygdala in the mouse: relation to GABAergic interneurons

GABAergic neurons of the amygdala are thought to play a critical role in establishing networks for feedback and feedforward inhibition and in mediating rhythmic network activity patterns relevant for emotional behavior, determination of stimulus salience, and memory strength under stressful experiences. These functions are typically fulfilled in interplay of amygdala and hippocampus. Therefore, we explored the putative connectivity of GABAergic neurons with the hippocampo-amygdalar projection with the anterograde tracers Phaseolus vulgaris leucoagglutinin (Phal) and Miniruby injected to GAD67-GFP knock-in mice in which GABAergic neurons are labeled by the expression of the gene for green fluorescent protein (GFP) inserted to the GAD1 gene locus (Tamamaki et al. J Comp Neurol 467:60-79, 2003). We found that, while hippocampal axons target all nuclei of the amygdala, the densest fiber plexus was found in the posterior basomedial nucleus. Electron microscopy revealed that the vast majority of contacts in this nucleus were formed by thin fibers making small asymmetrical contacts, predominantly on GFP-negative profiles. However, several asymmetrical contacts could also be seen on GFP-positive profiles. A surprising result was the occasional occurrence of anterogradely labeled symmetrical synapses indicating a GABAergic contribution to the projection from the hippocampus to the amygdala. While hippocampal input to the amygdala appears to be largely excitatory and targets non-GABAergic neurons, our data provide evidence for a direct involvement of GABAergic neurons in the interplay of these regions, either as target in the amygdala or as projection neurons from the hippocampus. These particular "interface neurons" may be of relevance for the information processing in the amygdalo-hippocampal system involved in emotional behavior and memory formation.