Alterations of AMPA-selected glutamate subtype immunoreactivity in the dentate gyrus after perforant pathway lesion

Immunocytochemical techniques were employed to examine the changes in immunolabeling of the α-amino-3-hydroxy-5-methyl-4-isoaxolepropionate (AMPA) receptor subunits GluR1 and GluR2/3 within the dentate gyrus 1, 3, 7, 14, 30, and 90 days after a unilateral perforant pathway lesion in the rat brain. Completeness of the lesion was confirmed following examination of Nissl-stained tissue sections at all times post-lesion and acetylcholinesterase (AChE)-stained sections 14, 30 and 90 days post-lesion, the latter providing evidence of compensatory sprouting of cholinergic fibers in the outer molecular layer of the dentate gyrus. Compared to the non-lesioned hippocampus there was no difference in the staining pattern of AMPA receptor subunits in the dentate gyrus of the deafferented hippocampus 1, 3, 7 and 14 days following lesioning of the perforant pathway. In contrast, 30 and 90 days post-lesion, GluR1 immunolabeling was increased in the outer molecular layer of the dentate gyrus (i.e., deafferented zone) ipsilateral to lesion. Likewise, GluR2/3 immunolabeling was increased within the same region although the intensity of the response was less than that which was observed for GluR1. These data suggest that the loss of the perforant pathway fibers results in a compensatory increase in GluR1 and to a lesser extent GluR2/3 immunolabeling of the outer molecular layer at 30 and 90 days post-lesion and further suggest that AMPA receptor subunits play a role in perforant pathway signal transduction.     

Distribution of glutamate receptor subunit NMDAR1 in the hippocampus of normal elderly and patients with Alzheimer's disease

Immunocytochemical techniques were employed to study the distribution and cytological features of NMDAR1-immunoreactive elements in the human hippocampal formation. Subjects with Alzheimer's disease (AD), presenting with a wide range of neuropathology and classified into six Braak stage (I-VI), and nondemented age-matched controls were examined. In control cases, the most intense NMDAR1 immunoreactivity was observed within the soma and dendrites of granule cells in the dentate gyrus and pyramidal neurons in Ammon's horn. Whereas small variations in the pattern of immunoreactivity were noted in control cases, AD subjects were characterized with intersubject variability which in most instances correlated with neuropathologic severity. For example, AD cases, particularly those with mild/modest pathology (Braak I-III), were indistinguishable from controls in the overall pattern of immunolabeling. In contrast, in those more severe AD cases (Braak IV-VI) the intensity of immunolabeling within the CA fields was greater than observed in controls and those with mild AD pathology. In addition, in pathologically severe cases numerous NMDAR1-positive pyramidal neurons were characterized by unique morphologic features including long and often tortuous apical dendrites. These latter findings were most prevalent in the CA1 region and subiculum. In contrast to the marked increase in immunolabeling in the CA fields, in the dentate gyrus we observed a reduction in NMDAR1 labeling particularly within the outer molecular layer (i.e., termination zone of the perforant pathway). This latter region was also the site of a number of NMDAR1-labeled plaques. Notably, the overall pattern of NMDAR1 immunoreactivity is distinct from that observed with antibodies against AMPA receptor subunits and suggests a differential role of various inotropic glutamate receptors in hippocampal plasticity in AD.     

Attenuated lesion-induced N-methyl-D-aspartate receptor (NMDAR) plasticity in the dentate gyrus of aged rats following perforant path lesions.

Young animals demonstrate a significant upregulation of N-methyl-d-aspartate receptor 1 (NMDAR1) in the outer molecular layer (OML) of the dentate gyrus following a total unilateral ablation of the perforant path, and this response presumably facilitates a degree of functional recovery. Aged animals have attenuated responses to lesion-induced synaptic plasticity as compared with young subjects, and in fact display decreased synaptogenesis and sprouting following a unilateral perforant path lesion. To investigate the response of NMDAR1 in the dentate gyrus of aged animals to perforant path ablation, 24-month-old Sprague-Dawley male rats received a unilateral knife cut of the angular bundle. Our results demonstrated that aged animals displayed a blunted response to lesion-induced NMDA receptor-mediated plasticity, suggesting that aged animals have an impaired ability to respond to deafferentation through an increase in NMDA receptor levels in the deafferented zone.     

AMPA-Selective glutamate receptor subtype immunoreactivity in the hippocampal dentate gyrus of patients with Alzheimer disease

Immunocytochemical techniques were employed in order to examine the distribution and relative intensity of immunolabeling of the α-amino-3-hydroxy-5-methyl-4-isoaxolepropionate (AMPA) receptor subunits GluR1 and GluR2/3 within the hippocampal formation of patients with Alzheimer disease (AD). Within sectors of the hippocampus that are particularly vulnerable to AD pathology (i.e., CA1, subiculum), we observed a variable loss of GluR1 and GluR2/3 immunolabeling correlating with the extent of cell loss and neurofibrillary pathology. In contrast, in less vulnerable sectors of the hippocampus (i.e., CA2/3, dentate gyrus), the intensity of immunolabeling was markedly increased in AD cases, particularly in the molecular and polymorphic layear of the dentate gyrus. Importantly, these latter regions correspond to termination zones of glutamatergic perforant pathway axons and mossy fiber collaterals, respectively. The increase in immunolabeling within these projection fields is hypothesized to occur in response to the deafferentation of selected glutamatergic pathways, and suggests a critical role for AMPA receptor subunits in hippocampal plasticity.     

AMPA-Selective glutamate receptor subtype immunoreactivity in the hippocampal dentate gyrus of patients with Alzheimer disease

Immunocytochemical techniques were employed in order to examine the distribution and relative intensity of immunolabeling of the α-amino-3-hydroxy-5-methyl-4-isoaxolepropionate (AMPA) receptor subunits GluR1 and GluR2/3 within the hippocampal formation of patients with Alzheimer disease (AD). Within sectors of the hippocampus that are particularly vulnerable to AD pathology (i.e., CA1, subiculum), we observed a variable loss of GluR1 and GluR2/3 immunolabeling correlating with the extent of cell loss and neurofibrillary pathology. In contrast, in less vulnerable sectors of the hippocampus (i.e., CA2/3, dentate gyrus), the intensity of immunolabeling was markedly increased in AD cases, particularly in the molecular and polymorphic layear of the dentate gyrus. Importantly, these latter regions correspond to termination zones of glutamatergic perforant pathway axons and mossy fiber collaterals, respectively. The increase in immunolabeling within these projection fields is hypothesized to occur in response to the deafferentation of selected glutamatergic pathways, and suggests a critical role for AMPA receptor subunits in hippocampal plasticity.     

Alterations of ubiquitin immunoreactivity in the hippocampal formation after perforant pathway lesion

Immunohistochemical techniques were employed to examine the changes in free ubiquitin within the hippocampus 1, 3, 7, 14, and 30 days after a unilateral perforant pathway lesion occurred in the rat brain. Immunoreactivity for ubiquitin was remarkably decreased in the cell body and proximal dendrites of neurons throughout the hippocampus ipsilateral to the lesion at 1 day post-lesion. At 3 days post-lesion, ubiquitin immunoreactivity was recovered in interneurons in the whole hippocampus as well as in mossy cells in the hilar region, although granule cells in the dentate gyrus and pyramidal cells in the CA1 subfield remained unlabeled, and pyramidal cells in the CA3 subfield demonstrated only weak immunoreactivity. In addition, we observed an increase in ubiquitin immunolabeling of the hilar neuropil ipsilateral to the lesion at 1 and 3 days post-lesion, and a decrease in immunolabeling in the inner portion of the molecular layer at 3 days post-lesion. All these alterations were transient, and by 7 days post-lesion, ubiquitin immunoreactivity was indistinguishable in the hippocampus ipsilateral to the lesion, compared to the controls. Immunoblot analysis also revealed a decrease in the amount of ubiquitin in the hippocampus ipsilateral to the lesion 1 and 3 days post-lesion. These data suggest that deafferentation of the perforant pathway results in transient reduction in free ubiquitin of the hippocampus, and that the ubiquitin system is involved in hippocampal plasticity following perforant lesions.     

AMPA-selective glutamate receptor subtype immunoreactivity in the hippocampal formation of patients with Alzheimer's disease

Immunocytochemical techniques were employed in order to examine the distribution and relative intensity of the AMPA receptor subunits GluR1 and GluR2/3 within the hippocampal formation of normal controls and Alzheimer's disease (AD) cases. Throughout our investigation we examined cases exhibiting a wide range of pathologic severity, thus allowing us to correlate our immunohistochemical data with the extent of pathology. Specifically, we investigated the distribution of these receptor subunits in hippocampal sectors that are particularly vulnerable to AD pathology (i. e., CA1 and subiculum) and compared these findings with those obtained following examination of sectors that are generally resistant to pathologic change (i. e., CA2/3, dentate gyrus). Within vulnerable sectors we observed a variable loss of GluR1 and GluR2/3 immunolabeling. The degree to which these proteins were reduced appeared to correlate with the extent of neurofibrillary pathology and cell loss. Despite the loss of labeled cells, the intensity of immunolabeling within the remaining neurons was comparable with, and in many instances even greater than, that observed in control cases. Within resistant sectors, the distribution of immunoreactive elements was comparable in both case groups yet the intensity of immunolabeling was markedly increased in AD cases, particularly in the molecular layer of the dentate gyrus and in the stratum lucidum of CA3 (i. e., the termination zones of perforant pathway and mossy fibers). In addition, within AD cases dramatic increases were observed within the supragranular and polymorphic layer of the dentate gyrus (i. e., the terminal zones of sprouting mossy fiber collaterals). The increase in GluR1 and GluR2/3 immunolabeling is hypothesized to occur in response to the deafferentation of selected glutamatergic pathways. Moreover, our data support that hippocampal plasticity is preserved, even in severe AD cases, and suggest a critical role for AMPA receptor subunits in this plasticity and in maintaining hippocampal functioning. © 1995 Wiley-Liss, Inc.     

Long‐lasting enhanced expression in the rat hippocampus of NMDAR1 splice variants in a kainate model of epilepsy

Chronic epilepsy is associated with increased excitability which may result from abnormal glutamatergic synaptic transmission involving altered properties of N-methyl-d -aspartate (NMDA) receptors. To date two gene families encoding NMDA receptor subunits have been cloned, NR1 and NR2. Eight NR1 mRNAs are generated by alternative splicing of exons 5, 21 and 22; the NR1–1 to NR1–4 C-terminal variants exist in the a or b version depending on the presence or absence of the domain encoded by exon 5. Epilepsy was induced in rats by unilateral intra-amygdalar injection of kainate and animals were killed from 6 h to 4 months following the injection. Increased NR1 mRNA levels were observed during status epilepticus (6–24 h after the injection), both ipsilateral and contralateral, while a second wave of NMDAR1 mRNA increase occurred in chronic epileptic animals, between 21 days and 4 months following kainate injection. Our data show: (i) a permanent increase of the NR1–2a and NR1–2b mRNA species (containing exon 22) in all hippocampal fields, both ipsilateral and contralateral, and (ii) an increase of the NR1–3 (a and b) mRNAs (containing exon 21) in the ipsilateral CA1, and NR1–3a mRNA in the ipsilateral dentate gyrus. No long-term changes were observed for the NR1–1 and NR1–4 splice variants. In the ipsilateral CA3 area a globally decreased mRNA expression was associated with neuronal loss. A possible contribution to the maintenance of the epileptic state by an increased expression of NMDA receptors is discussed.     

Asymmetrical reductions of hippocampal NMDAR1 glutamate receptor mRNA in the psychoses

The psychotomimetic properties of NMDA glutamate receptor antagonists suggest there may be disease related changes of this receptor in schizophrenia. Using in situ hybridisation histochemistry (ISHH), we measured mRNA for the obligatory NMDAR1 subunit of the NMDA glutamate receptor in post-mortem samples of hippocampus from schizophrenics, depressives, bipolar patients and normal controls. A significant main effect of diagnosis was observed in the dentate gyrus (ANOVA, p = 0.004) and a trend in the CA3 region (ANOVA, p = 0.06), with all psychiatric groups having reduced NMDAR1 mRNA levels compared to normal controls. In contrast to the affectively ill groups, the reductions in schizophrenics were more pronounced in the left side compared to the right. Expression of poly A mRNA also showed left-sided losses in the dentate gyrus in schizophrenia but reductions in NMDAR1 remained significant when expressed as a ratio of poly A. The findings confirm a recent report of reduced hippocampal NMDAR1 mRNA in schizophrenia. However, our new evidence suggests that this is a feature of both affective and schizophrenic disorders and that schizophrenia is distinguished from the others by left-sided reductions in hippocampal NMDAR1 gene expression.     

Bidirectional redistribution of AMPA but not NMDA receptors after perforant path simulation in the adult rat hippocampus in vivo

Long-term potentiation (LTP) in vitro reveals dynamic regulation of synaptic glutamate receptors. AMPA receptors may be inserted into synapses to increase neurotransmission, whereas NMDA receptors may redistribute within the synapse to alter the probability of subsequent plasticity. To date, the only evidence for these receptor dynamics in the hippocampus is from the studies of dissociated neurons and hippocampal slices taken from young animals. Although synaptic plasticity is induced easily, the extent of AMPA and NMDA receptor mobility after LTP is unknown in the adult, intact hippocampus. To test whether AMPA or NMDAR subunits undergo activity-dependent modifications in adult hippocampal synapses, we induced LTP at perforant path-dentate gyrus (DG) synapses in anesthetized adult rats, using high frequency stimulation (HFS), verified layer-specific Arc induction, and analyzed the distribution of postsynaptic AMPA and NMDAR subunits, using immunogold electron microscopy. The number of synapses with AMPA receptor labeling increased with LTP-inducing HFS in the stimulated region of the dendrite relative to the nonstimulated regions. The opposite trend was noted with low frequency stimulation (LFS). Moreover, HFS increased and LFS decreased the ratio of synaptic to extrasynaptic AMPA receptor labeling in the postsynaptic membrane. In contrast, HFS did not significantly alter NMDAR labeling. Thus, LTP in the adult hippocampus in vivo selectively enhanced AMPA but not NMDAR labeling specifically in synapses undergoing activity-dependent plasticity relative to the remainder of the dendritic tree. The results suggest a mechanism by which rapid adjustments in synaptic strength can occur through localized AMPA receptor mobility and that this process may be competitive across the dendritic tree.     

Synaptic coexistence of AMPA and NMDA receptors in the rat hippocampus: A postembedding immunogold study

Synaptic distributions of N-methyl-d -aspartate (NMDA) and α-amino-3-hydroxy-5-methylisoxazolepropionic acid (AMPA) receptor subunits, NMDAR1 and GluR2, respectively, were examined by electron microscopy with the high spatial resolution of postembedding immunogold localization. We provide direct evidence for colocalization at individual axodendritic asymmetric synapses within the CA1 subfield of rat hippocampus. AMPA/ NMDA receptor colocalization was found both in γ-aminobutyric acid (GABA)ergic dendrites and non-GABAergic dendritic shafts, as well as dendritic spines. Some asymmetric synapses were found to contain only NMDAR1 or GluR2; however, most immunopositive synapses contained both subunits. Many NMDAR1 and/ or GluR2 immunopositive profiles received GABAergic innervation at an adjacent synapse, providing a substrate for GABAergic modulation of both GluR classes. These data suggest that excitatory neuronal transmission in CA1 neurons may generally involve activation of both NMDA and AMPA receptor subunits at a single synapse, however, they also offer ultrastructural evidence for NMDAR1-only synapses that might represent silent synapses. J. Neurosci. Res. 54:444–449, 1998. © 1998 Wiley-Liss, Inc.     

Glutamate Receptor Activation in the Kindled Dentate Gyrus

Summary: Purpose : The contribution of α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA), N -methyl- d -aspartate (NMDA), and kainate receptor activation to the enhanced seizure susceptibility of the dentate gyrus was investigated in an experimental model of temporal lobe epilepsy.
Methods : Using the specific NMDA and AMPA receptor antagonists D-APV and SYM 2206, we examined alterations in glutamate receptor-dependent synaptic currents 48 hours and 28 days after kindling in field-potential and voltage-clamp recordings.
Results : Forty-eight hours after kindling, the fractions of AMPA and NMDA receptor-mediated excitatory postsynaptic current components shifted dramatically in favor of the NMDA receptor-mediated response. Four weeks after kindling, however, AMPA and NMDA receptor-mediated excitatory postsynaptic currents reverted to control-like values. Neither single nor repetitive perforant path stimuli evoked kainate receptor-mediated excitatory postsynaptic currents in dentate gyrus granule cells of control or kindled rats.
Conclusion : The enhanced excitability of the kindled dentate gyrus 48 hours after the last seizure most likely results from transiently enhanced NMDA receptor activation. The NMDA receptor seems to play a critical role in the induction of the kindled state rather than in the persistence of the enhanced seizure susceptibility.     

Distribution of N-methyl-D-aspartate and non-N-methyl-D-aspartate glutamate receptor subunits on respiratory motor and premotor neurons in the rat

Glutamate is required for the transmission of inspiratory drive in respiratory premotor and motor neurons. The glutamate receptors (GluRs) responsible for this essential function have yet to be anatomically characterized. We mapped the GluR subtypes expressed by respiratory premotor and motor neurons by using combined immunohistochemistry and retrograde labeling in adult rats. Phrenic motoneurons and bulbospinal ventral respiratory group (VRG) neurons were retrogradely labeled and immunolabeled with subunit-specific antibodies against the N-methyl-D-aspartate (NMDA) receptor subtype (NMDAR1) and the non-NMDA receptor subtypes, α-amino-3-hydroxy-5-methylisoxazole-4-proprionic acid (AMPA; GluR1, GluR2/3, GluR4) and kainate (GluR5–7). Phrenic motoneurons and bulbospinal VRG neurons showed positive immunolabeling for all five GluR subunits. These results support the hypothesis that NMDA and non-NMDA receptor subtypes underlie the excitation of bulbospinal VRG neurons and phrenic motoneurons. Furthermore, immunolabeling for each receptor subtype demonstrated a unique distribution along the neuronal membrane. Immunoreactivity for AMPA receptor subunits was distributed throughout somata and proximal dendrites, NMDAR1 subunit immunolabeling was localized to somata, and GluR5–7 subunit immunolabeling was confined largely to dendrites. The differential distribution of AMPA, kainate, and NMDA receptors on the somal and dendritic surface of respiratory neurons suggests that the location of glutamatergic synapses along the neuronal surface is an important determinant of glutamate-mediated postsynaptic currents. Consequently, different patterns of glutamatergic excitation of respiratory neurons could be achieved by selective activation of different profiles of GluR subtypes on different portions of the neuronal membrane. J. Comp. Neurol. 389:94–116, 1997. © 1997 Wiley-Liss, Inc.     

Sequential changes in AMPA and NMDA protein levels during Fe(3+)-induced epileptogenesis

Seizure susceptibility is related to enhanced glutamatergic excitatory synaptic transmission with alterations in the expressions of ionotropic glutamate receptors. We wondered if levels of AMPA and NMDA receptor subunits changed following epileptogenesis induced by amygdalar FeCl(3) injection. We used Western blots to measure levels of subunits in the ipsilateral and contralateral hippocampus at various times after FeCl(3) injection into the amygdaloid body. With acute seizures, at +5 days after the injection, levels of GluR1, NMDAR1, and NMDAR2 were markedly increased in both hippocampi, with quantities at least 2-4 times baseline. By +15 and +30 days after injection, when chronic spontaneous seizures were occurring, the levels of GluR2 were increased, while GluR1 and NMDAR1&2A/B were decreased. Increased NMDAR1&2A/B levels at +5 days are consistent with the occurrence of upregulation of NMDA receptor production in the early stages of epileptogenesis. Since GluR2 suppresses glutamate receptor-mediated Ca(2+)-influx, increased expression of GluR2 with development of chronic, recurrent seizures may be a compensatory effect during epileptogenesis from neural responses to propagated seizures.     

Synaptic recruitment during long-term potentiation at synapses of the medial perforant pathway in the dentate gyrus of the rat brain

Long-term potentiation (LTP) in synapses of the medial perforant pathway of the rat dentate gyrus has been studied using the whole-cell voltage clamp technique and a standard hippocampal slice preparation. The rate of LTP induction by 2–4 brief trains of stimuli at 100 Hz, paired with postsynaptic depolarization to −20 mV, in individual granule neurons was only 42% but the average magnitude was large. In a representative series of nine experiments the average potentiation was 339% (s.d. 255%). The variable magnitude of LTP appeared to be related to the relative size of the NMDA receptor dependent current in individual neurons. LTP was further characterized by the selective enhancement of the AMPA (but not the NMDA) component in the excitatory synaptic responses. This selective enhancement of the AMPA component and a graphical variance analysis suggest that the large magnitude of LTP in dentate gyrus can be best explained by recruitment of previously silent synapses by a combination of pre- and post-synaptic mechanisms. © 1996 Wiley-Liss, Inc.     

Ultrastructural evidence for presynaptic mu opioid receptor modulation of synaptic plasticity in NMDA-receptor-containing dendrites in the dentate gyrus

Physiological studies have demonstrated that long-term potentiation (LTP) induction in N-methyl-D-aspartate (NMDA) receptor containing dentate granule cells following lateral perforant path stimulation is opioid dependent, involving mu-opioid receptors (MORs) on gamma-aminobutyric acid (GABA)-ergic neurons. To determine the cellular relationships of MORs to postsynaptic NMDA receptor-containing dendrites, immunoreactivity (-I) against MOR and the NMDA receptor subunit 1 (NMDAR1) was examined in the outer molecular layer of the dentate gyrus using electron microscopy. MOR-I was predominantly in axons and axon terminals. NMDAR1-I was almost exclusively in spiny dendrites, but was also in a few terminals. Using immunogold particles to localize precisely NMDAR1, one-third of the NMDAR1-I was detected on the dendritic plasmalemma; in dendritic spines plasmalemmal immunogold particles were near synaptic densities. Many MOR-labeled axons and terminals contacted NMDAR1-labeled dendrites. MOR-labeled terminals formed symmetric (inhibitory-type) synapses on NMDAR1-labeled dendritic shafts or nonsynaptically contacted NMDAR1-labeled shafts and spines. MOR-labeled axons often abutted NMDAR1-containing dendritic spines which received asymmetric (excitatory-type) synapses from unlabeled terminals. Occasionally, MOR-labeled terminals and dendrites were apposed to NMDAR1-containing terminals. These results provide anatomical evidence that endogenous enkephalins or exogenous opioid agonists could inhibit GABAergic terminals that modulate granule cell dendrites, thus boosting depolarizing events in granule cells and facilitating the activation of NMDA receptors located on their dendrites.     

Long-term potentiation in the hippocampal CA1 area and dentate gyrus plays different roles in spatial learning

NMDA receptor-dependent long-term potentiation (LTP) at hippocampal synapses has been considered a crucial component of the cellular basis for learning and memory. This form of LTP occurs in excitatory synapses in both the CA1 area and the dentate gyrus in the hippocampus. However, differential roles of LTP in these areas have not yet been identified. To address this issue, we enhanced the degree of LTP by expressing Ca2+-permeable AMPA receptors at either hippocampal CA1 or dentate gyrus synapses using Sindbis viral vectors (SINs) encoding both green fluorescent proteins and unedited GluR2 (GluR2Q) subunits, and examined their effects on rat spatial learning. The viral vectors were locally injected into the 8-week-old-rat brain in vivo bilaterally. The postsynaptic expression of Ca2+-permeable AMPA receptors enhanced the degree of LTP, and induced NMDA receptor-independent LTP in the presence of the NMDA receptor antagonist in SIN-infected regions in both CA1 and dentate gyrus in hippocampal slice preparations. However, the regional expression of Ca2+-permeable AMPA receptors caused opposite behavioural consequences on the Morris water maze task: rats with SIN-infected CA1 pyramidal cells showed shorter escape latency and better probe test performance, whereas those with SIN-infected dentate gyrus granule cells showed impaired performance. Thus, it was demonstrated that CA1 and dentate gyrus synapses play different functional roles in spatial learning despite their similar mechanism for LTP induction.     

Spontaneous and synaptic input from granule cells and the perforant path to dentate basket cells in the rat hippocampus

To characterize excitatory inputs to dentate basket cells from dentate granule cells and the perforant path, the whole-cell recording technique was used in neonatal rat hippocampal slices. Spontaneous excitatory input to basket cells was also examined and compared to that of other interneurons in the dentate gyrus. Basket cells were separable from other neurons in the dentate gyrus based on morphology and location, as determined by biocytin staining following recording, and by resting membrane potential, propensity to fire action potentials spontaneously, and miniature excitatory postsynaptic current (EPSC) characteristics. Minimal electrical stimulation of the granule cell layer evoked in basket cells short latency EPSCs that were composed of both N-methyl-D-aspartate (NMDA) and α-amino-3-hydroxy-5-methyl-4-isoxazole-propionate (AMPA) components as judged by their time course, voltage dependence, and blockade by selective antagonists. Perforant path EPSCs exhibited slower kinetics than EPSCs evoked by granule cell stimulation. Like granule cell evoked EPSCs, however, perforant path EPSCs were composed of both NMDA and AMPA components. Minimal electrical stimulation of the granule cell layer and perforant path evoked monosynaptic EPSCs in only 67% and 62% of the trials, respectively, suggesting that these inputs are as unreliable as previously determined inputs from CA3 pyramidal cells (48%). Tetrodotoxin-insensitive spontaneous miniature EPSCs were frequent in basket cells and non-basket interneurons residing either at the border between the granule cell layer and the hilus or deep within the hilus. Miniature EPSCs recorded from all cells held at ?70 mV were blocked completely by 3 μSM 6-cyano-7-nitro-quinoxaline-2,3-dione (CNQX). Though a component of the miniature EPSCs recorded from border and deep hilar interneurons at +40 mV was blocked by the NMDA receptor antagonist D -2-amino-phosphonovaleric acid (D-APV), miniature EPSCs in basket cells were insensitive to D-APV. We conclude that input from granule cells and the perforant path results in activation of basket cells via glutamatergic synapses that employ both NMDA and AMPA receptors. These inputs to basket cells likely contribute to feedback and feedforward inhibition of granule cells. The absence of an NMDA receptor component in spontaneous miniature EPSCs of dentate basket cells implies a difference in organization of excitatory synapses made onto basket cells compared with other hilar interneurons. © 1995 Wiley-Liss, Inc.