AMPA and kainate receptors each mediate excitotoxicity in oligodendroglial cultures

Recent studies indicate that oligodendrocytes are vulnerable to excitotoxic insults mediated by glutamate receptors. The present study was carried out to characterize the type of glutamate receptors triggering cell death in optic nerve oligodendrocyte cultures. Acute activation of either AMPA or kainate receptors was toxic to oligodendrocytes, an effect that was prevented by CNQX. However, exposure to agonists of the NMDA and metabotropic glutamate receptors did not impair cell viability. Dose-response curves showed that toxicity was mediated by three distinct populations of receptors: an AMPA-type receptor and high- and low-affinity kainate-type receptors. Expression and immunocytochemical studies suggested that the glutamate receptor subunits give rise to the native receptors in each population. In all instances, Ca(2+) entry was a major determinant of glutamate receptor excitotoxicity. However, its influence varied for each receptor subtype. These results indicate that aberrantly enhanced activation of AMPA and/or kainate receptors may be involved in demyelinating diseases.     

The link between excitotoxic oligodendroglial death and demyelinating diseases

Oligodendrocytes, the myelinating cells of CNS axons, are highly vulnerable to excitotoxic signals mediated by glutamate receptors of the AMPA and kainate classes. Receptors in these cells are commonly activated by glutamate that is released from axons and glial cells. In addition, oligodendrocytes contribute to the control of extracellular glutamate levels by means of their own transporters. However, acute and chronic alterations in glutamate homeostasis can result in overactivation of AMPA and kainate receptors and subsequent excitotoxic oligodendroglial death. Furthermore, demyelinating lesions caused by excitotoxins can be similar to those observed in multiple sclerosis. This, together with the effect of AMPA and kainate receptor antagonists in ameliorating the neurological score of animals with experimental autoimmune encephalomyelitis (an animal model of multiple sclerosis), indicates that oligodendrocyte excitotoxicity could be involved in the pathogenesis of demyelinating disorders.     

AMPA receptors are the major mediators of excitotoxic death in mature oligodendrocytes

Myelination of axons is important for central nervous system function, but oligodendrocytes, which constitute CNS myelin, are vulnerable to excitotoxic injury and death. Although mature oligodendrocytes express functional alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionic (AMPA) and kainate-type glutamate receptors, the relative roles of these subtypes in excitotoxicity are not well understood. Using recently developed selective antagonists for subtypes of ionotropic non-NMDA receptors, we addressed this issue. By examining the pharmacological, biochemical, and morphologic features of kainite-induced excitotoxic death, we also determined whether it occurs by apoptosis, necrosis, or both. We conclude that when mature oligodendrocytes die after exposure to kainate: (1) AMPA receptors are the most important mediators, (2) kainate receptors play a smaller role, and (3) death occurs predominantly by necrosis, not apoptosis.     

Failure to form a stable topographic map during optic nerve regeneration: abnormal activity-dependent mechanisms

Visually evoked responses in the optic tectum are mediated by glutamate receptors. During development, there is a switch from N-methyl-d-aspartate (NMDA)- to alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA)-mediated activity as the retinotectal map refines and visual function ensues. A similar pattern is seen in goldfish as the map refines during optic nerve regeneration. Here we examined glutamate receptors during optic nerve regeneration in the lizard, Ctenophorus ornatus, in which an imprecise retinotopic map forms transiently but degrades, leaving animals blind via the experimental eye. Receptor function was examined using NMDA and AMPA/kainate antagonists during in vitro tectal recording of visually evoked post-synaptic extracellular responses. Expression of NR1 (NMDA) and GluR2 (AMPA) receptor subtypes was examined immunohistochemically. In unoperated control animals, responses were robust and AMPA/kainate receptor-mediated. When the imprecise map was present, responses were difficult to evoke and insecure; periods of spontaneous activity as well as inactivity were also noted. Although AMPA/kainate-mediated activity persisted and GluR2 immunoreactivity increased transiently, NMDA receptor-mediated activity was also consistently detected and NR1 expression increased. In the long term, when the map had degraded, responses were readily evoked and predominantly AMPA/kainate receptor-mediated although some NMDA-mediated activity and NR1 expression remained. We suggest that the asynchronous activity reaching the optic tectum results in an inability to recapitulate the appropriate functional sequences of expression of NMDA and AMPA/kainate receptors necessary to refine the retinotectal map.     

Glutamate-mediated excitotoxicity in neonatal hippocampal neurons is mediated by mGluR-induced release of Ca++ from intracellular stores and is prevented by estradiol

Hypoxic/ischemic (HI) brain injury in newborn full-term and premature infants is a common and pervasive source of life time disabilities in cognitive and locomotor function. In the adult, HI induces glutamate release and excitotoxic cell death dependent on NMDA receptor activation. In animal models of the premature human infant, glutamate is also released following HI, but neurons are largely insensitive to NMDA or AMPA/kainic acid (KA) receptor-mediated damage. Using primary cultured hippocampal neurons we have determined that glutamate increases intracellular calcium much more than kainic acid. Moreover, glutamate induces cell death by activating Type I metabotropic glutamate receptors (mGluRs). Pretreatment of neurons with the gonadal steroid estradiol reduces the level of the Type I metabotropic glutamate receptors and completely prevents cell death, suggesting a novel therapeutic approach to excitotoxic brain damage in the neonate.     

Mechanism of acute ischemic injury of oligodendroglia in early myelinating white matter: the importance of astrocyte injury and glutamate release

In developing CNS white matter (WM), the period of early myelination is characterized by a heightened sensitivity to ischemic injury. Using an in situ (isolated) preparation, we show that the mechanism of acute ischemic injury of immature WM oligodendroglial involves Ca2+ influx though non-NMDA type glutamate receptors (GluRs). The Ca2+-influx and acute cell death that was evoked by ischemic conditions (oxygen and glucose withdrawal) in identified P10 rat optic nerve oligodendroglia were blocked by removing extracellular Ca2+ or by CNQX, a selective non-NMDA GluR antagonist. The selective Na-K-Cl cotransport (NKCC) inhibitor bumetanide was also highly protective, even though NKCC expression is restricted to astrocytes in this tissue. Bumetanide largely prevented the non-NMDA GluR-mediated [Ca2+]i rise evoked by ischemia in oligodendroglia, suggesting that it interfered with ischemic glutamate release. In control WM, glutamate-like reactivity was located mainly in astrocytes and oligodendroglia identified using ultrastructural criteria. In ischemic WM, astrocyte glutamate-like reactivity was reduced, an effect countered by bumetanide. We suggest a model in which NKCC-dependent injury and release of glutamate from astrocytes activates glutamate receptors on oligodendroglia, resulting in Ca2+-influx and acute cell death.     

Testing NMDA receptor block as a therapeutic strategy for reducing ischaemic damage to CNS white matter

Damage to oligodendrocytes caused by glutamate release contributes to mental or physical handicap in periventricular leukomalacia, spinal cord injury, multiple sclerosis, and stroke, and has been attributed to activation of AMPA/kainate receptors. However, glutamate also activates unusual NMDA receptors in oligodendrocytes, which can generate an ion influx even at the resting potential in a physiological [Mg2+]. Here, we show that the clinically licensed NMDA receptor antagonist memantine blocks oligodendrocyte NMDA receptors at concentrations achieved therapeutically. Simulated ischaemia released glutamate which activated NMDA receptors, as well as AMPA/kainate receptors, on mature and precursor oligodendrocytes. Although blocking AMPA/kainate receptors alone during ischaemia had no effect, combining memantine with an AMPA/kainate receptor blocker, or applying the NMDA blocker MK-801 alone, improved recovery of the action potential in myelinated axons after the ischaemia. These data suggest NMDA receptor blockers as a potentially useful treatment for some white matter diseases and define conditions under which these blockers may be useful therapeutically. Our results highlight the importance of developing new antagonists selective for oligodendrocyte NMDA receptors based on their difference in subunit structure from most neuronal NMDA receptors.     

Astrocyte-Dependent Vulnerability to Excitotoxicity in Spermine Oxidase-Overexpressing Mouse

Transgenic mice overexpressing spermine oxidase (SMO) in the cerebral cortex (Dach-SMO mice) showed increased vulnerability to excitotoxic brain injury and kainate-induced epileptic seizures. To investigate the mechanisms by which SMO overexpression leads to increased susceptibility to kainate excitotoxicity and seizure, in the cerebral cortex of Dach-SMO and control mice we assessed markers for astrocyte proliferation and neuron loss, and the ability of kainate to evoke glutamate release from nerve terminals and astrocyte processes. Moreover, we assessed a possible role of astrocytes in an in vitro model of epileptic-like activity in combined cortico-hippocampal slices recorded with a multi-electrode array device. In parallel, as the brain is a major metabolizer of oxygen and yet has relatively feeble protective antioxidant mechanisms, we analyzed the oxidative status of the cerebral cortex of both SMO-overexpressing and control mice by evaluating enzymatic and non-enzymatic scavengers such as metallothioneins. The main findings in the cerebral cortex of Dach-SMO mice as compared to controls are the following: astrocyte activation and neuron loss; increased oxidative stress and activation of defense mechanisms involving both neurons and astrocytes; increased susceptibility to kainate-evoked cortical epileptogenic activity, dependent on astrocyte function; appearance of a glutamate-releasing response to kainate from astrocyte processes due to activation of Ca²⁺-permeable AMPA receptors in Dach-SMO mice. We conclude that reactive astrocytosis and activation of glutamate release from astrocyte processes might contribute, together with increased reactive oxygen species production, to the vulnerability to kainate excitotoxicity in Dach-SMO mice. This mouse model with a deregulated polyamine metabolism would shed light on roles for astrocytes in increasing vulnerability to excitotoxic neuron injury.     

A revised excitotoxic hypothesis of schizophrenia: therapeutic implications

A revision of an "excitotoxic hypothesis" of schizophrenia is summarized. The hypothesis suggests that in, at least, a subtype of patients with schizophrenia, progressive excitotoxic neuronal cell death in hippocampal and cortical areas occurs via "disinhibition" of glutamatergic projections to these areas. Patients who have excitotoxic damage would be expected to have poor outcomes characterized, perhaps, by anatomic evidence of progressive neurodegeneration, pronounced negative symptoms and cognitive deficits, and profound psychosocial deterioration. Disinhibited glutamatergic activity could result from inhibition of N-methyl-D-aspartate (NMDA) receptor-mediated neurotransmission and a consequent failure to stimulate inhibitory gamma-aminobutyric acid (GABA)-ergic interneurons, and/or anatomic degeneration of inhibitory GABAergic interneurons. The result of these hypothesized mechanisms is excessive stimulation of the alpha-amino-3-hydroxy-5-methylisoxazole-4-propionic acid (AMPA)/kainate class of glutamate receptor complexes. In turn, this excessive stimulation of AMPA/kainate receptors could lead to disruption of ionic gradients, depletion of energy reserves expended in an attempt to restore and maintain the ionic disequilibrium across neuronal membranes, generation of reactive oxygen species, and cell death from apoptotic and other mechanisms. The postulated existence of disinhibited glutamatergic neurotransmission and the subsequent cascade of excitotoxic events resulting from NMDA receptor hypofunction (NRH), anatomic degeneration of inhibitory GABAergic interneurons, or a combination of the two has suggested a diverse variety of experimental therapeutic interventions for schizophrenia. These interventions include facilitation of NMDA receptor-mediated neurotransmission, potentiation of GABAergic neurotransmission, antagonism of AMPA/kainate receptors, and "quenching" of locally generated reactive oxygen species. In fact, several of these approaches have already been pursued or are proposed as part of a systematic clinical investigation of the revised excitotoxic hypothesis of schizophrenia.     

Cleavage of the vesicular glutamate transporters under excitotoxic conditions

Glutamate is loaded into synaptic vesicles by vesicular glutamate transporters (VGLUTs), and alterations in the transporters expression directly regulate neurotransmitter release. We investigated changes in VGLUT1 and VGLUT2 protein levels after ischemic and excitotoxic insults. The results show that VGLUT2 is cleaved by calpains after excitotoxic stimulation of hippocampal neurons with glutamate, whereas VGLUT1 is downregulated to a lower extent. VGLUT2 was also cleaved by calpains after oxygen/glucose deprivation (OGD), and downregulated after middle cerebral artery occlusion (MCAO) and intrahippocampal injection of kainate. In contrast, VGLUT1 was not affected after OGD. Incubation of isolated synaptic vesicles with recombinant calpain also induced VGLUT2 cleavage, with a little effect observed for VGLUT1. N-terminal sequencing analysis showed that calpain cleaves VGLUT2 in the C-terminus, at Asn⁵³⁴ and Lys⁵⁴². The truncated GFP-VGLUT2 forms were found to a great extent in non-synaptic regions along neurites, when compared to GFP-VGLUT2. These findings show that excitotoxic and ischemic insults downregulate VGLUT2, which is likely to affect glutamatergic transmission and cell death, especially in the neonatal period when the transporter is expressed at higher levels.     

Differential role of mGlu1 and mGlu5 receptors in rat hippocampal slice models of ischemic tolerance

Activation of glutamate receptors has been proposed as a key factor in the induction of ischemic tolerance. We used organotypic rat hippocampal slices exposed to 30 min oxygen-glucose deprivation (OGD) to evaluate postischemic pyramidal cell death in the CA1 subregion. In this model, 10 min exposure to OGD 24 h before the exposure to toxic OGD was not lethal and reduced the subsequent OGD neurotoxicity by approximately 53% (ischemic preconditioning). Similarly, a 30 min exposure to the group I mGlu receptor agonist DHPG (10 microM) significantly reduced OGD neurotoxicity 24 h later (pharmacological preconditioning). Ischemic tolerance did not develop when either the selective mGlu1 antagonists LY367385 and 3-MATIDA or the AMPA/KA antagonist CNQX were present in the incubation medium during exposure to sublethal OGD. Neither the NMDA antagonist MK801 nor the mGlu5 antagonist MPEP affected the preconditioning process. On the other hand, pharmacological preconditioning was prevented not only by LY367385 or CNQX, but also by MPEP. In preconditioned slices, the toxic responses to AMPA or NMDA were reduced. The neurotoxicty of 100 microM DHPG in slices simultaneously exposed to a mild (20 min) OGD was differentially altered in the two preconditioning paradigms. After ischemic preconditioning, DHPG neurotoxicity was reduced in a manner that was sensitive to LY367385 but not to MPEP, whereas after pharmacological preconditioning it was enhanced in a manner that was sensitive to MPEP but not to LY367385. Our results show that mGlu1 and mGlu5 receptors are differentially involved in the induction and expression of ischemic tolerance following two diverse preconditioning stimuli.     

The balance of NMDA- and AMPA/kainate receptor-mediated activity in normal adult goldfish and during optic nerve regeneration

Retinotectal topography is established during development and relies on the sequential recruitment of glutamate receptors within postsynaptic tectal cells. NMDA receptors underpin plastic changes at early stages when retinal ganglion cell (RGC) terminal arbors are widespread and topography is coarse; AMPA/kainate receptors mediate fast secure neurotransmission characteristic of mature circuits once topography is refined. Here, we have examined the relative contributions of these receptors to visually evoked activity in normal adult goldfish, in which retinotectal topography is constantly adjusted to compensate for the continual neurogenesis and the addition of new RGC arbors. Furthermore, we examined animals at two stages of optic nerve regeneration. In the first, RGC arbors are widespread and receptive fields large resulting in coarse topography; in the second, RGC arbors are pruned to reduce receptive fields leading to refined topography. Antagonists were applied to the tectum during multiunit recording of postsynaptic responses. Normal goldfish have low levels of NMDA receptor-mediated activity and high levels of AMPA/kainate. When coarse topography has been restored, NMDA receptor-mediated activity is increased and that of AMPA/kainate decreased. Once topography has been refined, the balance of NMDA and AMPA/kainate receptor-mediated activity returns to normal. The data suggest that glutamatergic neurotransmission in normal adult goldfish is dual with NMDA receptors fine-tuning topography and AMPA receptors allowing stable synaptic function. Furthermore, the normal operation of both receptors allows a response to injury in which the balance can be transiently reversed to restore topography and vision.     

AMPA prevents glutamate-induced neurotoxicity and apoptosis in cultured cerebellar granule cell neurons

Exposure of cultured cerebellar neurons to ±-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) in the presence of aniracetam protects all of the vulnerable neurons against the excitotoxic actions of glutamate acting on N-methyl-D-aspartate receptors. The protective effect of AMPA was both time- and concentration-dependent. Aniracetam alone did not protect the neurons against the excitotoxic effects of glutamate. Pretreatment of cerebellar neurons with the AMPA antagonist, 6-cyano-7-nitroquinoxaline-2,3-dione blocked the neuroprotective effect mediated by AMPA indicating that the neuroprotective effect is mediated specifically by AMPA receptors. An excitotoxic concentration of glutamate, which killed between 60–80% of granule cell neurons on day 8in vitro, mediated its toxic effect via a time-dependent apoptotic pathway. Pretreatment of cerebellar granule cell neurons with AMPA (500 μM) completely blocked glutamate-mediated apoptosis. Our results suggest that AMPA receptors may play an important role in neuronal survival.     

TNF-α triggers rapid membrane insertion of Ca2 + permeable AMPA receptors into adult motor neurons and enhances their susceptibility to slow excitotoxic injury

Excitotoxicity (caused by over-activation of glutamate receptors) and inflammation both contribute to motor neuron (MN) damage in amyotrophic lateral sclerosis (ALS) and other diseases of the spinal cord. Microglial and astrocytic activation in these conditions results in release of inflammatory mediators, including the cytokine, tumor necrosis factor‐alpha (TNF-α). TNF-α has complex effects on neurons, one of which is to trigger rapid membrane insertion of α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) type glutamate receptors, and in some cases, specific insertion of GluA2 lacking, Ca^2 + permeable AMPA receptors (Ca-perm AMPAr). In the present study, we use a histochemical stain based upon kainate stimulated uptake of cobalt ions (“Co^2 + labeling”) to provide the first direct demonstration of the presence of substantial numbers of Ca-perm AMPAr in ventral horn MNs of adult rats under basal conditions. We further find that TNF-α exposure causes a rapid increase in the numbers of these receptors, via a phosphatidylinositol 3 kinase (PI3K) and protein kinase A (PKA) dependent mechanism. Finally, to assess the relevance of TNF-α to slow excitotoxic MN injury, we made use of organotypic spinal cord slice cultures. Co^2 + labeling revealed that MNs in these cultures possess Ca-perm AMPAr. Addition of either a low level of TNF-α, or of the glutamate uptake blocker, trans-pyrrolidine-2,4-dicarboxylic acid (PDC) to the cultures for 48 h resulted in little MN injury. However, when combined, TNF-α + PDC caused considerable MN degeneration, which was blocked by the AMPA/kainate receptor blocker, 2,3-Dihydroxy-6-nitro-7-sulfamoylbenzo (F) quinoxaline (NBQX), or the Ca-perm AMPAr selective blocker, 1-naphthyl acetylspermine (NASPM). Thus, these data support the idea that prolonged TNF-α elevation, as may be induced by glial activation, acts in part by increasing the numbers of Ca-perm AMPAr on MNs to enhance injurious excitotoxic effects of deficient astrocytic glutamate transport.     

Glutamate induces rapid loss of axonal neurofilament proteins from cortical neurons in vitro

One of the primary hallmarks of glutamate excitotoxicity is degradation of the neuronal cytoskeleton. Using a tissue culture approach, we have investigated the relationship between excitotoxicity and cytoskeletal degradation within axons, with particular reference to the axon specific neurofilament proteins. Neurofilaments were rapidly lost from axons over a 24-h period in response to excitotoxic insult (as observed by immunocytochemistry and western blotting), while other axonal cytoskeletal markers (such as betaIII-tubulin) remained intact. Treatment with kainic acid and NMDA, or complementary experiments using the pharmacological glutamate receptors blockers CNQX (kainate/AMPA receptor antagonist) and MK-801 (NMDA receptor antagonist), demonstrated that neurofilament degeneration was mediated primarily by NMDA receptor activity. This work suggests that excitotoxicity triggers a progressive pathway of cytoskeletal degeneration within axons, initially characterised by the loss of neurofilament proteins.     

Tissue plasminogen activator protects hippocampal neurons from oxygen-glucose deprivation injury

We have previously shown that tissue plasminogen activator (tPA) participates in the neurotoxicity of microglial conditioned medium (MgCM). Killing of hippocampal neurons by MgCM was prevented by both plasminogen activator inhibitor-1 (PAI-1) and anti-tPA antibody. An N-methyl-D-aspartate (NMDA) receptor blocker protected neurons from MgCM, suggesting that this subtype of glutamate receptor is involved. Whereas glutamate receptor-mediated events are important in cerebral ischemia and tPA has previously been shown to enhance excitotoxicity in hippocampus, we hypothesized that tPA would exaggerate oxygen glucose deprivation (OGD) injury in cultures of hippocampal neurons. Dissociated rat hippocampal cells were grown under conditions designed to optimize neuronal growth while minimizing glial replication. At 7--10 days, cultures were subjected to OGD for 2.5 hr. Recombinant human tPA (1,000 IU) was added immediately after OGD. Viability was assessed 24 hr later. Viable, apoptotic, and necrotic cells were classified and quantified based on staining patterns of acridine orange and ethidium bromide under fluorescence microscopy. tPA alone did not alter neuronal integrity. OGD produced significant neuronal death (viability reduced by 45%, P < 0.001). tPA completely protected OGD-exposed cultures. Potential mechanisms of tPA protection were explored. Whereas tPA antibody abolished the protective effect of tPA, its proteolytic inhibitor PAI-1 did not alter the effect. The effect of tPA was tested in separate free radical and excitatory amino acid insults. It did not protect neurons from hydrogen peroxide (1 microM), S-nitro-acetylpenicillamine (10 microM), glutamate (50 microM), or NMDA (10 microM) damage but significantly attenuated injury caused by 250 microM kainate. We conclude that tPA is capable of protecting hippocampal neurons from OGD by a nonproteolytic action. The mechanism of protection was not defined, although attenuation of AMPA/kainate glutamate receptors may play a role.     

Developmental dissociation of pharmacological and neurotoxic effects of excitatory amino acids.

The development of excitatory amino acid-(EAA)-induced cytotoxic cell death and [3H]gamma-aminobutyric acid ([3H]GABA) release were simultaneously examined in primary cultures of the rat cerebral cortex. Pronounced [3H]GABA release could already be evoked on day 3 by N-methyl-D-aspartate, quisqualate and kainate, whereas toxic cell death could first be induced on day 7, in vitro. EAA-induced GABA release declined between day 11 and 14, but the excitotoxic vulnerability of cells increased further during the same period. This dissociation of releasing and toxic responses indicates that functionally active EAA receptors do not necessarily mediate excitotoxic effects and suggests that the development of EAA receptors mediating release responses precedes the maturation of intracellular mechanisms involved in excitotoxic neuronal injury, at least in cultured cortical neurons.     

Concentrations of glutamate released following spinal cord injury kill oligodendrocytes in the spinal cord

We investigated in vivo in rats whether sufficient glutamate is released following spinal cord injury (SCI) to kill oligodendrocytes. Microdialysis sampling was used to establish the level of glutamate released (550 +/- 80 microM) in the white matter during SCI. This glutamate concentration was administered into the spinal cords of other rats and the densities of oligodendrocytes remaining 24 and 72 h later determined by counting cells immunostained with the oligodendrocyte marker CC-1. Administration of ACSF, 4.0 mM glutamate (estimated resulting tissue exposure 500 microM) and 10.0 mM glutamate by microdialysis reduced oligodendrocyte density 22%, 57%, and 74%, respectively, relative to normal at 24 h post-exposure. Therefore, sufficient glutamate is released following SCI to damage white matter. Oligodendrocyte densities near the fiber track were not significantly different at 72 h from 24 h post-exposure, so most glutamate-induced oligodendrocyte death occurs within 24 h after exposure. Injecting the AMPA/kainate receptor blocker NBQX into the spinal cord during glutamate administration reduced the glutamate-induced decrease in oligodendrocyte density, evidence for AMPA/kainate receptor involvement in glutamate-induced oligodendrocyte death. This work directly demonstrates in vivo that following SCI glutamate reaches concentrations toxic to white matter and that AMPA/kainate receptors mediate this glutamate toxicity to oligodendrocytes.     

Kindling enhances kainate receptor-mediated depression of GABAergic inhibition in rat granule cells

Several lines of evidence indicate a substantial contribution of kainate receptors to temporal lobe seizures. The activation of kainate receptors located on hippocampal inhibitory interneurons was shown to reduce GABA release. A reduced GABA release secondary to kainate receptor activation could contribute to an enhanced seizure susceptibility. As the dentate gyrus serves a pivotal gating function in the spread of limbic seizures, we tested the role of kainate receptors in the regulation of GABA release in the dentate gyrus of control and kindled animals. Application of glutamate (100 micro m) in the presence of the NMDA receptor antagonist d-APV and the AMPA receptor antagonist, SYM 2206 caused a slight depression of evoked monosynaptic inhibitory postsynaptic currents (IPSCs) in control, but a substantial decrease in kindled dentate granule cells. The observation that kainate receptor activation altered paired-pulse depression and reduced the frequency of TTX-insensitive miniature IPSCs without affecting their amplitude is consistent with a presynaptic action on the inhibitory terminal to reduce GABA release. In kindled preparations, neither glutamate (100 micro m) nor kainate (10 micro m) applied in a concentration known to depolarize hippocampal interneurons led to an increase of the TTX-sensitive spontaneous IPSC frequency nor to changes of the postsynaptic membrane properties. Consistently, the inhibitory effect on evoked IPSCs was not affected by the presence of the GABAB receptor antagonist, CGP55845A, thus excluding a depression by an enhanced release of GABA acting on presynaptic GABAB receptors. The enhanced inhibition of GABA release following presynaptic kainate receptor activation favours a use-dependent hyperexcitability in the epileptic dentate gyrus.