Kindling induces transient NMDA receptor-mediated facilitation of high-frequency input in the rat dentate gyrus

To elucidate the gating mechanism of the epileptic dentate gyrus on seizure -like input, we investigated dentate gyrus field potentials and granule cel l excitatory postsynaptic potentials (EPSPs) following high-frequency stimu lation (10-100 Hz) of the lateral perforant path in an experimental model o f temporal lobe epilepsy (i.e., kindled rats). Although control slices show ed steady EPSP depression at frequencies greater than 20 Hz, slices taken f rom animals 48 h after the last seizure presented pronounced EPSP facilitat ion at 50 and 100 Hz, followed by steady depression. However, 28 days after kindling, the EPSP facilitation was no longer detectable. Using the specif ic N-methyl-D-aspartate (NMDA) and RS-alpha -amino-3-hydroxy-5-methyl-4-iso xazoleproponic acid (AMPA) receptor antagonists 2-amino-5-phosphonovaleric acid and SYM 2206, we examined the time course of alterations in glutamate receptor-dependent synaptic currents that parallel transient EPSP facilitat ion. Forty-eight hours after kindling, the fractional AMPA and NMDA recepto r- mediated excitatory postsynaptic current (EPSC) components shifted drama tically in favor of the NMDA receptor- mediated response. Four weeks after kindling, however, AMPA and NMDA receptor- mediated EPSCs reverted to contr ol-like values. Although the granule cells of the dentate gyrus contain mRN A-encoding kainate receptors, neither single nor repetitive perforant path stimuli evoked kainate receptor- mediated EPSCs in control or in kindled ra ts. The enhanced excitability of the kindled dentate gyrus 48 h after the l ast seizure, as well as the breakdown of its gating function, appear to res ult from transiently enhanced NMDA receptor activation that provides signif icantly slower EPSC kinetics than those observed in control slices and in s lices from kindled animals with a 28-day seizure-free interval. Therefore, NMDA receptors seem to play a critical role in the acute throughput of seiz ure activity and in the induction of the kindled state but not in the persi stence of enhanced seizure susceptibility.