Low magnesium epileptogenesis in the rat hippocampal slice: electrophysiological and pharmacological features.

Extra- and intracellular recording techniques were used to study the epileptiform activity generated by rat hippocampal slices perfused with Mg2(+)-free artificial cerebrospinal fluid (ACSF). This procedure induced in both CA1 and CA3 subfields the appearance of synchronous, spontaneously occurring epileptiform discharges which consisted of extracellularly recorded 100-800 ms long, positive shifts with superimposed negative going population spikes. Simultaneous, extracellular recordings from CA1 and CA3 subfields revealed that the epileptiform discharges in CA3 preceded those occurring in CA1 by 5-25 ms. Surgical separation of the two areas led to the disappearance of spontaneous events in the CA1 but not in the CA3 subfield. In this type of experiment CA1 pyramidal cells still generated epileptiform discharges following orthodromic stimuli. The intracellular counterpart of both spontaneous and stimulus-induced epileptiform discharges in CA1 and CA3 pyramidal cells was a large amplitude depolarization with high frequency discharge of action potentials which closely resembled the paroxysmal depolarizing shift recorded in the experimental epileptogenic focus. A hyperpolarizing potential triggered by alvear stimuli was recorded in CA1 cells perfused with Mg2(+)-free ACSF. This hyperpolarization was blocked by bicuculline methiodide (BMI) indicating that it represented a GABAergic inhibitory postsynaptic potential (IPSP). BMI also caused a prolongation of both spontaneous and stimulus-induced Mg(+)-free epileptiform discharges. Perfusion of the slices with the N-methyl-D-aspartate (NMDA) receptor antagonist DL-2-amino-5-phosphono-valerate (APV) reduced and eventually abolished the Mg(+)-free epileptiform discharges. These effects were more pronounced in the CA1 than in the CA3 subfield. APV also reduced the amplitude and the duration of the alveus-induced IPSP. These data demonstrate that Mg(+)-free epileptiform activity is present in the hippocampal slice at a time when inhibitory GABAergic potentials are operant as well as that in the CA1 subfield this type of epileptiform activity is dependent upon NMDA-activated conductances. Our experiments also indicate that NMDA receptors might be involved in the neuronal circuit responsible for the hyperpolarizing IPSP generated by CA1 pyramidal neurons.