1. We made intracellular and extracellular field potential recordings and ion-selective measurements of extracellular Ca2+ concentration ([Ca2+](o)) and extracellular K+ concentration ([K+](o)) in human neocortical slices that were obtained in the course of epilepsy surgery. Slices were maintained in vitro at 34-35 degrees C and were perfused with Mg2+-free artificial cerebrospinal fluid (ACSF). 2. Spontaneous field potential epileptiform discharges (duration = 2.5-80 s) occurred in most of the slices studied (similar to 60%) after 1.5-2 h of perfusion with Mg2+-free ACSF. Intracellular recordings from regular-spiking neocortical neurons showed that epileptiform events consisted of large-amplitude( 15-30 mV) de polarizing shifts that were capped by bursts of fast action potentials. A decrease in [Ca2+](o) (change in [Ca2+](o) = 0.02-0.17 mM, 0.07 +/- 0.046 mM, mean +/- SD, from a baseline of 1.8 mM, n = 10 slices) and an increase in [K+](o) (change in [K+](o) = 0.5-3.8 mM, 1.6 +/- 1.24 mM, from a baseline of 3.25 mM, n = 10) were associated with each epileptiform discharge. 3. The epileptiform activity induced by Mg2+-free ACSF uas abolished by bath application of antagonists of the: N-methyl-D-aspnrtate (NMDA) receptor. This procedure also blocked the appearance of spreading depression-like episodes. By contrast, the rate of occurrence of epileptiform discharges was not significantly modified by antagonizing non-NMDA receptors. 4. We also observed spontaneous, rhythmic potentials of positive polarity during perfusion of Mg2+-free ACSF; the potentials became hyperpolarizing when the neuron membrane was made less negative than -75 mV with intracellular injection of depolarizing current, and they were decreased or abolished during application of the gamma-aminobutyric acid-A (GABA(A)) receptor antagonist bicuculline methiodide (BMI). The rate of occurrence and/or the amplitude of these presumably GABA(A)-mediated events decreased similar to 2 s before the onset of each epileptiform discharge. 5. Application of BMI prolonged the epileptiform discharges while decreasing their rate of occurrence. These changes were also accompanied by an increase in the amplitude of the epileptiform field potential DC shift, whereas the concomitant decreases in [Ca2+](o) and increases in [K+](o) became more pronounced than in control Mg2+-free medium (31.2% and 42.8%, respectively, n = 10 slices). 6. Intracellular analysis of regular-spiking neurons in slices that did not generate spontaneous epileptiform discharges after >2 h of perfusion with Mg2+-free ACSF showed all-or-none, variable latency epileptiform bursts that were induced by high-strength focal extracellular stimuli. Low-frequency (0.5-5 Hz) extracellular focal stimuli could also elicit epileptiform afterdischarges and longlasting potentiation of postsynaptic responses. 7. In these slices repetitive trains of focal stimuli caused spontaneous epileptiform discharges to occur. This phenomenon was accompanied by a decrease of the GABA(A)-mediated spontaneous rhythmic intracellular potentials whose rate of occurrence between one discharge and the next remained lower than under pretrain conditions (42-65%, n = 3). 8. These findings indicate that inhibitory potentials caused by the activation of the GABAA receptor are operant in the human neocortex maintained in vitro at the lime when Mg2+-free epileptiform discharges are recorded. A transient decrease in such GABA(A)-mediated mechanism might be involved in the onset of epileptiform discharges and might play a role in controlling the changes in [Ca2+](o) and [K+](o) that accompany each epileptiform event. 9. We also propose that in slices that did not generate spontaneous epileptiform activity, a decrease of GABA(A)-mediated inhibition can be induced by trains of focal stimuli. Such a change might lead to the appearance of spontaneous epileptiform activity.
