Search

All-or-none electrographic seizures (EGSs) were studied in hippocampal slices from young (21- to 38-day-old) rats in medium containing low (0 mM) or physiological (0.9 mM) levels of magnesium, with and without the GABAB agonist baclofen. Extracellular recording and stimulation were performed in stratum pyramidale and stratum radiatum of CA3, respectively. EGS activity was induced by exposure to low-Mg medium or by delivering repetitive stimulus trains in physiological Mg medium. After EGS activity had stabilized, the EGSs were tested for all-or-none behavior by varying the number of pulses in a train. An EGS was considered all-or-none if subthreshold stimulation produced no afterdischarge bursts, and if the EGS duration was largely independent of the number of suprathreshold stimulus pulses. According to this measure, EGSs in Mg-free + baclofen medium were all-or-none. EGSs evoked in physiological Mg medium were also all-or-none, although the threshold was higher, and the EGS duration lower, than in Mg-free medium. This all-or-none characteristic was observed whether the EGSs were induced by prior exposure to Mg-free medium or by repetitive stimulation, and in the presence and absence of baclofen. The all-or-none characteristic suggests that while the triggering mechanism for EGSs is strongly dependent on stimulus intensity, regenerative mechanisms — independent of stimulus intensity — are responsible for the maintenance of EGSs. EGSs are also terminated by mechanisms not dependent on stimulus intensity.

In this paper we present an in vitro model of epileptogenesis based on electrical stimulation rather than pharmacological or ionic manipulations. Hippocampal slices given a series of stimulus trains similar to those used in kindling exhibited 3 types of epileptiform activity in CA3: afterdischarges immediately following the trains; spontaneous bursts of multiple population spikes; and bursts triggered by single stimuli. The afterdischarges and spontaneous bursts may be comparable to those seen in vivo during kindling; also, the progression of these features in this model was similar to their progression during kindling. All epileptiform activities were long-lasting, persisting for up to 3.5 h following the last train. This stimulus train-induced population bursting should be valuable as an acute model of hippocampal epileptogenesis, and may also help elucidate hippocampal participation in the kindling process.