1. Physiological responses of hippocampal pyramidal neurons in primary culture to prolonged glutamate (GLU) exposure (500-mu-M in all experiments) were studied with the use of patch electrodes and whole-cell current-clamp recording techniques. In some experiments, perforated patch recordings were employed with electrodes containing the pore-forming antibiotic nystatin. 2. After washout of GLU after a 10-min exposure, pyramidal neurons remained depolarized by greater-than-or-equal-to 20 mV from rest for the duration of the recording (30 min to <4 h). This depolarization was accompanied by a 57.8% increase in membrane conductance and was termed an extended neuronal depolarization (END). The percentage of neurons in which END was induced varied with the duration of GLU exposure, with a 4-, 6-, 8-, 10-, and 20-min GLU exposure eliciting END in 12.5, 41.7, 81.8, 100, and 100% of neurons. EN D induction appeared to be an all-or-none phenomenon, because END levels did not differ when compared across GLU exposure times. 3. During the END, cells retained both the ability to fire action potentials and the ability to respond to GLU, appeared viable when examined anatomically, and still excluded vital dyes. This supports the conclusion that END is not a nonspecific consequence of cell death. Rather, END is a discrete physiological process triggered by prolonged GLU exposure. The results raise questions concerning the reversibility of END induction, i.e., can neurons be "rescued" once END is induced, or will these cells inevitably go on to die? 4. END induction was dependent on a rise in intracellular free calcium ([Ca]i). END was prevented by strong buffering of [Ca]i or by substitution of external Ba2+ for Ca2+. However, substitution of Mn2+ for Ca2+ still permitted END induction. In cells recorded with the perforated-patch technique, maintaining normal [Ca]i levels, END could be induced, but less readily than under unbuffered [Ca]i conditions. 5. END could not be induced by a 10- to 20-min current-clamp depolarization to 0 mV, nor by 10-min GLU application while the membrane potential was voltage clamped at rest in a solution containing 1 mM Mg2+. In addition, END induction by GLU could be blocked by application of MK-801 (10-30-mu-M) but not 6-cyano-7-nitroquinoxaline-2,3-dione [CNQX (100-200-mu-M)]. 6. The dependence of both delayed neuronal cell death and END induction on GLU exposure duration were similar. This correlation was derived by examining the physiology and toxicity of similar populations of neurons (medium to large pyramidal cells). This suggests that END induction may be a physiological triggering mechanism, responsible for subsequent delayed cell death. 7. These data are consistent with the hypothesis that END induction by GLU is dependent on depolarization and NMDA receptor activation, which raises [Ca]i in a [Ca]o-dependent manner, involving, at least in part, Ca2+ permeation through the N-methyl-D-aspartate (NMDA) channel. This rise in [Ca]i initiates intraneuronal calcium-dependent processes responsible for END induction. The prolonged END may allow more calcium entry through voltage- and ligand-dependent calcium channels, which may augment the toxicity initiated by GLU exposure. Future studies focusing on reversal of END once initiated may prove valuable in extending the therapeutic window for treatment of conditions related to excitotoxicity.
