We have used the rat hippocampal slice preparation as a model system for studying the epileptogenic consequences of a reduction in neuronal Na+-K+ pump activity. The cardiac glycosides (CGs) strophanthidin and dihydroouabain were used to inhibit the pump. These drugs had readily reversible effects, provided they were not applied for longer than 15-20 min. Hippocampal CA1 pyramidal cells were studied with intracellular recordings; population spike responses and changes in extracellular potassium concentration ([K+]o) were also measured in some experiments. This investigation focused on the possibility that intrinsic neuronal properties are affected by Na+-K+ pump inhibitors. The CGs altered the CA1 population response evoked by an orthodromic stimulus from a single spike to an epileptiform burst. Measurements of [K+]o showed that doses of CGs sufficient to cause bursting were associated with only minor (less than 1 mM) changes in resting [K+]o. However, the rate of K+ clearance from the extracellular space was moderately slowed, confirming that a decrease in pump activity had occurred. Intracellular recording indicated that CG application resulted in a small depolarization and apparent increase in resting input resistance of CA1 neurons. Although CGs caused a decrease in fast gamma-aminobutyric acid mediated inhibitory postsynaptic potentials (IPSPs), CGs could also enhance the latter part of the epileptiform burst induced by picrotoxin, an antagonist of these IPSPs. Since intrinsic Ca2+ conductances comprise a significant part of the burst, this suggested the possibility that Na+-K+ pump inhibitors affected an intrinsic neuronal conductance. CGs decreased the threshold for activation of Ca2+ spikes (recorded in TTX and TEA) without enhancing the spikes themselves, indicating that a voltage-dependent subthreshold conductance might be involved. The action of CGs on Ca2+ spike threshold could not be mimicked by increasing [K+]o up to 10 mM. A variety of K+ conductance antagonists, including TEA, 4-AP, Ba2+ (in zero Ca2+), and carbachol were ineffective in preventing the CG-induced threshold shift of the Ca2+ spike. The shift was also seen in the presence of a choline-substituted low Na+ saline. Enhancement of a slow inward Ca2+ current is a possible mechanism for the decrease in Ca2+ spike threshold; however, it is impossible to use the Ca2+ spike as an assay when testing the effects of blocking Ca2+ conductances. Therefore, we studied the influence of CGs on the membrane current-voltage (I-V) curve, since persistent voltage-dependent conductances appear as nonlinearities in the I-V plot obtained under current clamp.(ABSTRACT TRUNCATED AT 400 WORDS)
