EPILEPTIFORM ACTIVITY INDUCED BY LOW CHLORIDE MEDIUM IN THE CA1 SUBFIELD OF THE HIPPOCAMPAL SLICE

1. Extracellular and intracellular recordings and measurements of the extracellular concentration of free K+ ([K+]o) were performed in the CA1 subfield of the rat hippocampal slice during perfusion with artificial cerebrospinal fluid (ACSF) in which NaCl had been replaced with equimolar Na-isethionate or Na-methylsulfate (hereafter called low Cl- ASCF). 2. CA1 pyramidal cells perfused with low Cl- ACSF generated intracellular epileptiform potentials in response to orthodromic, single-shock stimuli delivered in stratum (S.) radiatum. Low-intensity stimuli evoked a short-lasting epileptiform burst (SB) of action potentials that lasted 40-150 ms and was followed by a prolonged hyperpolarization. When the stimulus strength was increased, a long-lasting epileptiform burst (LB) appeared; it had a duration of 4-15 s and consisted of an early discharge of action potentials similar to the SB, followed by a prolonged, large-amplitude depolarizing plateau. The refractory period of the LB was longer than 20 s. SB and LB were also seen after stimulation of the alveus. 3. Variations of the membrane potential with injection of steady, DC current modified the shape and SB and LB. When microelectrodes filled with the lidocaine derivative QX-314 were used, the amplitudes of both SB and LB increased in a linear fashion during changes of the baseline membrane potential in the hyperpolarizing direction. The membrane input resistance, as measured by injecting brief square pulses of hyperpolarizing current, decreased by 65-80% during the long-lasting depolarizing plateau of LB. 4. A synchronous field potential and a transient increase in [K+]o accompanied the epileptiform responses. The extracellular counterpart of the SB was a burst of three to six population spikes and a small increase in [K+]o (less-than-or-equal-to 2 mM from a resting value of approximately 2.5 mM). The LB was associated with a large-amplitude, biphasic, negative field potential and a large increase in [K+]o (up to 12.4 mM above the resting value). Changes in [K+]o during the LB were largest at the border between S. oriens and S. pyramidale. This was also the site where the field potentials measured 2-5 s after the stimulus attained their maximal amplitude. Conversely, field potentials associated with the early component of the LB or with the SB displayed a maximal amplitude in the S. radiatum. 5. Spontaneous SBs and LBs were at times recorded in the CA1 and in the CA3 subfield. After a cut that separated these two areas, spontaneous discharges disappeared in the CA1 but not in the CA3 subfield, suggesting that they originated presumably in the latter hippocampal region. Sb and LB could, however, be elicited in the isolated CA1 subfield by electrical stimuli. 6. The LB displayed features that differed from those of spreading depression seen during perfusion with low Cl- ACSF. In addition to having a longer duration, spreading depression was accompanied by increases in K+ > 40 mM (usually 50-90 mM) and propagated with a velocity of 1-2 X 10(-4) m/s. On the contrary, LBs traveled along the CA1 subfield of the hippocampal slice at a velocity of 0.016-0.16 m/s. 7. The late component of the LB was blocked by adding bicuculline methiodide (BMI, 2-20-mu-M) or picrotoxin (100-mu-M) to the low Cl- ACSF. These gamma-aminobutyric acid (GABA)A receptor antagonists also decreased the amplitude and the duration of the initial component of the LB as well as of the SB. High-strength stimuli could, however, elicit short-lasting epileptiform responses in the presence of BMI or picrotoxin. 8. The late-depolarizing component of the stimulus-induced LB disappeared during perfusion with low Cl- ACSF containing the N-methyl-D-asparate (NMDA) receptor antagonists 3-3(2-carboxypiperazine-4-yl)propyl-1-phosphonate (CPP, 1-5-mu-M) or DL-2-amino-5-phosphonovalerate (APV, 10-50-mu-M). However, it reappeared in this type of pharmacologic condition when the stimulus intensity was increased. The SB or the early component of the LB were not affected by these NMDA receptor antagonists. 9. These findings demonstrate that pyramidal cells in the CA1 subfield of hippocampal slices perfused with low Cl- ACSF generate two types of synchronous epileptiform activity after electrical stimuli. Both the laminar analysis of the extracellular field potentials and the pharmacologic data indicate that the SB represents a mixture of excitatory postsynaptic potentials (EPSP) and early inhibitory postsynaptic potentials (IPSP) that is inverted under these conditions. Conversely, the LB is presumably caused by the efflux of Cl- ions through channels coupled to GABA(A) receptors.