Dynamic coupling among neocortical neurons during evoked and spontaneous spike-wave seizure activity.

1. We investigated the development from patterns of electroencephalogram (EEG) synchronization to paroxysms consisting of spike-wave (SW) complexes at 2-4 Hz or to seizures at higher frequencies (7-15 Hz). We used multisite, simultaneous EEG, extracellular, and intracellular recordings from various neocortical areas and thalamic nuclei of anesthetized cats. 2. The seizures were observed in 25% of experimental animals, all maintained under ketamine and xylazine anesthesia, and were either induced by thalamocortical volleys and photic stimulation or occurred spontaneously. Out of unit and field potential recordings within 370 cortical and 65 thalamic sites, paroxysmal events occurred in 70 cortical and 8 thalamic sites (approximately 18% and 12%, respectively), within which a total of 181 neurons (143 extracellular and 38 intracellular) were simultaneously recorded in various combinations of cell groups. 3. Stimulus-elicited and spontaneous SW seizures at 2-4 Hz lasted for 15-35 s and consisted of barrages of action potentials related to the spiky depth-negative (surface-positive) field potentials, followed by neuronal silence during the depth-positive wave component of SW complexes. The duration of inhibitory periods progressively increased during the seizure, at the expense of the phasic excitatory phases. 4. Intracellular recordings showed that, during such paroxysms, cortical neurons displayed a tonic depolarization (approximately 10-20 mV), sculptured by rhythmic hyperpolarizations. 5. In all cases, measures of synchrony demonstrated time lags between discharges of simultaneously recorded cortical neurons, from as short as 3-10 ms up to 50 ms or even longer intervals. Synchrony was assessed by cross-correlograms, by a method termed first-spike-analysis designed to detect dynamic temporal relations between neurons and relying on the detection of the first action potential in a spike train, and by a method termed sequential-field-correlation that analyzed the time course of field potentials simultaneously recorded from different cortical areas. 6. The degree of synchrony progressively increased from preseizure sleep patterns to the early stage of the SW seizure and, further, to its late stage. In some cases the time relation between neurons during the early stages of seizures was inversed during late stages. 7. These data show that, although the common definition of SW seizures, regarded as suddenly generalized and bilaterally synchronous activities, may be valid at the macroscopic EEG level, cortical neurons display time lags between their rhythmic spike trains, progressively increased synchrony, and changes in the temporal relations between their discharges during the paroxysms.(ABSTRACT TRUNCATED AT 400 WORDS)