Neuronal synchrony and the transition to spontaneous seizures.

The role of inhibitory neuronal activity in the transition to seizure is unclear. On the one hand, seizures are associated with excessive neuronal activity that can spread across the brain, suggesting run-away excitation. On the other hand, recent in vitro studies suggest substantial activity of inhibitory interneurons prior to the onset of evoked seizure-like activity. Yet, little is known about the behavior of interneurons before and during spontaneous seizures in chronic temporal lobe epilepsy. Here, we examined the relationship between the on-going local field potential (LFP) and the activity of populations of hippocampal neurons during the transition to spontaneous seizures in the pilocarpine rat model of epilepsy. Pilocarpine treated rats that exhibited spontaneous seizures were implanted with drivable tetrodes including an LFP electrode and recordings were obtained from the CA3 region. For each recorded seizure, identified single units were classified into putative interneurons or pyramidal cells based on average firing rate, autocorrelation activity and waveform morphology. The onset of sustained ictal spiking, a consistent seizure event that occurred within seconds after the clinically defined seizure onset time, was used to align data from each seizure to a common reference point. Ictal spiking, in this paper, refers to spiking activity in the low-pass filtered LFP during seizures and not the neuronal action potentials. Results show that beginning minutes before the onset of sustained ictal spiking in the local field, subpopulations of putative interneurons displayed a sequence of synchronous behaviors. This includes progressive synchrony with local field oscillations at theta, gamma, and finally ictal spiking frequencies, and an increased firing rate seconds before the onset of ictal spiking. Conversely, putative pyramidal cells did not exhibit increased synchrony or firing rate until after ictal spiking had begun. Our data suggest that the transition to spontaneous seizure in this network is not mediated by increasing excitatory activity, but by distinct changes in the dynamical state of putative interneurons. While these states are not unique for seizure onset, they suggest a series of state transitions that continuously increase the likelihood of a seizure. These data help to interpret the link between in vitro studies demonstrating interneuron activation at the transition to seizure, and human studies demonstrating heterogeneous neuronal firing at this time.