Abstracts

Rationale: Rodent hippocampal CA1 region contains at least 20 different types of inhibitory cells (Somogyi and Klasuberger 2005). However, the role of distinct inhibitory cell subtypes in the dynamic process of seizures is unclear. We have previously shown a unique pattern of spiking interplay between pyramidal cells and the dendritically-innervating oriens-lacunosum moleculare (OLM) interneurons in vitro in 4-Aminopyridine (4-AP) induced seizure-like events (SLE; Ziburkus et al. 2006). The goal of the current study is to investigate the mechanisms of synchronization between the perisomatically-innervating parvalbumin (PV) and cholecystokinin (CCK) positive basket cells (BC), pyramidal cells (PC), and OLM cells during SLEs. Methods: We performed dual and triple whole cell and extracellular recordings in transverse slice preparations of the juvenile rat (P21-30) hippocampus CA1 region. Seizures were induced using bath application of an A-type potassium channel blocker 4-AP (200µM). During the recording, patched cells were filled with neurobiotin. Post hoc, slices were processed for triple immunofluorescence and biocytin stain retrieval to validate the identity of the patched cell subtype. The cells were then morphologically reconstructed using Neurolucida software (MicroBrightField, Inc). Action potential (spike) and overall signal crosscorrelation analysis was carried out to measure synchronization dynamics in these cell subtypes. Results: Dual whole cell recordings were made from putative BC-PC and BC-OLM pairs (n=33). A total of 122 SLEs were recorded from these pairs. Triple whole cell recordings of BC-OLM-PC triplets (n=2) yielded a total of 5 SLEs. To date, spike frequency analysis of 67 seizures in putative BC-PC pairs showed no difference in spike rate before and after the ictal event. However, BC spiking lagged behind PC firing and peaked towards the end of the ictal event. Our results suggest the following dynamical evolution of SLEs: OLM cells increase spiking first, followed by a relatively balanced spiking activity in the PC and BC cells during the initiation of the SLEs. During the ictal event (body of the SLEs) OLM spiking decreased or ceased during prolonged depolarization blocks, BC spike rates increased and PCs reached their peak firing rate. At the end of the ictal event PC firing decreases, whereas BC and OLM firing persists for a few more seconds. Crosscorrelation analysis showed an elevated correlation between PC and BC spikes at the beginning and the end of the ictal event, and decreased correlation during the ictal event. Conclusions: Overall signal synchronization patterns observed in this study are similar to our previous work (Ziburkus et al 2006), yet the pattern of spiking between the two interneuron subtypes with respect to PCs is distinct. OLM cells show increased spiking at the beginning of the SLE followed by spike failure, whereas BCs show increased spiking during and towards the end of the event. These findings suggest that impairments of both inhibitory feed-forward and feed-back microcircuits may accompany the pattern formation dynamics of seizures.