Abstracts

Rationale: Propagation of epileptic activity in the hippocampal slice seems to follow a sequential order in the hippocampal circuit. However, intra-regional and inter-layer activity propagation has been poorly characterized. To this aim, we conducted electrophysiological recordings from hippocampal slices with a high-density microelectrode matrix. We especially address how bursts of high frequency and voltage activity that propagate along the CA, in an apparent wave-like fashion, are related to their respective current density in high time and spatial resolution.Methods: We recorded 4-AP-induced epileptic activity from rat hippocampal slices (n=6) on a high-density array of 64X64 microelectrodes (BioCam). To determine the dynamics of activation of the CA region, we conducted a center-of-mass analysis (vector averaging of active regions) from the field activity and from the current source density derived from it, computed in sub-millisecond time frames. Because computing the center of mass can give poor results when used on concave geometries, we implemented a rotational quasi-invariant discrete Laplacian Operator. With it, we achieve a strict separation of the CSD data into two distinct sets: that of all sinks and that of all sources. Once they were obtained, we applied a disjoint component analysis that separates the areas in which groups of several components behave as a coordinated sink or source. Then, for each component a “center of mass” was computed by using the integrated CSD as weight. Re-joining each time frame, the succession of the centers of mass results on several disjoint trajectories, which have a precise moment of onset and offset with high spatial resolution. Thus, this indicates the loci, dynamics and relative strength of the locally averaged succession of activities.Results: With our method, we show that the initiation site of the epileptic activity was in CA3a-b and, form there, it propagated to CA1; importantely, with time, it also back-propagated to the dentate gyrus. However, once initiated, the epileptic activity does not travel along a single defined wave front, but it splits into disjoint components that run simultaneously along the different layers of CA. This suggests a sort of ""distributed signal analysis"" of hippocampal segments that might be unmasked by our analysis and that can be a characteristic of epileptic activity itself. Also, we observed reverberating hyperactivity in the site of initiation before it propagated to the rest of the hippocampus, which could also back-propagate to the DG.Conclusions: Our method provides a means to detect localized disjoint components that propagate in wavelike fashion. The better spatial resolution that provides the CSDA combined with a clear distinction of disjoint sources and sinks permits us to follow each of these coordinated groups. This paves the way for understanding how information is locally integrated and how it is conveyed along the hippocampal circuit. This investigation was funded by Consejo Nacional de Ciencia y Tecnología, México.