Abstracts

Rationale: To develop novel anti-seizure treatments for the population of patients with medically refractory epilepsy, we will need to generate new insights into the mechanisms of seizure initiation. Ictogenesis is difficult to study due to the rare and unpredictable nature of seizure onset. Furthermore, many electrophysiological measures have limited spatial resolution. To characterize neuronal patterns of activity during ictogenesis, we used optical methods to follow seizure activity over multiple time scales at cellular resolution, in both in vitro and in vivo models of epilepsy.


Methods: We employed both the genetically encoded calcium indicator GCaMP, and the genetically encoded voltage indicator VARNAM2, to sample the activity of hundreds of neurons during spontaneous seizures. Longitudinal imaging was performed in the in vitro organotypic hippocampal slice culture model of post-traumatic epilepsy. In vivo seizures were imaged in a model of acute focal injury and in the intrahippocampal kainate model of temporal lobe epilepsy.


Results: Ictogenesis was characterized by changes in neuronal calcium, the spiking activity of neurons, and sustained subthreshold depolarizations. Neuronal sequences of activation (as defined by the spiking activity) were highly variable and seizures were found to initiate from different neurons seizure-to-seizure. While ictogenesis was stochastic at the level of individual neurons, seizures did tend to begin from the same hippocampal subregion. Furthermore, we found that ictogenesis began with stereotyped patterns of spatially clustered subthreshold membrane depolarizations.


Conclusions: Our early findings indicate that the depolarized regions are the most consistent feature of seizure onset. This in turn raises the possibility that electrographic recordings do not capture the “prime mover” of ictogenesis.


Funding: NIH NINDS