Abstracts

Rationale: Post-traumatic epilepsy(PTE) is one of the most common causes of acquired epilepsy. Following traumatic brain injury (TBI), a sequence of neurological changes occur that lead to appearance of spontaneous seizures; however, the detailed mechanism of epileptogenesis is still unknown. One of the features of an epileptic network is excessive axon sprouting, but its causality in the development of epilepsy is not clearly understood. Methods: We used organotypic hippocampal cultures(OHCs) as an in-vitro model of PTE. We dissected whole hippocampal cultures into three sub-regions (CA1, CA3, and DG). The dissected parts were then oriented and cultured together. We were thereby able to make three groups of cultures: whole hippocampal cultures; lesioned cultures, where a gap was left between a cut edge at CA3b-c; and 'healed' cultures, where the sub-regions were allowed to fuse back together after lesion. The spontaneous activity of the cultures was recorded optically through jRGECO1a (red fluorescent genetically encoded [Ca2+] indicator) at different days in-vitro (DIV) from 8 DIV to 25 DIV. Each recording was 30 minutes long and ΔF/F values were assessed after baseline correction. Two ROIs were selected, one near the cut edge of lesioned cultures(CA3c) and a second away from the cut edge at CA3b. All ROIs from different OHCs had the same size and shape. Corresponding regions were selected in control OHCs. Activity was extracted by a simple thresholding method on the F/F traces. Sum of the values above the threshold from each ROI was calculated as well as the ratio from the two ROIs in the same culture. The ratios were then compared among three different experimental groups. The subiculum of whole hippocampal cultures was also compared through the same analysis against the organotypic cultures that contained both cortex and hippocampus. Results: Analysis of the optical recordings showed that in cut cultures, the area more proximal to the cut displays statistically significantly higher levels of paroxysmal calcium activity (p < 10E-6, paired t-test). This represents a reversal of the normally observed behavior in intact cultures, where ROI2 consistently exhibits higher levels of activity (P < 10E-5, paired t-test). Conclusions: Our results from the lesioned cultures suggest that the immediate location close to the cut edge becomes more excitable and sustains higher frequency of activity during paroxysmal events. In order to explain the observed behavior we implemented a progressive simulation to model axon sprouting and formation of new connections in time. We considered 900 nodes representing neurons. Each neuron had a maximum capacity for a total value of pre- and post-synaptic weights based on a normal distribution with a mean value of 270 (30% of total number of nodes) and SD of 9 to reach equilibrium. The connection probability followed a normal distribution versus Euclidean distance. This algorithm resulted in neurons at the network corners and edges to make more or stronger connections with the immediately neighboring neurons since there are fewer partners available. We hypothesize that this process may explain the hyperexcitability observed at the cut edge of OHCs. Funding: NIH/NINDS R33NS088358