Article
Neocortical layers 5 and 6 are the drivers of network activity in Micro-electrode array recordings of epileptic human brain slice cultures
Die neokortikalen Layer 5 und 6 sind die Treiber von Netzwerk-Aktivität in Micro-electrode Array Recordings in Epileptischen humanen Hirnschnittkulturen
Search Medline for
Authors
Published: | May 25, 2022 |
---|
Outline
Text
Objective: Epilepsy is a condition hallmarked by synchronized depolarizations of groups of neurons resulting in seizures. The underlying conditions can vary extensively ranging from structural pathologies to trauma. However, the electrophysiological and micro-anatomical correlates of seizures are yet to be described. Here, we leverage novel culturing methods for human brain slices and spatio-temporal electrophysiological recordings using micro-electrode arrays (MEA) to investigate features of synchronized network activities to describe epileptic discharges on a micro-network-level.
Methods: Neocortical human brain tissue samples from resective epilepsy surgeries were cultured using human cerebrospinal fluid (hCSF). MEA recordings were performed using a 256-MEA with artificial CSF (aCSF) and hCSF. Data analysis was performed using custom Python scripts including the libraries numpy and scipy. In short, spikes were extracted using a mean absolute deviation threshold-based approach. Network activities were defined as episodes of elevated activity with the mean moving average firing rate (FR) exceeding the mean FR of the whole recording. For each of these network events (n= 6043), we extracted a total of 62 features including interspike-interval distributions, layer-specific mean FR, and channels specific bursting times. Scipy’s Pearsonr-method was used to calculate correlations.
Results: Mean length of network activities was 1.765 seconds with a mean FR of 616.18 Hz. We found a strong correlation between the FR of layers 5 and 6 and the network FR (r = 0.993, p < 0.001). This was not found for any other neocortical layer (layer 1: r = - 0.125, p < 0.001, layer 2 and 3: r = 0.222, p < 0.001, layer 4: r = 0.541, p < 0.001). For layers 5 and 6, this effect was observed for experiments in hCSF and aCSF, which indicates a network-modulating role of these layers. Layers 2 and 3 only showed a strong correlation with the network burst FR for hCSF (r = 0.85, p < 0.001) but not for aCSF (r = 0.063, p < 0.001).
Conclusion: Our finding suggests a distinct role of layers 5 and 6 in the emergence of highly synchronized network spiking activity. Although the differentiation between the electrophysiological equivalent of interictal discharges and physiological network activity remains difficult, our results may help to understand epilepsy as a network disease with spatially distributed layer-specific properties.