Abstracts

Rationale:
Seizures arise not only as a consequence of abnormal activity at the single neuron level, but also by the abnormal propagation of activity through brain networks. Network pathology in epilepsy can be quantified at several levels of network organization. Large-scale measures of brain connectivity (“macro networks”) reveal distinct patterns of anatomical and functional connectivity in patients with refractory epilepsy. Anatomical macro networks are primarily measured using diffusion-weighted MRI or structural covariance MRI analysis to characterize structural changes associated with epilepsy syndromes. A somewhat broader field of literature describes functional macro networks, in patients with epilepsy, by inferring connectivity from correlated activity measured with electrical recordings or resting-state functional MRI. At the cellular level, the dysfunction of small-scale networks (“micro networks,” interactions between neurons) underlies the pathological macro networks observed at the whole brain level. Here, we quantify changes in functional connectivity of micro networks during both ictogenesis and epileptogenesis.

Methods:
In this study, we imaged neuronal activity in an ex vivo model of post-traumatic epileptogenesis: the organotypic hippocampal slice culture. Slices were prepared from P7 mice that were transduced intracerebroventricularly on P0 with a soma-targeted version of the genetically encoded calcium indicator GCaMP8m. We then used a novel imaging system constructed inside of a tissue culture incubator, to image slices continuously beginning shortly after the injury of slicing and continuing through the onset of spontaneous recurrent seizures (after ~seven days in vitro). Every four hours, a movie of calcium dynamics with cellular resolution and a field of view spanning the entire epileptic network was acquired. “Resting-state” functional network connectivity was quantified by identifying neurons that fired together at a rate above chance during non-epileptiform activity.

Results:
On the time scale of ictogenesis, we found that edge density and degree increased in the minutes leading up to seizure onset, indicating an increase network synchrony. On the time scale of epileptogenesis, we also observed a resting-state edge density increase of >100% (p< 0.01) during the first three days in vitro, when spontaneous seizures first emerged. Quantitative analysis is ongoing, but preliminary analysis indicates that incubating slices in chondroitinase ABC, which is anticipated to enhance inhibition, decreases functional network connectivity in epileptic slices.

Conclusions:
These results are consistent with increased functional connectivity in cellular neural networks associated with both short-term changes leading up to seizure onset in epileptic networks and long-term changes leading up to the first appearance of spontaneous seizures during early epileptogenesis. Uncovering the mechanisms that underlie these variations in functional cellular network connectivity will be critical to identifying drug targets to disrupt the pathological activity they produce.

Funding:
NIH R01NS112538, 5R35NS116852