Abstracts

Rationale:
Epileptic seizures are often caused by the hyper-synchronous activation of neurons in affected circuits. Prior work in absence seizures (Cacng2stg) mice has revealed lower activity and pairwise synchrony in most primary visual cortex (V1) neurons during spike and wave discharges (SWDs) starting several seconds before their appearance on intracranial electroencephalogram (iEEG) recordings. However, how neuronal networks are entrained in SWDs and other forms of epileptiform activity is not clear.
Method:
We use two-photon (2P) imaging of layer II/III retrosplenial cortex (RSC) neurons genetically expressing the neuronal td-Tomato marker in specific neuronal types and/or the neuronal activity indicator GCaMP combined with iEEG recordings in awake mice free to run on a Styrofoam wheel attached to a motion encoder at different stages of epileptogenesis induced by kainic acid (KA) injections in the dorsal hippocampus.
Results:
Paroxysmal limbic epileptiform activity induced by unilateral KA injection entrained both the ipsi- and the contra-lateral RSC. Layer II/III excitatory RSC neurons showed a biphasic entrainment constituted by a slow build-up and a transient drop of activity which preceded a larger transient corresponding to the invasion of the local RSC network. Parvalbumin- (PV) as well as somatostatin-expressing (SST) layer II/III interneurons in the RSC were also entrained by the KA-induced paroxysmal limbic epileptiform activity. Interestingly, the recruitment of layer II/III RSC neurons and interneurons lagged the appearance of the paroxysmal limbic epileptiform activity on iEEG recordings by several tens of seconds, whereas the entrainment of RSC neurons appeared to end in a more time-locked fashion with the paroxysmal limbic epileptiform activity on iEEG recordings. Finally, clusters of epileptiform activity in the form of high frequency oscillations (HFOs) in the fast ripples range, i.e. 250-550 Hz, in chronically epileptic animals, 8 weeks post-KA intra-hippocampal injections suppressed the activity of layer II/III excitatory RSC neurons.
Conclusion:
This work uses combined multi-cellular calcium imaging, iEEG, and genetically expressed activity and neuronal type indicators to demonstrate the impact that different forms of limbic epileptiform activity has on RSC neurons. This finding suggests that neocortical neurons can be modulated by limbic epileptiform activity differently: paroxysmal limbic epileptiform activity hyper-activate RSC neurons, whereas clusters of HFOs in the fast ripples range suppress them. In future studies, optogenetic experiments with single neuron-specific control will be used to understand how changes in the connectivity of RSC neurons during epileptogenesis might lead to the different modes of entrainment of RSC neurons by limbic epileptiform activity.
Funding:
:This work was supported by BLR&D Merit Award I01BX005221 from the VA administration.