Authors :
Anastasia Brodovskaya, PhD – University of Virginia; Huayu Sun, PhD – Neurology – University of Virginia; Nadia Adotevi, PhD – Neurology – University of Virginia; Ian Wenker, PhD – Anesthesiology – University of Virginia; Isabella Tucker, BS – Biology – University of Virginia; Keri Mitchell, BS – Chemistry – University of Virginia; Rachel Clements, PhD Candidate – Neuroscience – University of Virginia; Jaideep Kapur, MD, PhD – Neurology – University of Virginia
Rationale: Repeated generalized tonic-clonic seizures (GTCS) increase SUDEP risk. We tested whether seizure-induced plasticity of excitatory transmission plays a role in seizure-related death.
Methods: We induced 10 GTCS (PTZ, 50 mg/kg) every other day and simultaneously recorded EEG, breathing, and heartbeat. We performed patch-clamp recordings on seizure-activated neurons and tested a global knockout (KO) of AMPA receptor subunit GluA1. We compared EEG, breathing, and heartbeat in GluA1 KO and WT littermates during repeated seizures and confirmed our findings in the hippocampal kindling model. We then labeled activated neurons using activity reporter TRAP2 mice in the motor cortex and brainstem following apnea.
Results: C57BL/6 mice, during initial seizures, had brief primary apneas at the seizure onset (2.68
± 0.14 s; n=17). With repeated seizures, some mice developed post-ictal secondary apneas (19.29 ± 5.04 s) with recovery. Mice then had secondary apnea without recovery and gradual bradycardia until death (SUDEP-like event). Seizure latency decreased with repetition, from 5.35 to 2.9 min, and mice that died (33%) had longer seizures than survivors (24.73 ± 1.89s vs 19.90 ± 0.50 s). The gradual development of secondary apneas, faster seizure onset, and longer seizure duration indicated seizure-induced plasticity leading to death. To determine the cellular basis of this plasticity, we did patch-clamp recordings from motor cortical layer 2/3 in a TetTag mouse that utilizes c-Fos activity to tag activated neurons with eGFP. Seizure-activated neurons were more excitable than non-activated neurons (8 tagged, 5 untagged pyramidal neurons, n=6 mice) and had larger AMPA receptor-mediated sEPSCs (p=0.0336, 13 tagged, 16 untagged, n=12 mice), demonstrating enhanced AMPA transmission after a single seizure. We used an AMPA receptor- GluA1 subunit global knockout to determine the effect of repeated GTCS on apnea and death. 88.89% of GluA1 KO survived compared to 55.56% of WT littermates (died: 1 out of 8 mice for KO, 8 out of 19 for WT). Surviving KO mice never developed stage 6 seizures and had no secondary apneas compared to the WT mice. GluA1 KO had shorter seizures that remained relatively of the same duration (slope 0.16, 15.13 ± 0.44 s), whereas seizures of WT mice gradually became longer (slope 0.44, 20.21 ± 0.71s, p< 0.001). Hippocampal kindling confirmed that GluA1 KO did not die and had less severe seizures (n=10 each). Even when GluA1 KO mice achieved a stage 5 behavioral seizure score, they could not maintain it compared to WTs. We next used activity reporter TRAP2 mice to visualize activated neurons in the motor cortex and brainstem nuclei after one seizure versus secondary apnea caused by repeated GTCS. We found increased neuronal activation in the specific regions (quantification ongoing).