Abstracts

Rationale: People with epilepsy are at risk of sudden unexpected death in epilepsy (SUDEP). SUDEP is poorly understood, but it is thought a convulsive seizure leads to impaired awareness and respiratory demise. SUDEP typically occurs at night with victims found prone in bed. Impaired serotonin (5-HT) physiology has been implicated in SUDEP. Brain 5-HT neurons regulate seizure severity, respiratory physiology, and CO₂ chemosensitivity, are required for awakening to rising CO₂ (CO₂ arousal), and in diurnal humans and nocturnal rodents, 5-HT levels are lowest at night. Maximal electroshock (MES) seizures in C57BL6/J mice and spontaneous seizures in the Scn1a^R1407X/+ model of Dravet Syndrome (DS mice) are more fatal at night. The nighttime tendency of MES induced death is regulated by a circadian rhythm independent of sleep. Preliminary work from our lab suggests the timing of spontaneous seizure death is also under circadian control. We hypothesized that the timing of seizure-associated death is influenced by time-of-day variability in 5-HT. To address this, we employed three seizure models (DS mice, MES, and amygdala kindling) and two methods to eliminate brain 5-HT (genetic and chemical elimination).

Methods: Postnatal day (PND) 18 – 21 DS mice and 3 – 5-month-old Lmx1b^f/f/p (KO) and wild-type (WT) littermates were used. 5-HT KO mice lack >99% of brain 5-HT neurons. Singly housed DS mice were continuously monitored with infrared cameras. DS mice received daily injections of the tryptophan hydroxylase inhibitor para-chlorophenylalanine (800 mg/kg, PCPA) or volume-matched saline. KO mice and WT littermates were subjected to one MES seizure at different times of day. A group of WT mice was instrumented for amygdala kindling to examine the role of time-of-day seizures on CO₂ arousal.


Results: Saline-treated DS mice were more likely to die following spontaneous seizures during the night (8/10), compared to PCPA-treated mice, which died more during the day (7/10). Fatal seizures always progressed into full hindlimb extension (FHE). Non-fatal seizures with FHE were associated with prolonged recovery compared to seizures without FHE (p < 0.0001, two-way ANOVA with multiple comparisons). In the saline-treated group, recovery time was time-of-day-dependent (N = 10, MESOR = 142.8 ± 72.26 sec) with a nighttime zenith approaching baseline recovery in PCPA-treated animals (N = 4, MESOR 220.5 ± 138.1 sec). WT mice were more likely to die during the night (p = 0.0479, Barnard’s exact test). However, KO mice had high mortality regardless of time of day (p > 0.05, Barnard’s exact test). Daytime seizures in kindled mice (N = 3, 6 replicates/mouse) impaired CO₂ arousal (2.11-fold increase in arousal latency; p < 0.01, two-tailed t-test). Nighttime seizures profoundly impaired CO₂ arousal (14.81-fold increase; p < 0.05).


Conclusions: SUDEP occurs more frequently during the night and victims are often found in the prone position. Seizure-associated death in nocturnal mouse models is also more common at night. Nighttime seizures may exacerbate 5-HT-associated arousal physiology and increase risk of death.


Funding: R01NS129722, R01NS129722-01S1, Joanna Sophia Grant (CURE Epilepsy), Beth L. Tross Epilepsy Professorship.