Abstract
Sudden unexpected death in epilepsy (SUDEP) is the most extreme consequence of epilepsy. SUDEP typically occurs at night. Because humans sleep at night, these nighttime deaths are often attributed to seizures arising from sleep. Nocturnal mice also experience more seizure-associated deaths during the nighttime. This could represent timing that is under circadian control. To examine this, male and female Scn1a(R1407X/+) mice, a model of the epileptic encephalopathy Dravet syndrome, in which patients experience spontaneous seizures that often result in death, were housed in constant darkness and the timing of seizure associated death was assessed. We found that the timing of sudden death following seizures persists in constant darkness and peaks during the subjective nighttime. This circadian rhythm of death was independent of the timing of potentially fatal seizures and more frequently occurred while awake. Potentially fatal seizures resulted in prolonged unconsciousness, which also exhibited a circadian rhythm peaking during the subjective night. These findings provide support for circadian regulation, independent of seizure timing and sleep, in the nighttime risk of seizure-associated death. Nighttime seizures may increase risk of SUDEP via multiple mechanisms, as evident by peak spontaneous sudden death and profoundly impaired consciousness following seizures during the subjective night. KEY POINTS: Sudden unexpected death in epilepsy, or SUDEP, is a devastating outcome of intractable epilepsy. Converging lines of evidence indicate that there is a time-of-day preference for SUDEP, with more SUDEP occurring during the night. Several animal models of the epileptic encephalopathy Dravet syndrome (DS), including the one employed in our study, recapitulate key features of DS in patients, including a high rate of seizure-related death and more of the deaths occurring at night. Here, we removed light/dark photocycles, by housing animals in constant darkness, and identify nighttime preponderance of death, suggesting that this is under circadian regulation. We further carefully characterize fatal vs. non-fatal seizures in our animals and identify features that may prove to be useful biomarkers to predict which seizures may become fatal.