Abstracts

Rationale: Sleep cycles and architecture is bidirectionally affected in epilepsy (Sedigh-Sarvestani et al. 2014). Improved understanding of the interaction between sleep regulation and seizures provides opportunities for better seizure control and intervention. Sleep-wake states are predominantly characterized from EEG and EMG, and state transitions have been described as discrete mechanisms. Studies of these underlying mechanisms often invoke discrete co-inhibitory dynamics analogous to electrical flip-flops. But observations of discrete transitions are derived from experiments with head-fixed or lightly anesthetized animals, or short recordings in sensory deprived novel environments. These conditions are abnormal and inherently dysregulate sleep-wake patterns. In order to better understand the underlying shared circuitry between sleep-wake transitions and seizures, we need to be cognizant of the conditions of the experiment and their artifact-like effects on the result of the investigation. Methods: To study the relationship between sleep-wake transitions and seizure onset in normal conditions, we have obtained long-term continuous cortical and hippocampal local field potential measurements along with head acceleration, in freely behaving normal and epileptic rats. We categorized three main SOVs of REM, NREM, and Wake based on EEG rhythms, and head acceleration (Sunderam 2007): REM is characterized by a sharp spectral peak in the theta band of hippocampal activity and an absence of accelerometer activity except for brief muscle twitches. NREM is characterized by high amplitude, slow oscillations resulting in high delta band power in cortical activity and an absence of accelerometer activity. Wake is characterized by either low- or high-power accelerometer activity indicating small or large head-movement respectively. Results: We found that transitions between these states are not discrete. In normal animals, we characterized a significantly long intermediate state between NREM and REM with no accelerometer activity and no clear spectral peak at delta or theta frequencies. The existence of this state has been cited previously (Gottesmann 1996). However it seemed to be neglected in more recent sleep studies. Interestingly, epileptic rodents appeared to have discrete transitions between NREM and REM without any intermediate state. Conclusions: Our findings, in freely behaving animals, suggest that state transitions are governed by fluid interactions across multiple mechanisms rather than discrete on-or-off switches. This questions the conventional discrete definition of sleep-wake states used to investigate sleep-wake regulation in normal or diseased brain. Specifically in our epileptic animals, the intermediate state seemed to be eliminated. Funding: R01EB019804 04