Authors :
Anna Goforth, BS – University of Wisconsin Madison
Peizhen Yang, BS – Cornell University
Danny Lasky, BS – UC Davis
Presenting Author: Rama Maganti, MD – University of Wisconsin - Madison
Mathew Jones, PhD – University of Wisconsin Madison
Rationale: Sleep and epilepsy have a complex bidirectional relationship where seizures or epileptiform activity disrupt sleep and sleep disruptions in turn trigger breakthrough seizures. Prior work showed that sleep homeostasis, which is assessed by delta power in non-rapid eye movement (NREM) sleep and its overnight decline, is disrupted in human epilepsy. Furthermore, there are individual variations in how seizures affect sleep just as there are individual variations in susceptibility to acquired epilepsy following an insult, in humans and mouse strains. Here we investigated differences in susceptibility to acquired epilepsy, sleep homeostatic drive, association of interictal activity with NREM delta power and influences of sleep deprivation (SD) on interictal spikes in Kainic acid-treated epilepsy-resistant (C57) and epilepsy-susceptible (DBA) mouse strains.
Methods: We collected EEG data from 12-week-old C57BL/6 and DBA mice for 8 weeks following a single round of repeated low dose Kainic acid (KA) treatments (C57s n=8; DBAs n=12). Mice underwent SD for 4 hours a day x 3 days in week 4 of recording. Spontaneous seizures were manually scored across 8-weeks. Interictal spikes (IIS) were extracted from 24-hour EEG records of baseline, SD and recovery days using an automated spike detection algorithm (doi:
10.1371/journal.pone.0207158). Sleep was manually scored using Sirenia Sleep in a proportion of mice (DBA n=6; C57s, n=8). We then extracted the mean normalized delta power in every NREM sleep epoch in each hour and IIS from every hour on the baseline, SD and recovery days to assess the correlation between the two.
Results: Spontaneous seizures were seen in 3/8 KA-treated C57s and 9/12 KA-treated DBAs starting in week 2 (Fig. 1A). Mean weekly seizure frequency across weeks 2-8 after KA treatment was 9.4±5.06 (Range: 1-17) in C57s whereas it was 24.7±9.0 (Range: 11-38) in DBAs (p< 0.05). While no differences were seen at baseline, IIS significantly increased on SD days in DBA (ANOVA, F (2.818, 46.78) = 2.941; p=0.04), but not in C57s (p >0.05) (Fig. 1B). SD did not increase seizure frequency in either strain of mice. Sleep homeostatic drive measured by normalized NREM delta power rise and decline across 24-hours, was more attenuated in DBAs. The overall delta power was higher and the day-night oscillation was more robust in C57s. SD resulted in transient increases in normalized NREM delta power in both strains (Fig 2 A-F). Finally, IIS more significantly correlated with increased mean normalized NREM delta power on SD days, in C57s (Fig 2 D-F) (p< 0.001) compared to DBAs (Fig 2 J-L) (p=0.016).