The natural history and treatment of epilepsy in a murine model of tuberous sclerosis

Abstract

Purpose: Patients with tuberous sclerosis complex (TSC) often have severe epilepsy that is intractable to available therapies. The development of novel treatments for epilepsy in TSC would benefit greatly from a suitable animal model, but most animal models of TSC to date have few reported neurological abnormalities, such as epilepsy. We previously described a novel model of TSC, due to conditional inactivation of the Tsc1 gene in glia (Tsc1(GFAP)CKO mice), in which mice develop epilepsy and premature death. Here, we characterize the natural history of the epilepsy in Tsc1(GFAP)CKO mice in more detail and report acute effects of treatment with standard antiepileptic drugs on seizures in these mice.

Methods: Video-EEG recordings were obtained from Tsc1(GFAP)CKO mice on a weekly basis, starting at 4 weeks of age until death, to monitor progression of interictal EEG abnormalities and seizures. In separate experiments, Tsc1(GFAP)CKO mice were monitored for interictal EEG abnormalities and seizures before and during treatment with phenobarbital, phenytoin, or saline.

Results: Tsc1(GFAP)CKO mice developed seizures around 4-6 weeks of age and subsequently had progressive worsening of the interictal EEG background and seizure frequency over a month, culminating in death. Treatment with phenobarbital or phenytoin caused a reduction in seizure frequency, but did not improve EEG background or prevent death.

Conclusions: Tsc1(GFAP)CKO mice develop progressive epilepsy. Acute treatment with standard antiepileptic drugs suppresses seizures in these mice, but does not affect long-term prognosis. Tsc1(GFAP)CKO mice represent a good model to test other drugs that may have true antiepileptogenic actions in TSC.

Figure 1

The interictal EEG activity of Tsc1^GFAPCKO mice becomes progressively abnormal with age. (A) Representative examples of the 4 interictal EEG background grades, as described in the Methods, are shown: Grade 1 – normal background activity (+/− 6-8 Hz sinusoidal theta rhythm), no epileptiform spikes; 2 – mostly normal background activity, few epileptiform spikes; 3 – mostly abnormal background activity, many spikes; 4 – burst-suppression pattern. (B) Average interictal EEG background grade of Tsc1^GFAPCKO mice shows progressive worsening with age (p<0.01 by ANOVA; *post-test comparisons show weeks 8 and 9 to be significantly greater than week 4; n = 6-9 mice per time point). (C) Average interictal spike frequency of Tsc1^GFAPCKO mice also increased with age (p<0.001 by ANOVA; *post test comparisons show weeks 7-9 to be significantly greater than weeks 4 and 6; n = 6-9 mice per time point). (D) Fast Fourier Transform of interictal EEG background of Tsc1^GFAPCKO mice reveals different frequency distributions dependent on age. Average power spectra for 4 week and 9 week old mice are shown. Theta (4-8 Hz) frequencies predominated in 4 week old mice and slower delta (1-4 Hz) frequencies were more common in older 9 week old mice (For clarity, data for 5-8 week old mice are not shown). (E) The ratio of theta/delta power decreased with age (p<0.001 by ANOVA; *post test comparisons show weeks 8 and 9 to be significantly less than weeks 4-6; n = 6-9 mice per time point).

Figure 2

The progression of epilepsy in Tsc1^GFAPCKO mice. (A) A representative example of ictal EEG activity recorded during a typical seizure. Seizures usually started with bilateral fast tonic activity that evolved into higher amplitude spikes. The end of the seizure typically evolved into bursting spike discharges, followed by postictal suppression. (B) Average seizure frequency of Tsc1^GFAPCKO mice increased with age (p <0.05 by ANOVA; *post test comparisons show week 9 to be significantly greater than weeks 4-6; n = 6-9 mice per time point). (C) Average seizure duration did not change with age. Data for weeks 4-6 were combined due to the small number of mice with seizures at these ages. Seizure duration appeared to be higher at 9 weeks, but this difference was not statistically significant (p = 0.08 by ANOVA; n = 6-9 mice per time point).

Figure 3

Acute treatment with phenobarbital or phenytoin reduces seizure frequency, but does not affect seizure duration or interictal EEG background. (A) There was no significant difference in seizure frequency between pre-drug and drug periods in saline-injected Tsc1^GFAPCKO mice, but both phenobarbital and phenytoin significantly reduced seizure frequency in Tsc1^GFAPCKO mice (*p<0.01 for phenobarbital, p<0.05 for phenytoin by paired t-tests; n=10-11 mice per group). (B-D) Phenobarbital and phenytoin had no significant effect on seizure duration, interictal spike frequency or interictal EEG background grade (p>0.05 for all comparisons; n=10-11 mice per group). Pre – pre-drug monitoring period, Sal- saline, PB – phenobarbital, PHT – phenytoin.