Abstracts

Rationale: The high fat, low carbohydrate/protein ketogenic diet (KD) is an effective anti-seizure therapy for a broad spectrum of pediatric and adult epilepsies; although, due to the stringency of the diet, it is primarily used in pediatric patients refractory to current anti-seizure medications. The mechanism of KD anti-seizure efficacy is unclear, but it is known that the KD engages anti-inflammatory and anti-oxidant pathways and promotes mitochondrial health. Many of these effects mirror the downstream effects of the nutritionally-regulated transcription factor peroxisome proliferator activated receptor gamma, PPARγ. Here, we tested the hypothesis that PPARγ contributes to the anti-seizure efficacy of the KD. Methods: We treated wild-type (WT) and epileptic Kv1.1 knockout (KO) mice with either standard chow or KD with drinking water containing either DMSO vehicle or the PPARγ antagonist GW9662 for two weeks after weaning. Expression of nuclear PPARγ protein in brain homogenates was probed with western blots. Spontaneous seizures in Kv1.1KO mice was monitored with 24hr video-EEG. Acute seizures were evoked in control, PPARγ2KO and neuronal-specific PPARγKO (NKO) mice with flurothyl vapors. Results: We found that the KD differentially affected PPARγ isoform expression. The KD increased nuclear PPARγ2 in both WT and Kv1.1KO mice. In WT mice, PPARγ1 was predominant over PPARγ2 (γ2:γ1 ratio ~0.6), whereas in WT+KD mice, γ2 doubled and the two isoforms were expressed equally. Results in the epileptic brains were significantly different from WT: in Kv1.1KO mice, PPARγ2 was predominant and the γ2:γ1 ratio was three-fold greater than WT, which further increased to six fold in Kv1.1KO+KD mice. Co-administration of a PPARγ antagonist, GW9662, prevented KD-mediated changes in nuclear PPARγ2:γ1 ratios in both WT and Kv1.1KO mice and prevented the anti-seizure efficacy of the KD as measured with video-EEG recordings. We did not detect any seizures in WT mice given GW9662 and GW9662 did not worsen Kv1.1KO seizures. We found that the KD prolongs the latency to flurothyl-induced seizure in WT mice. Therefore, we further tested PPARγ involvement in the KD with PPARγ2KO and NKO mice. KD-mediated increases in flurothyl seizure latencies were lost in PPARγ2KO mice and NKO mice. Further, administration of a PPARγ agonist, pioglitazone (PIO), to Kv1.1KO mice mimicked the KD, inducing a similar increase in γ2:γ1 ratio and reducing seizures. In vivo PIO administration resulted in reduced Kv1.1KO hippocampal hyperexcitability as measured in vitro and improved function of Kv1.1KO hippocampal mitochondria. Moreover, WT mice given PIO exhibited increased flurothyl seizure latencies. Neither genetic nor pharmacologic manipulation of PPARγ affected the KD-induced changes in weight or blood glucose or ketone body concentrations. Conclusions: Collectively, we provide pharmacologic and genetic evidence that brain PPARγ is involved in the anti-seizure efficacy afforded by the KD in chronic epilepsy and acute seizures. Our results strongly support the investigation of PPARγ as a therapeutic target for severe, refractory epilepsy.