Abstracts

Rationale: The ketogenic diet (KD) is a proven treatment for medically refractory epilepsy, and results in prominent production of ketone bodies (KB) such as D-β-hydroxybutyrate (BHB), acetoacetate (ACA) and acetone. While the mechanisms of KD action remain unclear, it has long been hypothesized that KB might mediate the clinical effects of the KD. And although several studies have shown acute anticonvulsant properties of ACA and acetone, it remains unclear whether BHB has similar activity. Recent studies have shown that the KD can inhibit mTOR signaling and prevent hippocampal atrophy in an experimental model of multiple sclerosis (PMID: 21371020, 22567104). Here, we asked whether BHB is anticonvulsant in spontaneously epileptic Kcna1-null (KO) mice, and whether this action is associated with alterations in mTOR signaling and structural changes in hippocampus. Methods: KO mice (P31-37) were implanted s.c. with Alzet osmotic mini-pumps containing 10 mM BHB dissolved in saline. Seizure frequency was assessed with continuous video-EEG monitoring over 3-day periods beginning 1, 7 and 12 days after implantation. After 15 days of BHB treatment, hippocampal structural changes were studied with manganese-enhanced imaging (MnCl₂, 0.2 mM/kg, 1 day after i.p. injection) on a 7T MRI scanner. Hippocampal lysates were collected for Western blots from four groups of mice : saline-treated wild-type (WT); saline-treated KO; BHB-treated WT; and BHB-treated KO mice. Samples (n=8 per group) were probed with antibodies against upstream modulators of mTOR (e.g., Akt and AMPK) and downstream effectors of mTOR (e.g., S6 and 4EBP1). Results: Saline-treated KO expressed a mean daily tonic-clonic seizure frequency of 5.8 ± 0.6 at 12-15 days, and this effect was significantly reduced by BHB treatment (p<0.01). A similar change was seen in KO mice exposed to BHB for 7 days, but not after 1 day. Saline-treated KO mice showed a significant increase in phospho-S6 expression (31% increase compared to saline-treated WT), whereas BHB exposure led to a reduction in phospho-S6 in KO mice (20% decrease as compared to saline-treated KO). No differences in Akt, AMPK and 4EBP1 were seen amongst control and experimental groups. In saline-treated KO mice, MR imaging revealed increased signal intensities in dentate gyrus (DG) and CA3 hippocampus, but only slight increases in CA1. No significant differences were seen in DG between BHB-treated KO and saline-treated WT mice, but BHB treatment led to partial normalization of the MR signal in the CA3 subfield of KO animals. Conclusions: Chronic in vivo administration of BHB results in a significant anticonvulsant effect in Kcna1-null mice. Moreover, BHB alone appears to exert a structural neuroprotective effect in hippocampus, and inhibits aberrant mTOR signaling in these epileptic animals. The combination of morphological and functional neuroprotective effects of BHB in this genetic model of epilepsy suggests that KB may be beneficial for neurological conditions other than epilepsy.