Abstracts

Rationale: The ketogenic diet (KD) has been successfully used in the treatment of medically refractory epilepsy for decades, yet the underlying mechanisms remain elusive. An unresolved hypothesis is that ketone bodies (KBs; notably, ?-hydroxybutyrate [BHB] and acetoacetate [ACA]) may play a role. Recently, we demonstrated that KBs prevent mitochondrial permeability transition (mPT) in acutely isolated neocortical mitochondria and preserve synaptic plasticity in CA1 hippocampus of normal rodents. Here, we asked whether BHB or NIM811 (a selective cyclophilin D [CypD] inhibitor and a subunit of the mPT complex) alone can confer anticonvulsant activity and protection against intrinsic impairment of long-term potentiation (LTP) in spontaneously epileptic Kcna1-null mice.Methods: Kcna1-null mice (P31-37), generated using heterozygous breeding pairs, were implanted with Alzet osmotic mini-pumps SC containing either 10 mM BHB (N=3; effective blood concentration ~5 mM) in water or saline (N=4). Seizure frequency was assessed over a 3-day period after 2 weeks of treatment with video-EEG recordings. Long-term potentiation (LTP) an electrophysiological measure of memory consolidation was evoked by high-frequency stimulation (HFS, 100 Hz, 1 sec) in CA1 hippocampal slices from separately treated cohorts (N=6, each group). To assess thresholds for mPT, and to independently control for confounding effects of seizure activity, mitochondria were acutely isolated from CypD-null or wild-type (WT) mice (P25-30), loaded in a spectrofluorophotometer, and then subjected to increasing concentrations of Ca²⁺ in the presence or absence of KBs (BHB or ACA, each 1 mM) to provoke mPT opening.Results: Chronic BHB administration alone (superimposed on ad lib rodent chow) resulted in anticonvulsant effects, and also resulted in restoration of hippocampal LTP in in Kcna1-null mice (p<0.05). Intriguingly, short-term pre-incubation with either 5 M NIM 811 or 5 mM BHB (without in vivo administration) also resulted in reversal of LTP impairment in Kcna1-null animals. In mitochondria from WT mice, both BHB and ACA were effective in raising mPT threshold. However, in CypD-null mice, neither KB nor ACA were able to alter calcium-induced mPT; a similar lack of an effect was seen with cyclosporin A (1 mM), a known antagonist at the CypD site.Conclusions: Collectively, our data strongly suggest that the functional neuroprotective effects of KBs (and by extension, possibly the KD as well) are mediated by direct inhibition of the CypD subunit of the mPT pore. Inhibition of mPT represents a novel strategy for neuroprotection, and there is increasing evidence that modulation of this critical mitochondrial multimeric complex may produce broad protective effects against insults such as neurotrauma, cognitive dysfunction, and now epilepsy.