Abstracts

Rationale: Combining anti-seizure drugs with strict dietary adherence is often a successful strategy for achieving seizure control. However, little is understood regarding why certain diets provide superior seizure management. Moreover, some diets even appear to aggravate some seizures, notably those associated with childhood absence epilepsy. The ketogenic diet induces hypoglycemia and subsequent ketosis that promotes the use of ketone bodies as the brain's primary energy source. Herein, we investigate how acute hypoglycemia alters spike-and-wave discharge occurrence in rodent models of absence epilepsy.  Methods: Experiments were conducted on either WAG/Rij rats (W) or DBA/2J mice (D), two rodent species with electrographic spike-and-wave discharges (SWD). We evaluated SWD occurrence in response to acute hypoglycemia and insulin injection using combined video-EEG recordings. Blood glucose levels were measured with a human glucometer. Ketone body levels were measured using a beta-hydroxybutyrate colorimetric assay (Caymen Chemical). To evaluate specific contributions of thalamic circuits to SWD occurrence in WAG/Rij rats, we delivered drugs directly to the thalamus using cannulae while recording EEG. To gain a mechanistic understanding of glucose sensitivity in the thalamus, electrophysiological recordings in acute brains slices were performed. Whole-cell patch-clamp recordings were used to measure responses of single thalamic neurons to low glucose conditions, while extracellular field recordings were used to measure network-level activity changes induced by low glucose. Finally, immunoblotting was used to evaluate changes in AMPK and p-AMPK protein expression during hypoglycemia and pharmacological conditions. Results: SWD count increased after a 16-hour fast (D: p<0.05, n=12; W: p<0.05, n=13), a duration that also significantly decreased blood glucose levels (D: p<0.05, n=12; W: p<0.001, n=13) and increased ketone body levels (D: p<0.05, n=12; W: p<0.001, n=13)[MB1] . To disambiguate the combined effects of hypoglycemia and hyperketonemia on SWDs, we monitored seizures following insulin injection, a manipulation that results only in the former. Insulin injection increased SWD occurrence (D: p<0.05, n=11; W: p<0.05, n=9), suggesting that low glucose, not elevated ketone bodies, modulates SWDs. To support this hypothesis, we inhibited glycolysis specifically within the thalamus using 2-deoxyglucocse (2-DG), a manipulation that increased SWD count (W: p<0.05, n=6).  Phosphorylated AMPK (p-AMPK) levels rise during periods of cellular stress, including low energy states. Interestingly, p-AMPK has been shown to augment currents evoked by GABAB receptors, important molecular players in pathological thalamocortical oscillations. We therefore tested the hypothesis that increased thalamic p-AMPK levels following a fast amplifies GABAB receptor currents to lower the threshold for SWD generation. In support of this hypothesis, we show that p-AMPK levels in the thalamus are elevated following a fast, and that p-AMPK activators enhance GABAB-mediated currents in TC neurons (p < 0.05; n = 18W cells).     Conclusions: We demonstrate that glucose modulates thalamocortical excitability and SWDs, an effect that appears to involve elevated p-AMPK levels.  Funding: 5RO1NS099586 NIGMS T32 Gm007055