Abstracts

Rationale: Cognitive dysfunction is commonly observed in patients with medically refractory temporal lobe epilepsy, especially those who exhibit structural abnormalities on neuroimaging studies. It remains unclear whether such deficits result from the pathologic substrate, recurrent seizure activity, chronic anticonvulsant therapy, or a combination of these. Clinical reports have recently suggested that the ketogenic diet (KD) can enhance cognitive function in epileptic patients. To provide experimental evidence for this observation, we asked whether the KD can exert functional neuroprotective effects in epileptic Kcna1-null (KO) mice (lacking the voltage-gated Shaker-like potassium channel Kv1.1 alpha-subunit). Methods: Long-term potentiation (LTP) was used as an in vitro measure of synaptic plasticity and functional integrity in CA1 hippocampus. LTP was evoked by stimulation of Schaffer collaterals using high frequency stimulation (HFS, 100 Hz, 1 sec) and recorded in CA1 stratum radiatum. We also used the Barnes Maze behavioral paradigm to test spatial learning in our KO mice. Latencies to solve the maze were measured for each animal for 4 daily trials over 5 consecutive days. In both sets of experiments, we used animals (P31-P39) treated with the KD for 10-14 days. Results: HFS evoked robust LTP in CA1 pyramidal neurons in the hippocampus of wild-type (WT) mice fed a standard diet (SD). LTP induction was significantly impaired in KO mice fed the same SD; LTP induction measured 140 ± 8.0 % (WT) and 100 ± 5.5 % (KO) of baseline excitatory postsynaptic potential amplitudes (EPSPs) at 60 min post-HFS. KO animals on the KD showed a restoration of LTP induction, resembling that of the WT mice (143 ± 8.3% of baseline at the same time-point after HFS). In the Barnes maze test, both SD-fed WT and SD-treated KO mice exhibited spatial learning capabilities by improving their average time latency over the 5 day period. By day 5, KO mice solved the maze in a mean time of 69 sec, compared to 37 sec of the WT mice. On a KD, KO mice solved the maze in 46 sec on average. Conclusions: Using both in vitro and in vivo assessments of hippocampal learning and memory, we found that KD treatment can help restore impaired LTP in CA1 hippocampus and can preserve spatial learning of epileptic Kcna1-null mice. We conclude that the KD has functional, neuroprotective effects against the intrinsic impairment of the hippocampus in epileptic mice thus supporting previous suggestions that KD treatment can enhance cognitive function in epileptic patients.