Abstracts

Rationale: We hypothesized that some epilepsy-associated ion channel variants may act as seizure-protective alleles in the appropriate genetic context. We examined this idea using a new digenic mouse model of human idiopathic epilepsy composed of two epilepsy-associated ion channel mutations with mutually opposing excitability defects and overlapping sub-cellular localization: Kv1.1 (a Kcna1 Shaker-like potassium channel mutation) and tottering (tg; a Cacna1a P/Q-type calcium channel mutation). The Kv1.1 mutation causes sudden death and limbic epilepsy characterized by generalized tonic-clonic (TC) seizures, while the tg mutation produces absence epilepsy associated with spike-wave (SW) seizures. We hypothesized that increasing axon terminal excitability by eliminating Kcna1 channels would improve synaptic transmission at terminals associated with Cacna1a channel dysfunction, and conversely, that impairing transmission by expression of mutant Cacna1a channels might dampen the hyperexcitability associated with Kcna1 defects.Methods: Double mutant mice for experiments were derived from crosses between doubly heterozygous mice (Kv1.1/+; tg/+). We performed EEG recordings using bilateral electrodes implanted into the subdural space over the frontal, temporal, parietal, and occipital cortices. We evoked in vitro epileptiform discharges in hippocampal brain slices with 7.5 mM [K⁺]_o solutions and extracellularly measured the frequency and duration of the resulting discharges in CA3. We measured CA1 fiber volley (FV) facilitation by determining the normalized amplitude of the presynaptic FV (while blocking postsynaptic responses with CNQX and D-APV) in response to stimulation (10 pulses at 33 Hz) of the Schaffer collateral pathway.Results: We found that when the Kv1.1 and tg mutations are combined, they partially mask one another with respect to in vivo seizure activity and in vitro excitability of hippocampal networks and presynapses. The two mutations masked the in vivo EEG seizure phenotypes of one another in Kv1.1/Kv1.1; tg/tg mice, as evidenced by an essentially complete lack of tg-like SW seizures and a 60% decrease in Kv1.1-like TC seizure occurrence. Additionally, tg dramatically rescued the sudden death phenotype of Kv1.1 mice. In the hippocampus, we found that CA3 interictal discharges evoked by high potassium exhibited intermediate burst frequencies in Kv1.1/Kv1.1; tg/tg mice that were partially restored to the wild-type value. When we analyzed the specific contribution of presynaptic excitability and synaptic transmission by measuring CA1 fiber volley facilitation, we found that double mutant brain slices exhibited normal axon fiber recruitment, whereas tg and Kv1.1 brain slices showed significant reductions in facilitation.Conclusions: Our findings demonstrate for the first time that epilepsy-associated ion channel mutations can be seizure protective in the right genomic milieu. These results highlight the importance of comprehensive evaluation of an individual’s channel profile (“channotype”) for improving genetic diagnosis and risk assessment in complex human diseases involving channelopathies, such as epilepsy.