KNa1.1 gain-of-function preferentially dampens excitability of murine parvalbumin-positive interneurons

KCNT1 encodes the sodium-activated potassium channel KNa1.1, expressed preferentially in the frontal cortex, hippocampus, cerebellum, and brainstem. Pathogenic missense variants in KCNT1 are associated with intractable epilepsy, namely epilepsy of infancy with migrating focal seizures (EIMFS), and sleep-related hypermotor epilepsy (SHE). In vitro studies of pathogenic KCNT1 variants support predominantly a gain-of-function molecular mechanism, but how these variants behave in a neuron or ultimately drive formation of an epileptogenic circuit is an important and timely question. Using CRISPR/Cas9 gene editing, we introduced a gain-of-function variant into the endogenous mouse Kcnt1 gene. Compared to wild-type (WT) littermates, heterozygous and homozygous knock-in mice displayed greater seizure susceptibility to the chemoconvulsants kainate and pentylenetetrazole (PTZ), but not to flurothyl. Using acute slice electrophysiology in heterozygous and homozygous Kcnt1 knock-in and WT littermates, we demonstrated that CA1 hippocampal pyramidal neurons exhibit greater amplitude of miniature inhibitory postsynaptic currents in mutant mice with no difference in frequency, suggesting greater inhibitory tone associated with the Kcnt1 mutation. To address alterations in GABAergic signaling, we bred Kcnt1 knock-in mice to a parvalbumin-tdTomato reporter line, and found that parvalbumin-expressing (PV+) interneurons failed to fire repetitively with large amplitude current injections and were more prone to depolarization block. These alterations in firing can be recapitulated by direct application of the KNa1.1 channel activator loxapine in WT but are occluded in knock-in littermates, supporting a direct channel gain-of-function mechanism. Taken together, these results suggest that KNa1.1 gain-of-function dampens interneuron excitability to a greater extent than it impacts pyramidal neuron excitability, driving seizure susceptibility in a mouse model of KCNT1-associated epilepsy.