Abstracts

Rationale: KCNT1 epileptic encephalopathies arise due to mutations in an outwardly rectifying, sodium-activated, potassium channel known as Slack, K_Na1.1 or Slo2.2. This channel is widely expressed in the brain and contributes to slow hyperpolarization following repetitive action potentials. KCNT1 mutations are typically characterized as gain-of-function and patients present with a variety of seizure disorders and profound developmental delay. Seizures tend to be refractory to anti-epileptic drugs (AEDs), underscoring the need to identify novel compounds that have inhibitory activity against KCNT1 mutations. Methods: A wild-type or P924L mutant KCNT1 gene were transfected into HEK293 cells to generate models capable of expressing a functional K⁺ current. The cell lines were evaluated with electrophysiology to characterize the effects of the mutation on K⁺-related currents. A rubidium (⁸⁶Rb⁺) efflux assay was then adapted for high-throughput drug screening; ⁸⁶Rb⁺ serves as a charge and size proxy for K⁺ flow. Wild-type and P924L cell lines were screened against a proprietary library of over 1,290 approved drugs as well as a panel of AEDs and known K⁺ channel modulators to identify compounds with inhibitory activity. Results: Using patch-clamp electrophysiology, Slack potassium channels with the P924L mutation showed significantly increased potassium current compared to wild-type Slack channels, confirming a gain-of-function molecular phenotype (Milligan et al, Ann Neurol 75:581-590, 2014). A comprehensive, high-throughput drug repurposing screen was performed to find compounds that inhibited ion flow in the P924L or wild-type KCNT1 cell models. From the drug library, 55 compounds were identified in the wild-type and 40 compounds in the P924L cell model that significantly inhibited efflux. These hit compounds were then evaluated across a series of concentrations in the ⁸⁶Rb⁺ efflux assay and with electrophysiology to confirm activity and to derive an IC₅₀ and maximal effectiveness. Compounds with a high degree of novel inhibitory activity in both cell models included ezetimibe (cholesterol-lowering), raloxifene (hormone modulator) and doxazosin mesylate (antihypertensive). Ritonavir (anti-viral) and carvedilol (beta-blocker) were two drugs with activity only against the P924L variant cell model. Although there has been interest in quinidine as an inhibitor of KCNT1 channels, the drug displayed only a modest level of inhibitory activity that was substantially less than several other compounds identified by the drug repurposing screen. Conclusions: The present study demonstrates the utility of comprehensive screening with a large library of approved drugs in order to identify compounds with significant activity against cellular models of epilepsy-related genes. These results provide an enriched pool of new drug candidates for further development and therapeutic evaluation against gain-of-function KCNT1 epileptic encephalopathies. Funding: This work was sponsored by Pairnomix, LLC (Minneapolis, MN).