Abstracts

Rationale: The voltage-gated potassium channel Kv2.1, encoded by KCNB1, is primarily responsible for delayed rectifier potassium currents, which are important regulators of neuronal excitability. Previous work identified de novo KCNB1 mutations in individuals with epileptic encephalopathy type 26 (EIEE26). One of the novel mutations identified, G379R, located within the pore selectivity domain, resulted in an epilepsy phenotype which included infantile spasms, atonic, tonic-clonic and focal seizures. EEG monitoring in infancy revealed hypsarrhythmia and childhood monitoring showed multifocal diffuse polyspikes, polyspike-wave, right temporal spike and wave, left occipital spikes and diffuse polyspike bursts. In order to further investigate the G379R mutation we generated a mouse model using CRISPR/Cas9 genome engineering and performed initial characterization of spontaneous seizures, electroclinical features, induced seizure thresholds and neurobehavioral phenotypes.  Methods: Mice were generated and maintained on the C57BL/6J background. Kcnb1R/+, Kcnb1R/R and wild-type (WT) littermate controls of both sexes were evaluated. Beginning from postnatal day 18 (P18) through P84, Kcnb1R/+, Kcnb1R/R and WT littermates were continuously video monitored to evaluate spontaneous seizure activity and home cage behavior. Selected behaviors (continuous circling, grooming, jumping) were quantified offline by an observer blinded to genotype. A separate cohort of Kcnb1R/+, Kcnb1R/R and WT littermates underwent EEG implantation between P18 and P42 and video-EEG was collected for at least two weeks per subject. Epileptiform activity was scored manually by an observer blinded to genotype. Thresholds to a minimal clonic seizure (6 hz) or generalized tonic-clonic seizure (flurothyl) were determined in P18-84 mice. Neurobehavioral testing consisted of a neurological exam, open field, zero maze, cliff avoidance and marble burying performed on separate days with Kcnb1R/+, Kcnb1R/R and WT littermates at P70 to P105. Prior to each assay, mice were acclimated to the testing facility for at least an hour prior to testing. Males and females were tested separately with ≥1 hour delay between sexes.  Results: Kcnb1R/+ and Kcnb1R/R mice were viable and fertile, although a small percentage (~2%) experienced sporadic death starting as early as P21. Spontaneous seizures, characterized by rearing and paddling followed by wild running, were often observed in Kcnb1R/R mice upon placement in a novel environment. Video-EEG monitoring of Kcnb1R/+ and Kcnb1R/R mice revealed spike and wave discharges, polyspikes, slow disorganized baseline activity and abnormal sleep architecture. Relative to WT littermates, both Kcnb1R/+ and Kcnb1R/R mice exhibited significantly reduced thresholds to minimal clonic seizures in the 6 hz assay and GTC seizures induced by flurothyl (p< 0.0006, student's t test). Continuous home-cage video monitoring revealed abnormal repetitive behaviors in Kcnb1R/+ and Kcnb1R/R mice, including excessive rearing and/or paddling, excessive grooming and circling. In addition, Kcnb1R/R mice exhibited extended bouts of vertical jumping. In neurobehavioral assays, compared to WT littermates, Kcnb1R/+ and Kcnb1R/R mice exhibited hyperactivity, increased impulsivity and inattention to novel objects. Conclusions: The mouse models Kcnb1R/+ and Kcnb1R/R recapitulate many neurologic and neurobehavioral features observed in individuals with KCNB1-associated encephalopathy, including spontaneous seizures, abnormal EEG, repetitive behaviors and hyperactivity/impulsivity. This new mouse model will be useful for better understanding the underlying pathophysiology and evaluating potential treatments. Funding: NS053792