Abstracts

Rationale: Dravet syndrome (DS) is a severe pediatric epileptic encephalopathy predominantly linked to heterozygous loss-of-function mutations in SCN1A (encoding the voltage-gated sodium channel [VGSC] α subunit, Nav1.1). A subset of DS patients has homozygous loss-of-function mutations in SCN1B (encoding VGSC β1/β1B subunits) which functions in modulation of sodium current (INa) and cell adhesion. Scn1b-/- mice model DS, with 100% lethality by postnatal day 21 (P21). We previously showed that these mice have both neuronal excitability and development defects. Here we compared constitutive and inducible Scn1b-/- mouse models to test the impact of altered brain development vs. impaired modulation of pyramidal neuron INa in the mechanism of seizures and death in Scn1b DS. Methods: We used P14-20 Scn1b-/- mice with wildtype controls or P14-19 (juvenile) and P42-117 (adult) tamoxifen (TMX)-inducible, SLICK-H (Thy1)-CreERT2 mice crossed with Scn1bFlox/Flox mice, and Cre-negative controls. SLICK-H mice express YFP upon Cre recombinase activation. Seizures were monitored by video EEG. From acute brain slices we measured action potentials (APs), spontaneous excitatory post synaptic currents (sEPSCS), somatic INa using nucleated patch clamp, and isolated axonal INa by subtracting somatic from whole cell INa in dissociated neurons. We used a biophysical computational single neuron and network models incorporating measured INa. Results: Immunohistochemistry of TMX-treated SLICK-H mice showed YFP expressed in pyramidal neurons of cortical layers (L) 3, 5, and 6, but not in GABA+ neurons (n=3). Both juvenile (n=5) and adult (n=12) Scn1bFlox/Flox–SLICK-H mice exhibited seizures and 100% lethality by 20 days post TMX treatment. These results in adult mice, which developed normally prior to gene deletion, suggest altered brain development does not play a role in seizure generation and death in Scn1b-linked DS. Both adult Scn1bFlox/Flox–SLICK-H mice and juvenile Scn1b-/- mice show AP defects (Table 1) but L3 and 5 were unaffected. Scn1b-/- L6 nucleated patches showed reductions in INa density and inactivation recovery rate while axonal INa activation was depolarized and recovery from inactivation was accelerated compared to WT (Table 2). A L6 pyramidal neuron model with these parameters predicted a phase response curve transition from type 1 (desynchronizing) to type 2 (synchronizing) firing patterns. Excitatory network modeling with single point neurons revealed spontaneous network bursting with Scn1b-/- sodium channel properties. sEPSC recordings in L6 in Scn1b-/- mice had increased network bursts in 5/7 cells vs 1/8 in WT (p < 0.05) with unchanged total frequency. Conclusions: We show that, while pyramidal neuron hyperexcitability is observed in Scn1b-/- mice and may result from altered brain development, it is likely not the origin of seizures and death in Scn1b-linked DS. Instead, we propose that cell type and subcellular domain specific INa defects favor synchronization even in the presence of hypoexcitability. Since L6 is involved in cortico-thalamic feedback, our results suggest a potential origin of generalized seizures in DS. This work may guide development of DS therapies targeting synchronization in thalamocortical networks. Funding: NIH R37NS076752 to LLI and a Michigan Brain Initiative Predoctoral Fellowship to JMH