Specific deletion of NaV1.1 sodium channels in inhibitory interneurons causes seizures and premature death in a mouse model of Dravet syndrome

Abstract

Heterozygous loss-of-function mutations in the brain sodium channel Na(V)1.1 cause Dravet syndrome (DS), a pharmacoresistant infantile-onset epilepsy syndrome with comorbidities of cognitive impairment and premature death. Previous studies using a mouse model of DS revealed reduced sodium currents and impaired excitability in GABAergic interneurons in the hippocampus, leading to the hypothesis that impaired excitability of GABAergic inhibitory neurons is the cause of epilepsy and premature death in DS. However, other classes of GABAergic interneurons are less impaired, so the direct cause of hyperexcitability, epilepsy, and premature death has remained unresolved. We generated a floxed Scn1a mouse line and used the Cre-Lox method driven by an enhancer from the Dlx1,2 locus for conditional deletion of Scn1a in forebrain GABAergic neurons. Immunocytochemical studies demonstrated selective loss of Na(V)1.1 channels in GABAergic interneurons in cerebral cortex and hippocampus. Mice with this deletion died prematurely following generalized tonic-clonic seizures, and they were equally susceptible to thermal induction of seizures as mice with global deletion of Scn1a. Evidently, loss of Na(V)1.1 channels in forebrain GABAergic neurons is both necessary and sufficient to cause epilepsy and premature death in DS.

Fig. 1.

Floxed Scn1a mouse has reduced Na_V1.1 expression in presence of Cre. (A) Schematic representation of the targeting construct used to generate our floxed Scn1a mouse. Exon 25 is flanked on the 5′ and 3′ ends by LoxP sites in combination with an FRT flanked neomycin selection marker (Top). The neomycin cassette was removed in breeding cross of heterozygous Flox animals with a flippase-expressing animal, leaving behind a single FRT site (Middle). In the presence of Cre recombinase exon 25, the FRT site and one LoxP site are excised leaving behind a single LoxP site (Bottom). (B) Representative genomic DNA gel of PCR products from +/+, F/F, and F/+ tail DNA in the absence of Meox2-Cre and F/+:Meox2-Cre⁺, demonstrating an excision of exon 25 and a corresponding decrease in PCR product size. WT allele, 846 bp; Flox allele, 1019 bp; and excised allele, 258 bp. (C) Western blot for Na_V1.1 protein expression in whole brain membrane proteins (150 μg) from P14 F/+:Meox-Cre⁻, F/+:Meox2-Cre⁺, and F/F:Meox2-Cre⁺, demonstrating a stepwise decline in Na_V1.1 protein levels with the successive removal of exon 25 in the presence of Meox2-Cre. The F/F:Meox2-Cre⁺ sample contains no Na_V1.1 protein and recreates the global knockout expression profile as expected. (D) Following mating a F/F female with a Dlx-Cre⁺ male, a visually detectable decrease in Na_V1.1 immunoreactive staining in deep somatosensory cortex at P14 can be seen. Left, F/F:Dlx-Cre⁻. Immunostaining of neuronal cell bodies for Na_V1.1 channels. Center, F/F:Dlx-Cre⁺. Immunostaining of Na_V1.1 channels in layer V pyramidal cells appears intact, but decreases in immunostaining for Na_V1.1 channels in layer IV and VI cells can be detected. Right, global Na_V1.1 knockout, included as a negative control for antibody specificity. Na_V1.1 protein expression is not observed in an age- and gain-matched global knockout at P14.

Fig. 2.

Dlx1/2-I12b-Cre preferentially decreases Na_V1.1 expression in interneurons. (A–F) Representative immunostaining for Na_V1.1 (green), GABA (red), and co-localized (yellow) in the dentate gyrus from F/F:Dlx-Cre⁻ (A–C) and F/F:Dlx-Cre⁺ (D–F) demonstrate a selective loss of cell body Na_V1.1 staining in F/F:Dlx-Cre⁺ slices in GABA⁺ cells of the hilus (arrowheads) and molecular layer (arrows). (G) Fraction of neurons in F/F:Dlx-Cre⁺ tissue coimmunostained for GABA and Na_V1.1 channels quantified across a range of brain regions relative to GABA and Na_V1.1 coimmunostained cells from F/F:Dlx-Cre⁻ tissue. FC, frontal cortex; MC, motor cortex; SS, somatosensory cortex; VC, visual cortex; s, superficial layers; d, deep layers; DG, dentate gyrus; CA1, hippocampal CA1; and CA3, hippocampal CA3 (*P < 0.05, **P < 0.01, °P = 0.08). (H) Number of excitatory pyramidal neurons immunostained for Na_V1.1 channels but not GABA in the pyramidal layers of SS and VC. (I) Decreased Na_V1.1 immunostaining intensity is observed only in F/F:Dlx-Cre⁺ interneurons also coexpressing GABA (IN) and not in pyramidal cells (Pyr) or dentate granule neurons (DGN) expressing Na_V1.1 alone. (Scale bar, 100 μm.)

Fig. 3.

Conditional heterozygotes experience premature death and spontaneous seizures. (A) Survival of F/+:Dlx-Cre⁻ and F/+:Dlx-Cre⁺ mice. The fraction of each genotype surviving is plotted versus postnatal day (F/+:Dlx-Cre⁻, n = 45; F/+:Dlx-Cre⁺, n = 27). (B) Representative example of spontaneous seizure progression in a F/+:Dlx-Cre⁺ mouse. Racine score of each seizure is plotted as a function of the time before death (t = 0 at time of death). (C) Graphical representation of the age of death in animals monitored for spontaneous seizure. Average age of death: 20.6 ± 0.4 d (n = 10). (D) Graphical representation of time elapsed from initial seizure onset to death in animals monitored for spontaneous seizures. Average latency to death: 17.9 ± 3.9 h (n = 10). (E) Graphical representation of the number of seizures preceding death in animals monitored for spontaneous seizures. Average number of seizures before death: 9 ± 1 seizures (n = 10).

Fig. 4.

Conditional heterozygotes experience evoked behavioral and electrographic seizures. (A) Thermally induced seizures were evoked in all F/+:Dlx-Cre⁺ animals with a mean temperature of 39 °C. No F/+:Dlx-Cre⁻ mice presented with seizure (F/+:Dlx-Cre⁻, n = 10; F/+:Dlx-Cre⁺, n = 17). (B) Distribution of thermally induced seizure severity in F/+:Dlx-Cre⁺ mice. (C) Representative EEG traces of F/+:Dlx-Cre⁻ (n = 5) and F/+:Dlx-Cre⁺ (n = 4) mice during thermal induction at P24. Top, F/+:Dlx-Cre⁻ mouse at 39.5 °C. Bottom, F/+:Dlx-Cre⁺ mouse during GTC seizure at 39.5 °C. (D) Representative ECG traces of F/+:Dlx-Cre⁻ (n = 5) and F/+:Dlx-Cre⁺ (n = 6) mice at rest. Top, F/+:Dlx-Cre⁻ mouse: heart rate, 672.2 ± 24; PR interval, 32.8 ± 2; QRS interval, 9.4 ± 1; and QT interval, 23.5 ± 5.5. Bottom, F/+:Dlx-Cre⁺ mouse: heart rate, 658.4 ± 57; PR interval, 32.3 ± 2; QRS interval, 8.3 ± 1; and QT interval, 19.7 ± 1.5 (P > 0.05 for all parameters).