Developmental changes in brain activity of heterozygous Scn1a knockout rats

Abstract

Introduction: Dravet syndrome (DS) is an infantile-onset developmental and epileptic encephalopathy characterized by an age-dependent evolution of drug-resistant seizures and poor developmental outcomes. Functional impairment of gamma-aminobutyric acid (GABA)ergic interneurons due to loss-of-function mutation of SCN1A is currently considered the main pathogenesis. In this study, to better understand the age-dependent changes in the pathogenesis of DS, we characterized the activity of different brain regions in Scn1a knockout rats at each developmental stage.

Methods: We established an Scn1a knockout rat model and examined brain activity from postnatal day (P) 15 to 38 using a manganese-enhanced magnetic resonance imaging technique (MEMRI).

Results: Scn1a heterozygous knockout (Scn1a ^+/-) rats showed a reduced expression of voltage-gated sodium channel alpha subunit 1 protein in the brain and heat-induced seizures. Neural activity was significantly higher in widespread brain regions of Scn1a ^+/- rats than in wild-type rats from P19 to P22, but this difference did not persist thereafter. Bumetanide, a Na⁺-K⁺-2Cl^- cotransporter 1 inhibitor, mitigated hyperactivity to the wild-type level, although no change was observed in the fourth postnatal week. Bumetanide also increased heat-induced seizure thresholds of Scn1a ^+/- rats at P21.

Conclusions: In Scn1a ^+/- rats, neural activity in widespread brain regions increased during the third postnatal week, corresponding to approximately 6 months of age in humans, when seizures most commonly develop in DS. In addition to impairment of GABAergic interneurons, the effects of bumetanide suggest a possible contribution of immature type A gamma-aminobutyric acid receptor signaling to transient hyperactivity and seizure susceptibility during the early stage of DS. This hypothesis should be addressed in the future. MEMRI is a potential technique for visualizing changes in basal brain activity in developmental and epileptic encephalopathies.

Keywords: Dravet syndrome (DS); bumetanide (BTN); developmental and epileptic encephalopathy (DEE); functional neuroimaging; gamma-aminobutyric acid (GABA).

Figure 1

Targeting construct and generation of global Scn1a knockout rats. (A) The arrows show the position of the genotyping primer set (forward 5′-TAATAACTTTTAATGCTATC-3′, reverse 5′-CTTCCCAGCTTCCAAGTCAC-3′). Genotype analysis shows a 272 bp band in the wild-type allele and a 178 bp band in the Scn1a^em1kyo mutant. The upper band of the Scn1a^em1kyo mutant is nonspecific. Genotypes are indicated by +/+ for wild-type, +/– for heterozygous, and –/– for homozygous animals. (B) Western blotting of brain membrane proteins from wild-type (+/+), heterozygous (+/–), and homozygous (–/–) Scn1a rats at P13 using an anti- voltage-gated sodium channel alpha subunit 1 (Na_V1.1) antibody. β-actin was used as the internal control. Original blots/gels are presented in Supplementary Figure 1. (C) Relative Na_V1.1 protein levels normalized to β-actin. (D) Survival curves of wild-type (n = 15), Scn1a⁺/− (n =15), and Scn1a^−/− (n = 20) rats. Scn1a^−/− rats exhibited ataxia and seizures from postnatal day (P) 10, gradually progressing to weight loss and complete loss of postural control. They became inactive and did not survive beyond P18. In wild-type and Scn1a⁺/− rats, there were no spontaneous deaths. (E) Body weight curves of wild-type (blue circle), Scn1a⁺/− (red square), and Scn1a^−/− (black triangle) rats. The sample size is shown below the graph. Scn1a⁺/− rats showed no difference in body weight compared to wild-type rats. Data are presented as mean ± standard deviation. (F) Ictal electroencephalography (EEG) recording of heat-induced seizures from Scn1a⁺/− rats at P21. Spiking activity was recorded, and generalized tonic-clonic seizures were observed. Open triangle indicates seizure onset. Asterisk indicates the expanded EEG trace of spiking activity. (G) Representative interictal EEG recordings from wild-type and Scn1a⁺/− rats at P21. No interictal epileptic discharge was found.

Figure 2

Schematic diagram of the experiments. (A) Manganese-enhanced magnetic resonance imaging (MEMRI) experiment. Manganese chloride was intraperitoneally administered immediately after the first T1 mapping, which was reconfirmed 24 h later. (B) MEMRI experiment in rats treated with bumetanide (BTN) or normal saline (NS). BTN solution (0.2 mg/kg) or an equal volume of NS (4 mL/kg) was injected intraperitoneally twice a day beginning at postnatal day (P) 12, and MEMRI was performed for each rat on P21 or P27. (C) Heat-induced seizure experiment in rats treated with BTN or NS. After daily injections of BTN or NS, seizures were induced in rats by a warm bath at P21 or P27. (D) Representative coronal T1 map and regions of interest. Regions were manually obtained using the Paxinos and Watson atlas. The image shows the average of T1 maps. Colors indicate 17 different brain regions, which are overlaid on averaged images of the MEMRI T1 map. M1, primary motor cortex; M2, secondary motor cortex; CPu, caudate-putamen (striatum); LGP, lateral globus pallidus; RSGb, retrosplenial granular b cortex; DG, dentate gyrus; S1BF, primary somatosensory cortex, barrel field; Au1, primary auditory cortex; VPM, ventral posteromedial thalamic nucleus; VPL, ventral posterolateral thalamic nucleus; nRT, reticular thalamic nucleus; LHb, lateral habenular nucleus; CA3, CA3 field of the hippocampus; V1B, primary visual cortex, binocular area; V2L, secondary visual cortex, lateral area; Vermis, cerebellar vermis; Hemisphere, cerebellar hemisphere.

Figure 3

Age-dependent changes in T1_post in each brain region of wild-type and Scn1a^+/− rats. (A) T1_post showed age-dependent changes in each region of wild-type (WT, green circle) and Scn1a^+/− rats (Scn1a^+/−, magenta circle). Tukey multiple-comparison test showed a significant decrease in T1_post of Scn1a^+/− rats during P19–22: M1(mean difference 0.205s; 95% CI 0.073–0.337s), M2 (0.198s; 0.066–0.330s), CPu (0.355s;−0.203–0.467s), LGP (0.247s; 0.115–0.380s), RSGb (0.318s; 0.203–0.467s), DG (-0.272s; 0.140–0.404s), S1BF (0.299s; 0.167–0.431s), Au1 (0.269s; 0.137–0.401s), VPM (0.286s; 0.154–0.419s), VPL (0.316s; 0.184–0.448s), nRT (0.280s; 0.142–0.418s), LHb (0.202s; 0.070–0.334s), CA3 (0.277s; 0.145–0.409s), V1B (0.278s; 0.147–0.410s), V2L (0.291s; 0.159–0.423s), Vermis (0.126s;−0.006–0.258s), and Hemisphere (0.229s; 0.097–0.361s). These differences disappeared after P23, but are significantly decreased during P35–38: CPu (0.148s; 0.028–0.268s), LGP (0.120s; 0.000–0.240s), LHb (0.136s; 0.012–0.260s), and CA3 (0.143s; 0.023–0.263s). Tukey's multiple-comparison test; *p < 0.05, ** p < 0.01, *** p < 0.001. Data are presented as mean ± standard error of the mean (SEM). Dots represent individual data points. (B) Representative MEMRI T1 map at P21. Yellow lines show part of regions of interest. MEMRI, manganese-enhanced magnetic resonance imaging; M1, primary motor cortex; CI, confidence interval; M2, secondary motor cortex; CPu, caudate-putamen (striatum); LGP, lateral globus pallidus; RSGb, retrosplenial granular b cortex; DG, dentate gyrus; S1BF, primary somatosensory cortex, barrel field; Au1, primary auditory cortex; VPM, ventral posteromedial thalamic nucleus; VPL, ventral posterolateral thalamic nucleus; nRT, reticular thalamic nucleus; LHb, lateral habenular nucleus; CA3, CA3 field of the hippocampus; V1B, primary visual cortex, binocular area; V2L, secondary visual cortex, lateral area; Vermis, cerebellar vermis; Hemisphere, cerebellar hemisphere; P, postnatal day.

Figure 4

T1_post and seizure characteristics after administration of BTN or NS. (A) Scheffé multiple-comparison test of T1_post in rats at P21 after the administration of BTN (orange circle) or NS (purple square). Scn1a^+/− rats showed a significant increase in the T1_post of widespread brain regions in the BTN-treated group compared to the NS-treated group: M1 (mean difference−0.225s; 95% CI−0.423-−0.027s), M2 (-0.210s;−0.408−0.011s), CPu (-0.260s;−0.458−0.062s), LGP (-0.288s;−0.486−0.090s), RSGb (-0.236s;−0.435−0.037s), DG (-0.265s;−0.463−0.067s), S1BF (-0.235s;−0.435−0.035s), Au1 (-0.222s;−0.421−0.024s), VPM (-0.216s;−0.414−0.018s), VPL (-0.223s;−0.420−0.025s), nRT (-0.272s;−0.470−0.074s), LHb (-0.197s;−0.396−0.002s), CA3 (-0.200s;−0.397−0.002s), V1B (-0.222s;−0.420−0.023s), V2L (-0.223s;−0.422−0.024s), Vermis (-0.213s;−0.412−0.015s), and Hemisphere (-0.272s;−0.471−0.073s). There was no difference between the T1_post of NS-treated and BTN-treated WT rats. n = 6 per group, Scheffé multiple-comparison test; *p < 0.05, **p < 0.01, ***p < 0.001. Data are presented as mean ± SEM. (B) Comparison of seizure latency, seizure duration, and seizure score between BTN-treated and NS-treated Scn1a^+/− rats at P21. BTN significantly increased seizure latency of Scn1a^+/− rats at P21. n = 6 per group, Mann–Whitney U test; *p < 0.05. Data are presented as mean ± SEM. BTN, bumetanide; NS, normal saline; P, postnatal day; M1, primary motor cortex; CI, confidence interval; M2, secondary motor cortex; CPu, caudate-putamen (striatum); LGP, lateral globus pallidus; RSGb, retrosplenial granular b cortex; DG, dentate gyrus; S1BF, primary somatosensory cortex, barrel field; Au1, primary auditory cortex; VPM, ventral posteromedial thalamic nucleus; VPL, ventral posterolateral thalamic nucleus; nRT, reticular thalamic nucleus; LHb, lateral habenular nucleus; CA3, CA3 field of the hippocampus; V1B, primary visual cortex, binocular area; V2L, secondary visual cortex, lateral area; Vermis, cerebellar vermis; Hemisphere, cerebellar hemisphere; SEM, standard error of the mean.

Figure 5

T1_post at P27 of wild-type and Scn1a^+/− rats after administration of BTN or NS. (A) Tukey multiple-comparison test of T1_post in rats at P27 after administration of BTN (orange circle) or NS (purple square). There was no significant difference in T1_post between BTN-treated and NS-treated Scn1a^+/− rats. n = 6 per group (Tukey multiple-comparison test). Data are presented as mean ± SEM. (B) Comparison of seizure latency, seizure duration, and seizure score of Scn1a^+/− rats at P27 between the BTN-treated and NS-treated groups. BTN did not affect the seizure characteristics at P27. n = 7 per group, Mann–Whitney U test. Data are presented as mean ± SEM. BTN, bumetanide; NS, normal saline; P, postnatal day; SEM, standard error of the mean.