Epilepsy and neuropsychiatric comorbidities in mice carrying a recurrent Dravet syndrome SCN1A missense mutation

Abstract

Dravet Syndrome (DS) is an encephalopathy with epilepsy associated with multiple neuropsychiatric comorbidities. In up to 90% of cases, it is caused by functional happloinsufficiency of the SCN1A gene, which encodes the alpha subunit of a voltage-dependent sodium channel (Nav1.1). Preclinical development of new targeted therapies requires accessible animal models which recapitulate the disease at the genetic and clinical levels. Here we describe that a C57BL/6 J knock-in mouse strain carrying a heterozygous, clinically relevant SCN1A mutation (A1783V) presents a full spectrum of DS manifestations. This includes 70% mortality rate during the first 8 weeks of age, reduced threshold for heat-induced seizures (4.7 °C lower compared with control littermates), cognitive impairment, motor disturbances, anxiety, hyperactive behavior and defects in the interaction with the environment. In contrast, sociability was relatively preserved. Electrophysiological studies showed spontaneous interictal epileptiform discharges, which increased in a temperature-dependent manner. Seizures were multifocal, with different origins within and across individuals. They showed intra/inter-hemispheric propagation and often resulted in generalized tonic-clonic seizures. ¹⁸F-labelled flourodeoxyglucose positron emission tomography (FDG-PET) revealed a global increase in glucose uptake in the brain of Scn1a^WT/A1783V mice. We conclude that the Scn1a^WT/A1783V model is a robust research platform for the evaluation of new therapies against DS.

Figure 1

Scn1a^WT/A1783V mice present reduced body weight and high mortality during the first 6 weeks of life. (a) Animals were weighted once per week since weaning. Points represent mean values ± SEM (females Scn1a^WT/WT average n = 30; females Scn1a^WT/A1783V n = 20; males Scn1a^WT/WT n = 40; males Scn1a^WT/A1783V n = 15). Weight curves of Scn1a^WT/A1783V and control littermates mice show significant differences in elevation (p < 0.0001) but not in slope (linear regression analysis). (b) Survival curves of mice after weaning (postnatal day 21) are significantly different (p < 0.0001 log-rank test). Of note, Scn1a^WT/A1783V mice present a 24% mortality rate before weaning.

Figure 2

Scn1a^WT/A1783V mice present reduced Scn1a mRNA in hippocampus, but no significant differences in any brain region at the protein levels. (a) mRNA levels of Scn1a were determined by qRT-PCR in the cortex (Ctx), hippocampus (HC) and cerebellum (Cb) of Scn1a^WT/A1783V and aged-matched littermates (5–8 months). Values (normalized against GAPDH mRNA levels) were calculated as percentage vs control littermates and data represented as mean ± SEM (n = 6 for Scn1a^WT/WT mice and n = 8 for Scn1a^WT/A1783V mice). *p < 0.05 Kruskal-Wallis with Dunn’s post-test. (b) Membrane-enriched protein extracts prepared from Ctx, HC and Cb were used to evaluate Nav1.1 protein content by Western blotting. Bars represent the densitometric analysis of individual determinations normalized to GAPDH values. Data are represented as mean percentage ± SEM of values normalized to control mice (n = 6 for Scn1a^WT/WT mice and n = 8 for Scn1a^WT/A1783V mice). No statistical differences were found between control and Scn1a^WT/A1783V mice. The right panel corresponds to representative blots showing Nav1.1 and GAPDH bands. Original blots were cropped and re-arranged to display grouped Scn1a^WT/A1783V and control littermates. Full-length blots are available in supplemental material. (c) Additional mice were sacrificed for analysis of Nav1.1 by immunofluorescence (green). Nuclei are stained with DAPI (blue). The image shows the indicated brain areas: prefrontal cortex (pCtx); dentate gyrus of the HC (HC-DG); and Cb of representative mice (n = 4 for both groups of animals). The different regions of each structure were delimited by dotted lines based on the Allen adult mouse brain reference atlas. In pCtx roman numbers indicate its different layers. In HC-DG the following regions are included: stratum lacunosum-molecurare of CA1 (CA1-slm); and molecular layer (GD-mo), granule cell layer (DG-sg) and polymorph layer (DG-po) of dentate gyrus. In Cb: molecular layer (ml), white matter (wm), purkinje cell layer (pcl) and granule layer (gl). Scale bar 200 µm. (d) In order to study the subcellular localization of Nav1.1 channel, tissue sections were visualized using a confocal laser scanning microscope. The panel shows the labelling observed in cortical later V (Ctx-LV), HC-DG and Cb of representative mice (n = 4 for both groups of animals). Scale bar 20 µm.

Figure 3

Scn1a^WT/A1783V are prone to suffer heat-induced seizures. Mice of the indicated age ranges were exposed to controlled hyperthermia in a chamber with 0.5 °C increments in temperature every 30 s up to 45 °C or until a generalized seizure was reached. (a) Thresholds were significantly lower in Scn1a^WT/A1783V compared with their control littermates at all age ranges tested. These differences were maintained throughout all the age range tested (Scn1a^WT/WT: 1–2 mo n = 23, 2–4 mo n = 27 and 4–6 mo n = 9; Scn1a^WT/A1783V: 1–2 mo n = 26, 2–4 mo n = 20 and 4–6 mo n = 13). ***p < 0.001, One-way ANOVA with Tukey’s post-test. (b) Cumulative seizure probability showing that the risk of seizures is confined between 40–45 °C in all ages in Scn1a^WT/WT mice, whereas Scn1a^WT/A1783V present a wider temperature range at young ages.

Figure 4

Electrophysiological characterization of freely moving Scn1a^WT/A1783V mice. (a) Example of the electrical activity recorded in a Scn1a^WT/WT (top) and Scn1a^WT/A1783V (bottom) mouse during the thermal challenge. Two months-old animals (n = 5) were introduced in the heating chamber, and temperature was increased gradually while electrical activity was recorded in the prefrontal cortex/frontal associative cortex (PFC), CA1 and dentate gyrus (DG) regions. The heat source was switched on at t = 800 s. In the case of the Scn1a^WT/A1783V mouse the seizure starts around t = 1,700 s, at T = 38.94 °C. Despite the heat source was immediately disconnected, seizures persisted and were organized in clusters, even when the animal was removed from the recording chamber (around t = 2,250 s). Note the presence of IEDs at RT (t < 1,000 s) and how they increase in frequency as the temperature rises, reaching a maximum right before the seizure (bottom). In contrast, neither IEDs nor seizures were observed in the case of the Scn1a^WT/WT mouse (top). (b) Example of a seizure with focal origin in the DG that is further generalized (see arrow).

Figure 5

Scn1a^WT/A1783V mice show cognitive alterations involving task learning and visuospatial memory. (a) The MWM test performed at different ages showed increased escape latency in the visible platform (VP) in the Scn1a^WT/A1783V mice compared with their littermate controls, although significant learning was present in both groups of mice (p < 0.001 Friedman test). In contrast, the performance in the invisible platform (IP) showed differences both in absolute latencies and the slope of curves, indicating that Scn1a^WT/A1783V mice are unable to improve their escape latency throughout the training (p > 0.05 and p < 0.001 for Scn1a^WT/A1783V and Scn1a^WT/WT mice, respectively, Friedman). In concordance with the lack of spatial learning, Scn1a^WT/A1783V mice showed no preference for the target quadrant in the probe test, consistent with altered retention. Values are represented as mean ± SEM (Scn1a^WT/WT: 1–3 mo n = 11, 3–5 mo n = 12 and 5–8 mo n = 27; and Scn1a^WT/A1783V: 1–3 mo n = 17, 3–5 mo n = 10 and 5–8 mo n = 24). (b) The NOR test was applied to mice in the same age groups. Although no differences were observed 1 h after training (left panel), a significant reduction in long-term memory was noted in Scn1a^WT/A1783V mice (NOR 24 h), revealing a defect in memory consolidation. Values are represented as mean ± SEM (Scn1a^WT/WT: 1–3 mo n = 37, 3–5 mo n = 26 and 5–8 mo n = 27; Scn1a^WT/A1783V: 1–3 mo n = 28, 3–5 mo n = 30 and 5–8 mo n = 37). Statistical relevance was assessed applying one-way ANOVA with Tukey’s post-test for VP1–2, IP5, IP8, Probe 15 day 4, Probe 15 day 7, NOR 1 h and NOR 24 h or Kruskal-Wallis with Dunn’s post-test for VP3–5, IP1–4, IP6–7 and Probe 15 day 9. *p < 0.05, **p < 0.01, ***p < 0.001 for comparison of Scn1a^WT/A1783V and Scn1a^WT/WT mice. Symbols *, ⁺ and ^# correspond to the 1–3, 3–5 and 5–8 months age ranges, respectively.

Figure 6

Scn1A^WT/A1783V mice show motor impairment. (a–c) Mice of the indicated age ranges were subjected to the rotarod, inverted grid and elevated beam tests, as indicated. The latency to fall from the rotatory rod was decreased in the Scn1A^WT/A1783V mice compared with their control littermates, with significant differences observed from 3 months of age. For the other two tests, a significant reduction was observed in all age groups. Each bar represents the mean ± SEM of elapsed time (s) and are the mean of at least two trials (Scn1a^WT/WT: 1–3 mo n = 31/25/19, 3–5 mo n = 16/19/14 and 5–8 mo n = 24/14/26; Scn1a^WT/A1783V: 1–3 mo n = 18/13/11, 3–5 mo n = 16/29/14 and 5–8 mo n = 25/16/31; respectively in each of the tests performed). *p < 0.05, **p < 0.01, and ***p < 0.001. One-way ANOVA with Tukey’s post-test for (a) and Kruskal-Wallis with Dunn’s post-test for (b,c).

Figure 7

Scn1A^WT/A1783V mice show an altered interaction with the environment. Mice of the indicated age ranges were subjected to the open-field test and showed a reduction in the time spent in the center of the arena -indicative of anxiety- (a), hyperkinesia (b) and an increased number of stereotypies (c). No differences in total distance moved were detected in any age-range (data not shown). Other alterations of animal behavior included a reduction in the normal tendency to hide objects in the marble burying test (d) and a poor performance in the nest building test (e). The graph represents the percentage of mice that complete, initiate or fail to initiate the task during one night (dark, medium and light colors, respectively). In contrast, Scn1A^WT/A1783V performed relatively well in the social interaction task (f), with a reduction in the number of contacts only observed in the 3–5 months age range (left Y axis), which coincided with a significantly higher latency to approach the unfamiliar mouse for the first time (right Y axis), compared with their control littermates. Values are represented as mean ± SEM (Scn1a^WT/WT: 1–3 mo n = 44/44/14/26/16/16, 3–5 mo n = 31/30/13/34/6/11, and 5–8 mo n = 29/27/17/8/17/18; Scn1a^WT/A1783V: 1–3 mo n = 26/29/14/20/6/15, 3–5 mo n = 33/34/9/37/12/11 and 5–8 mo n = 35/36/24/18/20/15; respectively in each of the test performed). *p < 0.05; ***p < 0.001, Kruskal-Wallis with Dunn’s post-test.

Figure 8

Scn1a^WT/A1783V mice show increased glucose uptake in the brain. Mice were subjected to ¹⁸F-FDG PET at the indicated ages. (a) Left panel correspond to maximum intensity projection PET images of representative mice showing brain ¹⁸F-FDG uptake. The quantification of positron emission (mean SUV) is represented in the right panel. (b) Mice in the 5–8 age group were sacrificed after PET, and isotope incorporation was quantified in a gamma counter (expressed as % of the injected dose). (1–2 mo n = 7; 5–8 mo n = 14). *p < 0.05; **p < 0.01. Mann Whitney U test.