Sex differences in physiological response to increased neuronal excitability in a knockin mouse model of pediatric epilepsy

Abstract

Background: Epilepsy is a common neurological disease; however, few if any of the currently marketed antiseizure medications prevent or cure epilepsy. Discovery of pathological processes in the early stages of epileptogenesis has been challenging given the common use of preclinical models that induce seizures in physiologically normal animals. Moreover, despite known sex dimorphism in neurological diseases, females are rarely included in preclinical epilepsy models.

Methods: We characterized sex differences in mice carrying a pathogenic knockin variant (p.N1768D) in the Scn8a gene that causes spontaneous tonic-clonic seizures (TCs) at ∼3 months of age and found that heterozygous females are more resilient than males in mortality and morbidity. To investigate the cellular mechanisms that underlie female resilience, we utilized blood-brain barrier (BBB) and hippocampal transcriptomic analyses in heterozygous mice before seizure onset (pre-TC) and in mice that experienced ∼20 TCs (post-TC).

Results: In the pre-TC latent phase, both sexes exhibited leaky BBB; however, patterns of gene expression were sexually dimorphic. Females exhibited enhanced oxidative phosphorylation and protein biogenesis, while males activated gliosis and CREB signaling. After seizure onset (chronic phase), females exhibited a metabolic switch to lipid metabolism, while males exhibited increased gliosis and BBB dysfunction and a strong activation of neuroinflammatory pathways.

Conclusion: The results underscore the central role of oxidative stress and BBB permeability in the early stages of epileptogenesis, as well as sex dimorphism in response to increasing neuronal hyperexcitability. Our results also highlight the need to include both sexes in preclinical studies to effectively translate results of drug efficacy studies.

Keywords: Scn8a; blood brain barrier; model organisms; sex dimporphism; transcriptomics; voltage-gated sodium channel.

Figure 1. Lifetime number of TCs, survival, seizure bouts and gaps

(A) Kaplan–Meier survival curves (solid lines) for 17 females and 17 males. Log rank test P-value <0.0001. Cumulative number of TCs is shown in dotted lines. (B) Pattern of TC frequency by day for a representative female (red vertical lines) and male (blue vertical lines) illustrating seizure onset, bouts, gaps and time of death. Time-points for sampling are indicated with bold arrow (pre-TC) or with bracket (post-TC).

Figure 2. BBB paracellular permeability to sucrose is increased in male and female D/+ mice both before and after seizure onset

In situ brain perfusion with 14C-sucrose as a vascular permeability marker shows significantly elevated radioactivity represented by brain-to-perfusate radioactivity ratios (RBr %) in brains of mice before and after detection of seizures and compared with controls (+/+ mice). Data are expressed as mean ± SD of 5–6 animals per treatment group (*P<0.05, **P<0.01, ***P<0.001, ****P<0.0001; ns = not significant).

Figure 3. Genes and pathways identified by analysis of RNAseq data that are shared by the sexes or unique to each sex

Number of differentially expressed genes (DEGs) identified in (A) pre-TC and (B) post-TC females and males. Number of significantly enriched canonical pathways identified by IPA in (C) pre-TC or (D) post-TC females and males. Female and male symbols and up or down arrows refer to number of genes or pathways that were up/down regulated or activated/deactivated, respectively.

Figure 4. Canonical pathways in (A) pre-TC females and (B) males

Bubble charts showing significantly enriched pathways (−log [P-value-] ≥ 1.3) that are color-coded by strength of activation (positive z-score ≥ +2.0) or deactivation (negative z-score ≤ −2.0). Circumference of pie chart represents ratio of number of overlapping genes (DEGs) to total number of genes in pathway. Dendrograms embedded within each bubble chart were constructed from distance matrices based on the Jaccard similarity index, calculated as the intersection over union for each pair of gene sets (see Materials and Methods).

Figure 5. Canonical pathways in (A) post-TC females and (B) males

See Figure 4 for the legend.

Figure 6. Integrative model for sex differences in epileptogenesis

Impaired inactivation gating caused by the N1768D Na_V1.6 variant is shown at the top, resulting in persistent sodium current, (1) glutamate excitotoxicity, and (2) downstream oxidative stress (OS). This process continues throughout pre-TC and post-TC stages, as shown centrally above and below the seizure onset lines, respectively. Reactive oxygen species (ROS) shown as potential inducers of (3) astrogliosis in pre-TC males, whereas (4) estrogen may in part mitigate effects of OS and extent of BBBD in females (see text), as well as influence increased protein biogenesis and OXPHOS. CREB signaling in neurons may have a direct effect in further enhancing (5) neuronal excitability and (6) BBB permeability (see text), whereas in females BBBD may be caused by (7) NET signaling. BBBD may lead to seizure onset (higher line in chart for (8) males than (9) females), marking the transition to more pathological conditions. In post-TC males, albumin extravasation via TGFβ/BMP signaling increases astrogliosis (10) and increased (11) MAPK/ERK/CREB signaling leads to activation of (12) neuroinflammatory pathways both of which increase (13) BBB permeability. Similar signaling in post-TC females is (13) suppressed (see text). Cytoskeletal remodeling associated with increased (13) BBBD occurs in both sexes. Activities that are part of the ‘tetrad of pathogenic processes’ common to neurological diseases (see text) are indicated in dark gray filled shapes. Feedback loops are indicated in arrows with dotted lines. Hexagonals, observed processes; rounded rectangles, inferred processes; rectangle with square corners, identified canonical pathways; ovals, inferred molecules.