Age- and sex-dependent susceptibility to phenobarbital-resistant neonatal seizures: role of chloride co-transporters

Abstract

Ischemia in the immature brain is an important cause of neonatal seizures. Temporal evolution of acquired neonatal seizures and their response to anticonvulsants are of great interest, given the unreliability of the clinical correlates and poor efficacy of first-line anti-seizure drugs. The expression and function of the electroneutral chloride co-transporters KCC2 and NKCC1 influence the anti-seizure efficacy of GABAA-agonists. To investigate ischemia-induced seizure susceptibility and efficacy of the GABAA-agonist phenobarbital (PB), with NKCC1 antagonist bumetanide (BTN) as an adjunct treatment, we utilized permanent unilateral carotid-ligation to produce acute ischemic-seizures in post-natal day 7, 10, and 12 CD1 mice. Immediate post-ligation video-electroencephalograms (EEGs) quantitatively evaluated baseline and post-treatment seizure burdens. Brains were examined for stroke-injury and western blot analyses to evaluate the expression of KCC2 and NKCC1. Severity of acute ischemic seizures post-ligation was highest at P7. PB was an efficacious anti-seizure agent at P10 and P12, but not at P7. BTN failed as an adjunct, at all ages tested and significantly blunted PB-efficacy at P10. Significant acute post-ischemic downregulation of KCC2 was detected at all ages. At P7, males displayed higher age-dependent seizure susceptibility, associated with a significant developmental lag in their KCC2 expression. This study established a novel neonatal mouse model of PB-resistant seizures that demonstrates age/sex-dependent susceptibility. The age-dependent profile of KCC2 expression and its post-insult downregulation may underlie the PB-resistance reported in this model. Blocking NKCC1 with low-dose BTN following PB treatment failed to improve PB-efficacy.

Keywords: KCC2; NKCC1; bumetanide; ischemia; neonatal seizures; phenobarbital.

Figure 1

(A) Representative video-frame from a ligated pup having an ischemic tonic-clonic seizure. (B,C) Schematics of the time-line and experimental design.

Figure 2

Age-dependent seizure burden and PB-efficacy. (A–C) Representative electrographic traces of ischemic seizures recorded with sub-dermal scalp electrodes at P7, P10, and P12 and their associated behavioral grades on video. Arrowheads show the start and end of ictal events. (D–F) EEG seizure burden, mean number of ictal events, and ictal durations in ligated-controls that received saline injections at each hour after ligation. Seizure burden after ischemia at baseline recording was highest at P7, and was significantly more severe than at P10 (p = 0.01) and P12 (p = 0.03): pairwise t-test (G–I). Electrographic seizure burden, mean number of ictal events, and ictal durations in ligated-treated mice that received PB (1 h post-ligation) and BTN (2 h post-ligation; adjunct to PB). PB (25 mg/kg; IP) was inefficacious as an anti-seizure agent at P7. At the same loading dose, PB was significantly efficacious as an anti-seizure agent at P10 and P12. BTN as an adjunct failed to improve PB-efficacy at any age tested, and significantly blunted PB-efficacy at P10. PB-efficacy at P10 and P12 (G) was due to significant reduction in mean number of ictal events (H). Ictal durations were not significantly different at any age tested (I).

Figure 3

Behavioral correlates of EEG seizures and PB-efficacy. (A) Electrographic seizures (grade 0–2) occurred at all ages tested, and responded well to PB treatment at P10 and P12. This response was not significant at P7 (repeated measures ANOVA). (B) Likewise, the convulsive seizures (grades 3–6) also showed significant PB-efficacy at P10 and P12 only with significant BTN aggravation of the convulsive seizures at P10 (repeated measures ANOVA). Gray ^* = P10, ^# = P12, represent significant p-values for pairwise t-tests. Therefore, PB failed to block either electrographic or convulsive seizures at P7.

Figure 4

Ictal events vs. PB-efficacy. (A–C) Correlations of the number of ictal events before treatment (i.e., during baseline EEG recording) to the number of ictal events post-PB treatment at P7, P10, and P12. Brackets in (B,C) show the difference in the PB-efficacy at P10 vs. P12. Post-PB seizure suppression at P12 was consistently significant and uniform regardless of baseline seizure severities (p = 0.09). In contrast, at P10, PB-efficacy was significantly dependent on the baseline seizure severity (i.e., better efficacy with lower baseline seizure loads compared to higher baseline seizure loads). Additional correlations run for baseline vs. post-PB seizure burdens showed similar results (see brackets in B; first bracket ≤20 i vs. second bracket >20 ictal events); low baseline seizure burdens and PB-efficacy were not significantly correlated (for baseline seizure burden ≤250 s; r = 0.35, p = 0.39); high baseline seizure burdens showed significant positive correlations to their post-PB seizure burdens (for baseline seizure burden ≤250 s; r = 0.61, p = 0.03). This may indicate the potential role of the number of ictal events that have occurred before treatment on the efficacy of anti-seizure agents at P10 (p = 0.001).

Figure 5

Age-dependent stroke injury. (A,B) Severity of the ischemic injury evaluated at P18 in ligated-treated mice. Histopathological analyses were performed on a series of coronal brain sections harvested at P18 for all the ages investigated. Post-ischemic P7 brains were significantly less vulnerable to necrotic infract injury compared to P10 and P12. Both hemispheric and hippocampal atrophies associated with stroke injury were significantly higher at P10 and P12. The stroke severities between P10 and P12 were not significantly different. ^*p < 0.05.

Figure 6

Developmental profile of KCC2 expression. (A) IHC for KCC2 in cortex as a function of postnatal age in naïve brains respectively (Scale bar = 250 um) showed increased neuronal expression with age. (B) Western blot quantitation of KCC2 expression as a function of postnatal age in naïve brains (n = 3 for each age). Bar graphs show the co-transporter expression normalized to actin expression of the same brains. (C) A significant sex-dependent lag of KCC2 expression was detected at P7 in naïve males compared to age-matched naïve females (^*p < 0.05), and this dimorphism was not significant at older ages [M, male; F, female (n = 2 each at every age)]. Analyses for NKCC1 in the same brain samples are shown in Supplementary Figures 4A–C.

Figure 7

Seizure severity by age and sex. (A,B) Baseline seizure burdens pooled for ligated-control and ligated-treated pups showed an age-dependent susceptibility for ischemic seizures that was significant in males but not in females. ^*p < 0.05

Figure 8

Post-ischemic KCC2 downregulation. Western blot quantification of post-ligation expression of KCC2 in P7, P10, and P12 ligated pups at acute and sub-acute time-points after ischemia (n = 3 each) in ipsi- and contralateral (i.e., injured and uninjured hemispheres, respectively) hemispheres; (A) At P7, the significant downregulation of KCC2 detected within 6–8 h of ligation (^*p = 0.004) showed a complete recovery by 96 h. (B) At P10, KCC2 downregulation showed the same trend as at P7 at 48 h. (C) P12 pups showed a significant downregulation of KCC2 at 24 h (^*p < 0.02). (D) Scatter plot of KCC2 expression levels, normalized to actin and shown as the percent of levels in their respective contralateral uninjured hemispheres. All acute time-points (6–48 h) were pooled from all three age groups (n = 13). Data show that ischemia in the CD1 mice results in an acute KCC2 downregulation in the ischemic injured hemispheres with approximately 45.35% reduction in mean expression (pairwise t-test, p = 0.0002; contralateral control is 100%, which is represented as a dotted gray line as a ratio of 1).