Infantile status epilepticus disrupts myelin development

Temporal lobe epilepsy (TLE) is the most prevalent type of epilepsy in adults; it often starts in infancy or early childhood. Although TLE is primarily considered to be a grey matter pathology, a growing body of evidence links this disease with white matter abnormalities. In this study, we explore the impact of TLE onset and progression in the immature brain on white matter integrity and development utilising the rat model of Li-pilocarpine-induced TLE at the 12th postnatal day (P). Diffusion tensor imaging (DTI) and Black-Gold II histology uncovered disruptions in major white matter tracks (corpus callosum, internal and external capsules, and deep cerebral white matter) spreading through the whole brain at P28. These abnormalities were mostly not present any longer at three months after TLE induction, with only limited abnormalities detectable in the external capsule and deep cerebral white matter. Relaxation Along a Fictitious Field in the rotating frame of rank 4 indicated that white matter changes observed at both timepoints, P28 and P72, are consistent with decreased myelin content. The animals affected by TLE-induced white matter abnormalities exhibited increased functional connectivity between the thalamus and medial prefrontal and somatosensory cortex in adulthood. Furthermore, histological analyses of additional animal groups at P15 and P18 showed only mild changes in white matter integrity, suggesting a gradual age-dependent impact of TLE progression. Taken together, TLE progression in the immature brain distorts white matter development with a peak around postnatal day 28, followed by substantial recovery in adulthood. This developmental delay might give rise to cognitive and behavioural comorbidities typical for early-onset TLE.


Supplementary methods
On postnatal day (P) 11, the animals were marked for identification and randomly assigned to SE or control groups, ensuring that each group contained animals from multiple litters to exclude a litter effect. Each litter consisted of both controls and SE exposed animals. In the group of additional animals for histology, the rats were at the same time randomly allocated for tissue collection at postnatal days 15, 18, or 28 (P15, P18, and P28) while controlling for the litter effect. All pups received LiCl (127 mg/kg; LiCl 127 mg was dissolved in 5 ml of water for injection and injected intraperitoneally 127 mg/5ml/kg) and were returned to their dams. On P12 pups were separated from their dams and transferred into a special silent room with controlled conditions. Experiments were always performed at the same time of the day, between 10 am and 2 pm (i.e. during the light period). Animals were placed individually into small containers and they were maintained at +33±1°C to compensate for the immature thermoregulation at this age (Conklin and Heggeness, 1971) during the entire period of separation from their mothers. SE was induced with a single intraperitoneal injection of pilocarpine (35mg/kg; 35 mg of pilocarpine was dissolved in 5 ml of saline and injected). Controls received corresponding volume of saline instead.
The severity of motor SE was assessed using the following scoring system: 0 -normal behaviour 1 -stereotypic behaviour (face washing, scratching), isolated myoclonic jerks 2 -head bobbing, pivoting, swimming movements 3 -clonic seizures with preserved righting reflexes 4 -repeated periods of wild running 5 -generalized tonic-clonic seizures with loss of righting reflexes.
Animals were assigned a score for the most severe behavioural characteristics. Latency to the onset of motor seizures was recorded. Mortality was recorded throughout the entire experimental period. Only rats that exhibited behavioural manifestations of seizures progressing to forelimbs clonus (i.e., score 3) for at least 1 hour and without periods of wild running and generalized tonic-clonic seizures (score 4-5) were used for further experiments.
After 1.5 hours of convulsive SE, animals were given a single dose of paraldehyde (0.07ml/kg; 0.07 ml of paraldehyde was mixed with 5 ml of saline and injected intraperitoneally). Controls received paraldehyde in the same dose. Approximately 30 min later they were injected subcutaneously with 0.5 ml saline to restore the volume loss. After the brief recovery, pups were returned to their dams (the duration of isolation from mothers in the control and SE groups was the same ~ 4 hours).
The weight of pups was checked daily and animals that did not gain any weight within 24hrs after SE were given 0.5ml of saline with 10% of glucose subcutaneously to prevent dehydration and further weight loss. To minimize the effects of variability in individual groups, the data were used to calculate relative body weight (body weight at P11 was taken as 100%). The difference in the relative body weights between two consecutive days (starting at P11) was used as a measure of weight gain ( Figure  S4). Statistical comparison of relative weight gains between controls and SE animals was done using RStudio (version 1.2.5033). Data were compared using Mixed-effects analysis with False Discovery Rate (FDR) correction for multiple comparisons. Mixed effect analysis was used instead of One-way repeated measure ANOVA because of missing values.

Statistical disclaimer:
Multiple testing correction techniques are known to reduce false-positive discovery while increasing the risk of type II error (false-negative results) (Drachman, 2012;Ranganathan et al., 2016;White et al., 2019). To avoid the possible unnecessary introduction of false-negative values in cases when the likelihood of false-positive discovery is below 1 per compared set, we consider p-value without adjustment (displayed in table S1 and S2) a true p-value in our ROI statistics with less than 20 comparisons per method (S1) or position (S2).   Figure S1 -ROIs representing the default mode network. Locations of regions of interest used for the analysis of default mode network connectivity in rat brain fMRI; slice thickness 1 mm. ROIs are overlaid on the reference brain, which was acquired with a functional spin-echo echo planar imaging sequence. Numbers located under individual brains indicate distance from the bregma in mm.

Figure S2 -Brain sections for analysis of optical density.
Illustrative positions of Black-Gold IIstained brain sections selected for analysis of optical density in large white matter structures with indicated approximate distance from the bregma in the adult animal. Coronal brain sections from 15-, 18-, 28-, and 72-day-old rats are displayed.

Figure S3 -Position of regions of interest (ROIs) for the optical density quantification.
Representative positions of ROIs in white matter structures in left and right hemisphere Figure S4 -Animal weight.
Daily weight percentage change (A) of rats following the induction of status epilepticus (SE) and controls (Ctrl). Daily follow-up is displayed across 16 days starting at the day of SE induction (0). The symbols represent the mean value and the error bars represent the mean standard deviation. The red asterisk (*) indicates significantly different values when comparing SE and control rats using Mixedeffects analysis. Displayed data include weight information from all animals used in the study (both MRI and histology groups). Graph (B) shows absolute animal weight [g] on the day of MRI & fMRI data collection (postnatal day (P) 28, 72 and 75) for animals after SE and controls within the MRI group.