X-linked infantile spasms syndrome (ISSX) is a clinically devastating developmental epileptic encephalopathy with life-long impact. Arx^(GCG)10+7, a mouse model of the most common triplet-repeat expansion mutation of ARX, exhibits neonatal spasms, electrographic phenotypes and abnormal migration of GABAergic interneuron subtypes. Neonatal presymptomatic treatment with 17β-estradiol (E2) in Arx^(GCG)10+7 reduces spasms and modifies progression of epilepsy. Cortical pathology during this period, a crucial point for clinical intervention in ISSX, has largely been unexplored, and the pathogenic cellular defects that are targeted by early interventions are unknown. In the first postnatal week, we identified a transient wave of elevated apoptosis in Arx^(GCG)10+7 mouse cortex that is non-Arx cell autonomous, since mutant Arx-immunoreactive (Arx+) cells are not preferentially impacted by cell death. NeuN+ (also known as Rbfox3) survival was also not impacted, suggesting a vulnerable subpopulation in the immature Arx^(GCG)10+7 cortex. Inflammatory processes during this period might explain this transient elevation in apoptosis; however, transcriptomic and immunohistochemical profiling of several markers of inflammation revealed no innate immune activation in Arx^(GCG)10+7 cortex. Neither neonatal E2 hormone therapy, nor ACTH(1-24), the frontline clinical therapy for ISSX, diminished the augmented apoptosis in Arx^(GCG)10+7, but both rescued neocortical Arx+ cell density. Since early E2 treatment effectively prevents seizures in this model, enhanced apoptosis does not solely account for the seizure phenotype, but may contribute to other aberrant brain function in ISSX. However, since both hormone therapies, E2 and ACTH(1-24), elevate the density of cortical Arx+-interneurons, their early therapeutic role in other neurological disorders hallmarked by interneuronopathy should be explored.

Nanostring mRNA expression profiling was conducted using nCounter® mouse Neuropathology Plus Panel, which contains probe sets to query the expression of 800 unique mRNAs (https://www.nanostring.com/products/gene-expression-panels/ncounter-neuropathology-panels). Table S3 (Siehr et al., 2020) contains a list of 29 additional probes that were designed to profile mRNA from Arx targets, and other factors found to be dysregulated in the Arx^(GCG)7/Y model (Kitamura et al., 2009; Mattiske et al., 2016). RNA was extracted from P7 cortical (neocortex and hippocampus) from the left hemisphere using a Qiagen RNeasy Plus kit with guide DNA eliminator column (74134). Cortices from Arx^(GCG)10+7 and WT male littermates were used (six WT and six Arx^(GCG)10+7). The BCM Genome and RNA Profiling Core performed the assay on an nCounterMAX5s and RNA quality control using a NanoDrop One and Agilent 2100 Bioanalyzer. Only RNA samples that met strict quality and purity criteria (RIN>8 and no anomalies in Bioanalyzer spectrograph) were submitted for analysis. Output data were analyzed and quality controlled using Nanostring NSolver software (version 4.0.70). Differential expression between genotypes was conducted using a linear regression model in the Advanced Analysis software (version 2.0.115) in nSolver and corrected for batch and litter effects (see https://www.nanostring.com/products/analysis-software/advanced-analysis for more information). mRNAs that were below the detection limit of nCounter MAX 5 s were excluded from analysis. Pathway scoring analysis for cytokines (see Table S1 in Siehr 2020 for list of genes) was also performed in Advanced Analysis software (version 2.0.115) in nSolver, which uses the first principal component of expression data to determine pathway-level information from a group of genes (Tomfohr et al., 2005). We are grateful for the additional support of the Genomic and RNA Profiling Core at Baylor College of Medicine and the expert assistance of the core director, Lisa D. White, and Daniel Kraushaar.

To query and analyze these datasets, nSolver (version 4.0) with Advanced Analysis software (version 2.0.115) or updated versions of these software must be used. Software is free to download and manuals and instructional videos are available here https://www.nanostring.com/products/analysis-software/nsolver.