Cell-Type-Specific Transcriptomic Alterations in the Epileptic Prefrontal Cortex and Complementary Neuroimaging...

Epilepsy is a complex neurological disorder involving distributed neuronal and glial networks. The cellular transcriptional states associated with epilepsy and their relationship to macroscopic brain abnormalities remain incompletely characterized. We therefore performed a two-part exploratory study using a public postmortem prefrontal cortex (PFC) single-nucleus RNA sequencing (snRNA-seq) dataset and an independent clinical neuroimaging cohort.

Public snRNA-seq data from eight individuals classified as having epilepsy and seven controls were processed with CellRanger (10x Genomics, Pleasanton, California) and Seurat (Satija Lab, New York Genome Center, New York, New York). Cell-type composition, exploratory nucleus-level differential expression, Gene Ontology (GO)/Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment, Monocle pseudotime analysis, and SCENIC (Single-Cell rEgulatory Network Inference and Clustering) regulon inference were evaluated. Separately, 18F-fluorodeoxyglucose positron emission tomography-computed tomography (18F-FDG PET-CT) and diffusion tensor imaging (DTI) data from 16 patients with neocortical epilepsy were analyzed using within-patient lesion-versus-contralateral comparisons. Because the sequencing and imaging cohorts were independent, the two components were interpreted as complementary observations rather than direct cross-modal validation.

After quality control, 16,132 nuclei were retained (11,587 control and 4,545 epilepsy nuclei). Oligodendrocyte-lineage representation was lower in epilepsy than in controls (5.3% vs. 14.7%; exploratory P = 0.023), whereas layer 2/3 (L2/3) excitatory neurons showed higher representation (47.7% vs. 29.5%). Exploratory cell-level enrichment analysis identified MAPK-related signaling as the top-ranked KEGG pathway among upregulated genes in L2/3 neurons, and SCENIC identified a STAT3 (signal transducer and activator of transcription 3)-associated reactive astrocytic regulon pattern. Oligodendrocyte-lineage nuclei showed altered gene-expression patterns and different distributions along an inferred pseudotime continuum; these observations were not interpreted as direct evidence of impaired differentiation. In the independent imaging cohort, PET signal was lower in epileptic lesions than in contralateral regions (P = 0.007), and fractional anisotropy (FA) was reduced (P < 0.001), whereas mean diffusivity (MD) was not significantly different (P = 0.101). PET and FA findings remained significant after Bonferroni correction for three prespecified imaging comparisons (adjusted threshold P < 0.0167).

The study identifies exploratory cell-type-specific transcriptional alterations in the public PFC dataset and, independently, metabolic and white matter microstructural abnormalities in a separate neocortical epilepsy cohort. These findings represent complementary observations across different biological scales rather than direct molecular-imaging validation. Larger donor-level transcriptomic studies, matched molecular-imaging cohorts, and functional experiments are required to validate the candidate cellular and molecular signatures.