GENOMIC AND REGULATORY NON-CODING ALTERATIONS IN DRUG-RESISTANT EPILEPSY

Drug-resistant epilepsy (DRE) presents a major clinical challenge, with marked heterogeneity and persistent treatment failure, suggesting molecular dysregulation. Structural genomic variations and non-coding RNAs, which can exert broad downstream effects on gene dosage and regulatory networks, remain incompletely characterized in DRE.
This study investigates DRE-associated genomic and transcriptomic alterations in patient cohorts (n=341). Genome-wide copy number variants (CNVs) and single nucleotide polymorphisms (SNPs) were assessed from an existing whole-genome sequencing dataset comprising patients with drug-resistant and drug-sensitive epilepsy. Recurrent DRE-enriched exonic and noncoding-CNVs were identified, and selected variants were validated using TaqMan-based assays. Two large variants (9q11-13 loss and 16p12.2-p11.2 gain) showed significant associations with seizure frequency and age of onset. While DRE-specific exonic-SNPs showed limited association with epilepsy, noncoding-SNPs were enriched in regulatory regions.
Long non-coding RNA (lncRNA) dysregulation was examined in brain tissue (DRE patients and non-epileptic autopsy controls) and peripheral blood (DRE, drug-sensitive epilepsy, and healthy controls) using candidate-based quantitative-PCR alongside whole-transcriptome and small-RNA-sequencing. Several lncRNAs, including UCA1, H19, NEAT1, MALAT1, were dysregulated in brain tissue and showed clinicopathological associations with seizure frequency and age of onset. In peripheral blood, a subset of lncRNAs was dysregulated, with the antisense lncRNA BDNF-AS showing moderate diagnostic potential. Based on expression patterns and clinicopathological relevance, UCA1 was knocked down in differentiated SH-SY5Y cells to assess its functional impact on epilepsy-relevant molecular pathways.
Together, these findings highlight the contribution of structural genomic variations and non-coding RNAs to epileptogenesis and pharmacoresistance in DRE, with implications for biomarker discovery and therapeutic target identification.