Abstracts

Rationale: Long-term epilepsy-associated tumors (LEATs) presents with a long history of drug-resistant epilepsy in early pediatric patients. LEATs had frequent mutations in BRAF (Nat. Rev. Neurol. 2016; 12(12); 732-740), newly implying intrinsic epileptogenicity to contribute to drug-resistant epilepsy besides tumorigenesis. In spite of strong association between epilepsy and LEAT, there was no study on animal model and causal relationship about intrinsic molecular epileptogenic mechanism of LEATs. Interestingly, our group reported that PIK3CA, MTOR, TSC1, or TSC2 tumor driver mutations directly cause intractable epilepsy in focal cortical malformation patients (Nat. Med. 2015; 21(4); 395-400). Thus, we examined whether the oncogenic mutation specific to LEATs serve as the epileptogenic mutation during the brain development and identified the molecular mechanism of how such an oncogenic mutation causes to the epileptogenicity. Methods: We examined BRAF somatic mutation in LEATs (n=733) compared to other brain tumors (n=5484) registered in COSMIC. To exclude other epilepsy-prone somatic mutations with low-allelic frequency, we analyzed deeply whole exome sequencing (~620X) using fresh frozen ganglioglioma (GG), a prominent example of LEAT, and paired blood samples (Fig.1 and Table.1). Then, laser capture microdissection (LCM) of patient's tissue was used to specifically localize mutation in different cell types. The mouse harboring BRAFV600E somatic mutation was made by in utero electroporation (IUE) in BRAFLSL-V600E mouse reflecting GG patients' clonal condition (n=30). Spike activity was confirmed by electroencephalography (EEG) and electrophysiology. Immunostaining and biochemical study characterized GG-like feature including dysplastic neuron, neoplastic glial cells, CD34 positivity and cortical dyslamination. BRAFV600E specific inhibitor was chronically infused into brain to reduce ictogenesis. Finally, RNA sequencing of BRAFV600E(+) GG tissue (n=3) was analyzed to explain molecular mechanism of epileptogenesis. Next, the dominant negative form of epileptogenic candidate was transfected to reduce seizure. Results: BRAFV600E mutation was frequent in LEATs and found as only one shared mutation in 3/5 (60%) GG in deep sequencing without no any other epilepsy related molecule. Interestingly, this was independently coexisted in neuron and glia, which implying mutation arising in neural progenitor during corticogenesis. Likewise, over 90% of our animal model gained somatic BRAF mutation uniquely at brain development led to intrinsic epileptogenic property derived from dysmorphic neuronal lineage cells (Fig.2), whereas the tumorigenic property was attributed to highly-proliferative GFAP+ and OLIG2+ glial lineage cells (n=23, p < 0.05). BRAFV600E affected tissue also had high CD34 positivity and cortical dyslamination (n=7, p < 0.005). Moreover, RNA-seq of patient’s tissue suggested that increased expression of RE1-silencing transcription factor (REST) was sustained and also remarkably shown in neuron in our mouse model (n=7, p < 0.001). Finally, we found that seizure was significantly alleviated by the intraventricular injection of clinically available BRAFV600E inhibitor as well as the dominant-negative of REST (n=7, p < 0.05). Conclusions: Therefore, this study provides the direct evidence that BRAF somatic mutations in developing neurons lead to the epileptogenicity and identifies BRAF and REST as a treatment target for intractable epilepsy in LEATs. Funding: 2015 Pediatric Epilepsies Award, CURE (Citizens United for Research on Epilepsy, USA)The Korean Health Technology R&D Project of the Ministry of Health & Welfare (H16C0415)