Abstracts

Rationale: Focal epilepsies comprise the majority of epilepsy cases, and focal cortical dysplasia (FCD) is a common cause, occurring in 5-25% of these patients. Unfortunately, many experimental barriers hinder FCD gene discovery. Unlike germline epilepsy mutations that can be sequenced from blood samples, FCD mutations are often somatic in nature, and, therefore, only occur within the cortical malformation. When samples are available, the mutations are still only found in a small percentage (1-6%) of cells from surgical resections. Therefore, identifying novel genes is extremely challenging and may not be possible in many cases using patient tissues. Currently, ~30% of FCD samples have identified mutations, and these are found exclusively in a few key genes in the mTOR pathway (mTOR, TSC1/2, and GATOR1 complex genes). Most of these discoveries have occurred due to hypothesis-driven targeted sequencing. Therefore, by combining several recent technological advances, we are developing a platform for identifying FCD genes in vitro using an unbiased inhibitory CRISPR library screen. Methods: To perform the library screen, we reprogrammed human iPSC lines stably expressing both KRAB-dCAS9 and Neurogenin (NGN) 1,2 genes under control of a DOX-inducible promoter and containing either +/+ or +/- DEPDC5 genotypes. The inducible expression of NGN1,2 allows for the efficient, uniform differentiation of human excitatory cortical neurons in culture. The KRAB-dCAS9 inhibits the expression of genes targeted by the guide RNA (gRNA) sequence. Using a previously published lentiviral CRISPR inhibitory guide RNA (gRNA) library containing 5 unique sequences for each human gene (>100,000 total gRNAs), we are able to randomly inhibit a single gene in individual clones of developing neurons. After treatment with doxycycline, we will sort for mature neurons with elevated phospho-S6 and increased soma size, known FCD type II markers. DNA will be collected and sequenced to identify enrichment for specific gRNAs. Recently, germline DEPDC5 gene mutations were found to cause FCD type II. We hypothesize that “second-hit” somatic mutations lead to these FCD lesions. The heterozygous DEPDC5 knockout should increase our ability to identify “second-hit” genes that may not be sufficient to generate FCD alone. Results: Since ~30% of FCD causative genes are thought to be known, we expect to identify most or all of these known genes in our screen. In fact, we will first perform a small test screen with positive (FCD-causing) and negative control (scrambled and non-essential) genes. For the large screen, we will use a one-sided Kolmogorov-Smirnov test to determine statistically significant candidate genes and confirm by individual Cas9 inhibition experiments. Furthermore, we will in utero electroporate CRISPR plasmids for top candidate genes into the developing rodent brain to identify FCD-like lesions and seizure induction. Conclusions: Validated FCD gene candidates discovered in this study can be used for targeted resequencing in resected patient brain tissues, which may overcome the barrier of mosaicism that precludes the use of whole exome or whole genome sequencing for gene discovery. Defining the spectrum of FCD-causing genes may also open the door to novel druggable targets. Funding: CURE Innovator Award