Abstracts

Rationale: Loss-of-function mutations in synaptic genes can lead to epileptic disorders that are poorly managed by current approved antiepileptic drugs. Haploinsufficiency of STXBP1 leads to developmental and epileptic encephalopathy 4 (DEE4), which is characterized by early onset epilepsy that is refractory to standard anti-epileptics in approximately a quarter of patients. Similarly, haploinsufficiency of SYNGAP1 causes developmental and epileptic encephalopathy, and approximately half of the patients with loss-of-function mutations in this gene respond poorly to standard antiepileptics. To identify potential therapeutics for these patient populations, we generated human neuronal models of STXBP1 and SYNGAP1 genetic disruption, characterized their synaptic behavior and designed antisense oligonucleotides (ASOs) to boost levels of the target proteins.

Methods: We used our BRITE^TM system, which integrates patient-derived neuronal models, multi-omics characterization of neuronal cell types, and machine-learning based analytics to uncover disease relevant cellular phenotypes and assess therapeutic candidates. CRISPR/Cas9 was used to generate >50 cell lines used for cellular phenotyping of neurons associated with DEE4 and SYNGAP1-related developmental and epileptic encephalopathy. Neuronal synaptic transmission was characterized using all-optical electrophysiology, which characterizes >500,000 neurons per day with single cell resolution. In parallel, we designed ASOs targeting regulatory sites in STXBP1 or SYNGAP1 transcripts predicted to ultimately boost levels of their respective protein products. We treated human iPS cell-derived cortical excitatory neurons with these ASOs and used immunoblotting to evaluate levels of the target proteins.

Results: We found that human neurons lacking STXBP1 showed a complete loss of synaptic transmission, while neurons heterozygous for the functional allele differ from wild-type cells in their evoked synaptic potential decay. Phenotypes were rescued by reintroduction of STXBP1 with lentivirus. SYNGAP1 mutant neurons show a faster decay time, slightly less postsynaptic potential area and an earlier amplitude peak time than neurons generated from control cell lines. These synaptic phenotypes provide a human neuronal model to evaluate the potential of therapeutic candidates to rescue neuronal behavior in a disease context. In addition, we identified several ASOs that increase levels of STXBP1 or SYNGAP1 protein up to 60% in human iPSC-derived neurons.

Conclusions: Here we demonstrate the application of human cellular models, high-throughput electrophysiology and ASO technology to epilepsies caused by mutations in key synaptic genes. These molecules are promising candidates to correct deficiency of these key proteins in the CNS. We are currently optimizing ASO lead candidates for in vivo tolerability screening and functional rescue assessment with the goal of bringing ASO therapeutics to patients with genetically-driven epileptic synaptopathies.

Funding: NIH R44MH112273