Abstracts

Rationale: N-methyl-D-aspartate receptors (NMDARs), ionotropic glutamate receptors, are heterotetramers comprised of two GluN1 subunits (one gene, GRIN1) and two GluN2 subunits (four genes, GRIN2A-D). These receptors are vital for normal brain function, and when mutated, can contribute to a host of severe neurological diseases, including epilepsy. Recently, whole exome sequencing has identified GRIN2A mutations to play a causal role in the development of epileptic encephalopathy. Here, we investigate the functional effects, both in vitro and in vivo, of a de novo GRIN2A gain-of-function missense variant (c.1930 A>G; p.Ser644Gly, S644G) identified in a child with epilepsy, developmental delay, absence of speech, and intellectual disability.  Methods: Electrophysiology was conducted in heterologous expression systems to determine receptor kinetics and pharmacology. Additionally, in collaboration with the Jackson Laboratory Center for Precision Genetics, a Grin2aS644G (S644G) knock-in mouse model was made. Hippocampal slices from these S644G mice were evaluated for synaptic decay time course of the NMDAR-mediated current and cultured cortical pyramidal cells were subjected to network burst firing assays using multielectrode arrays (MEAs). In vivo animal testing included assays for seizure threshold and various behavioral tasks such as acoustic startle, ambulation, and repetitive behaviors.  Results: Two-electrode voltage clamp recordings on oocytes revealed that the S644G variant enhanced glutamate and glycine potency by 17-fold (EC50 of 0.18 µM vs 3.0 µM of wild type, WT; p<0.001) and 10-fold (EC50 of 0.08 µM vs 0.86 µM of WT; p<0.001). Patch clamp recordings of transfected HEK cells containing 0 (wildtype GluN2A), 1 (heterozygous S644G), or 2 (homozygous S644G) copies of the GluN2A-S644G subunit suggests significantly prolonged deactivation time course in a gene-dependent manner. CA1 pyramidal cells show a significantly prolonged NMDAR-mediated synaptic decay in heterozygous S644G mice compared to WT littermates (150 ms vs. 104 ms of WT; p=0.000045). MEA analysis from cultured cortical neurons indicated both mutant heterozygous and homozygous networks displayed significantly increased bursting kinetics compared to WT controls. Finally, homozygous S644G mice showed severe spontaneous seizures and died between postnatal day (PND) 15 and PND17, while heterozygous mice had abnormal seizure threshold. Heterozygous S644G mice also exhibited increased ambulation, decreased acoustic startle response, and an increase in repetitive behaviors. Application of FDA-approved, NMDAR-specific drugs showed that Nuedexta (a combination of quinidine and dextromethorphan) to be the most effective at delaying the onset of lethal seizures in homozygous S644G mice.  Conclusions: Taken together, these data indicate that the GluN2A-S644G mutation is a strong gain-of-function NMDAR variant. The S644G mouse model was successful at recapitulating both seizure and secondary symptom phenotypes. Our data suggest that Nuedexta may be the most beneficial FDA-approved drug for this patient that is specific for NMDARs. However, the recovery from secondary symptoms with Nuedexta is unknown, but also unlikely. Given the role of NMDARs in neuronal development and synaptic plasticity, this mutation has likely altered circuit formation throughout the brain. Future studies will center around defining a critical window for therapeutic intervention and fully understanding how this genetic mutation rewires the brain to drive epileptic encephalopathy.  Funding: No funding