Neurodevelopmental disorder due to a frameshift mutation in the GRIN2A gene: a case report

Introduction

The GRIN genes (GRIN1, GRIN2A-D, GRIN3A-B) encode the GluN subunits (GluN1, GluN2A-D, and GluN3A-B), which assemble into a heterotetrameric N-methyl-D-aspartate receptor (NMDAR) composed of two GluN1 subunits and two additional GluN2 or GluN3 subunits (1-3). The distribution of these subunits is tightly regulated in a spatiotemporal manner. NMDARs are critical for excitatory neurotransmission, neuronal development, and synaptic plasticity, all of which are fundamental to learning and memory (4). Moreover, specific isoforms of the NMDAR have been implicated in the pathogenesis of various neurological disorders. The clinical manifestations of these neurodevelopmental disorders are highly variable and may include intellectual disability (ID), epilepsy, speech and language impairments, autism spectrum disorders (ASDs), movement and sleep disorders, and gastrointestinal issues (5-7). Mutations in GRIN2A have been associated with a variety of epilepsy phenotypes, including electrical status epilepticus during slow-wave sleep (ESES), epileptic encephalopathy, Landau-Kleffner syndrome, and rolandic epilepsy (8).

A previous study has examined 248 cases with or at risk for GRIN2A pathogenic variants, which are associated with more discrete phenotypes characterized by a range of epilepsy disorders, including Landau-Kleffner syndrome and benign childhood epilepsy (9). These disorders are often associated with ID or developmental delay (9). Current research on the GRIN2A gene aims to clarify the correlation between mutation patterns and the spectrum of disease phenotypes. In addition, the study includes relevant treatment options. The specific treatment depends on the location of the gene, the type of mutation, and the clinical phenotype of the patient.

An increasing number of studies on GRIN2A gene mutations have enhanced our understanding of the underlying pathological mechanisms (10-12). This case report presents a patient with a potentially pathogenic GRIN2A mutation who initially exhibited ataxia without seizures or epileptiform discharges on electroencephalogram (EEG). The patient’s condition improved with immunotherapy alone, without the addition of antiepileptic drugs. To our knowledge, no previous reports have described GRIN2A-associated disease presenting with ataxia and EEG abnormalities in the absence of seizures. A genome-wide assay identified a potential disease-causing mutation in the GRIN2A, expanding the phenotypic spectrum of the disease and suggesting new therapeutic approaches for this disorder. We present this case in accordance with the CARE reporting checklist (available at https://tp.amegroups.com/article/view/10.21037/tp-2025-93/rc).

Case presentation

A 23-month-old boy presented with a two-month history of predominantly unsteady gait, without slurred speech or upper limb dysmetria. His prenatal and perinatal history was unremarkable. The patient’s mother reported a history of childhood seizures of unspecified type and frequency; she has been seizure-free since the age of 10 months, and currently exhibits mild speech dysfluency without signs of ataxia. The father and grandparents were healthy with no significant medical histories. No dysmorphic features were observed in the patient.

Upon initial admission, comprehensive neurological examination, cerebrospinal fluid (CSF) analysis, and brain magnetic resonance imaging (MRI) were all normal, effectively ruling out acute or chronic inflammatory processes. Based on the clinical presentation and available ancillary findings, a presumptive diagnosis of immune-mediated cerebellar ataxia was made. EEG was not performed, which we acknowledge as a limitation. Screening for paraneoplastic and autoimmune encephalitis antibodies was negative, and metabolic evaluations were unremarkable. Notably, the patient had experienced a viral respiratory infection approximately one week prior to the onset of ataxia, supporting a diagnosis of post-infectious immune-related ataxia. High-dose corticosteroids and intravenous immunoglobulin (IVIG) were administered with parental consent, resulting in improvement in ataxia. Upon discharge, a tapering course of oral prednisone was prescribed for six weeks.

At this stage, the patient’s ataxia primarily manifested as cerebellar gait disturbance with preserved speech and fine motor coordination, suggesting a relatively localized dysfunction.

Approximately five months later (early April 2024), the patient developed recurrent ataxia following another viral respiratory infection. Despite worsening symptoms—persistent ataxia, speech regression, and drooling—the parents declined hospital readmission. Instead, the patient was evaluated at external medical centers. Respiratory pathogen screening, cranial and spinal MRI, and video EEG were performed; no definitive pathogen was identified, and no new immunotherapy or antiepileptic therapy was initiated. The patient continued oral prednisone until returning to our institution.

Compared with the initial episode, the recurrent ataxia was more severe. In addition to exacerbated gait instability, the patient showed marked speech regression, persistent drooling, and developmental delays, suggesting progression from isolated cerebellar to widespread cortical and subcortical involvement.

During the second hospitalization at our hospital, medical records indicated that the patient underwent lumbar puncture, video EEG, blood tests, and whole genome sequencing. He exhibited significant gross motor delay and moderate delays in adaptive behavior, fine motor skills, language, and social domains. No clinical seizures were observed. CSF analysis, brain and spinal MRI, and a comprehensive autoimmune encephalitis antibody panel—including anti-NMDAR, α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid receptor (subunits 1 and 2) (AMPAR1/2), gamma-aminobutyric acid B receptor (GABABR), leucine-rich glioma-inactivated 1 (LGI1), contactin-associated protein-like 2 (CASPR2), glutamic acid decarboxylase 65 (GAD65), myelin oligodendrocyte glycoprotein (MOG), and oligoclonal bands (OCB)—were all normal.

Pretreatment video EEG showed bilateral moderate-amplitude 4–5 Hz theta and 3 Hz delta activity, with prolonged discharges of high- to very-high-amplitude 1.5–2.5 Hz spike-slow wave complexes. Diffuse poly spike-slow wave discharges were recorded across all leads, with interlead asymmetry most prominent in the frontal, parietal, occipital, and posterior temporal regions, especially during non-rapid eye movement (NREM) sleep. These findings were consistent with ESES, with a spike-wave index (SWI) exceeding 85% during NREM sleep (Figure 1).

Figure 1 EEG findings in a patient with a GRIN2A mutation. (A) Waking EEG before immunotherapy shows incompletely synchronized sharp-slow wave and spike–slow wave complexes across all leads. (B) Sleep EEG before immunotherapy reveals widespread asynchronous sharp-slow and spike-slow wave discharges. (C) Waking EEG after immunotherapy demonstrates high-amplitude slow waves over the bilateral frontal poles and anterior temporal regions. (D) Sleep EEG after immunotherapy shows sharp-slow wave discharges predominantly over the right central, parietal, occipital, and temporal regions, including the parietal-midline. EEG, electroencephalographic.

Genetic analysis revealed a heterozygous frameshift variant in the GRIN2A gene (c.1717delG, p.Val573Phefs*16 [NM_001134407.3]), a novel mutation not previously reported in the literature (Figure 2). Sanger sequencing confirmed maternal heterozygosity and paternal wild-type genotype. According to the American College of Medical Genetics and Genomics (ACMG) guidelines, the variant was classified as likely pathogenic. The mutation is predicted to result in nonsense-mediated mRNA (messenger ribonucleic acid) decay, leading to loss of functional GRIN2A protein.

Figure 2 Illustrates the GRIN2A mutation. (A) This shows the pedigree structure of the patient and both parents. The square symbolizes a male, while the circle represents a female. (B) A map illustrating the Sanger sequencing of GRIN2A reveals a guanine deletion at base 1717, which results in the substitution of valine by phenylalanine at position 573. Bases in green represent the reference (wild-type) sequences, while the gray boxes highlight the mutant segment, indicating the mutation at the locus (G→C change).

Based on the genetic findings, clinical phenotype, and ancillary investigations, a diagnosis of GRIN2A-related disorder was established. The patient was treated with high-dose corticosteroids (20 mg/kg) and IVIG (2 g/kg). During corticosteroid tapering, notable clinical improvement was observed, including resolution of drooling, improved gait stability, and increased vocalization. However, speech output remained limited to simple words or superlative expressions.

A review of the video EEG conducted before discharge showed slow fundamental waves in both cerebral hemispheres, including medium-high amplitude delta waves (2–3 Hz) and occasional low-moderate amplitude theta waves (4–7 Hz). Low-to-moderate amplitude sharp and spiky wave bursts (10–15 Hz) were observed in the right central, parietal, mesial, and posterior temporal regions, as well as in the parietal midline, during both waking and sleeping phases. Additionally, 2–5 Hz wave bursts were noted in the bilateral frontal poles and the prefrontal regions.

During follow-up, the patient continued oral prednisone without recurrence of ataxia. He achieved independent ambulation and cessation of drooling; however, expressive language development remained severely delayed. No clinical seizures were observed throughout the follow-up period.

The patient’s clinical evolution, diagnostic findings, and therapeutic interventions are summarized in Table 1.

This study was approved by the Ethics Committee of Children’s Hospital of Soochow University. All procedures performed in this study were in accordance with the ethical standards of the institutional and/or national research committee(s) and with the Declaration of Helsinki and its subsequent amendments. Written informed consent was obtained from the patient’s legal guardians for the publication of this case report and any accompanying images. All data were analyzed anonymously. A copy of the written consent is available for review by the editorial office of this journal.

Whole genome sequencing

The whole genome was sequenced using a high-throughput sequencing platform and the TruSeq Library Construction Kit. The paired-end sequencing approach was used. The original sequencing FASTQ file was processed with Cutadapt software to remove junction primer sequences, low-quality bases (Q<30), reads of more than 5 N bases, and sequences less than or equal to 35 bp after quality control. The Sentieon tool used the integrated accelerated BWA module for mass conversion. The reference genome (GRCh38), including the mitochondrial NC_012920.1 reference sequence, was compared to the quality-controlled FASTQ data. The Sentieon tool performed genome-wide detection of single nucleotide variants (SNVs) and insertion-deletion mutations (InDels), followed by the GATK process for filtering and screening, resulting in Variant Call Format (VCF) files. Copy number variation (CNV) variant detection was performed across the genome using the CNVkit process. Structural variations in the samples were detected using Manta software, and annotations were provided using Annovar software, combined with open-source data for comprehensive analysis.

In silico structural modeling

To assess the structural impact of the GRIN2A mutation (c.1717delG, p.Val573Phefs*16), protein structure prediction was conducted using AlphaFold2 (DeepMind, 2021) via the ColabFold notebook (Mirdita et al., Nat Methods, 2022) with default parameters. The input sequences included both wild-type and mutant GRIN2A isoforms. Structural confidence was evaluated with the predicted Local Distance Difference Test (pLDDT) score, and visualization was done in PyMOL (v2.5, Schrödinger, LLC). The wild-type protein showed a complete, high-confidence domain structure, including the ligand-binding and transmembrane regions, while the mutant protein was truncated at residue 589, missing all transmembrane domains and displaying widespread low-confidence, disordered regions, indicative of a loss-of-function phenotype.

Discussion

This case illustrates the progressive nature of neurological symptoms associated with GRIN2A-related disorders, initially presenting as isolated cerebellar ataxia and subsequently evolving into a complex phenotype involving speech regression, persistent drooling, and ESES. This clinical evolution suggests broader cortical and subcortical involvement beyond the cerebellum. Early-onset ataxia may serve as an early indicator of an evolving neurodevelopmental disorder, emphasizing the need for prompt and comprehensive evaluation.

Given the patient’s clinical course, it is crucial to consider the molecular mechanisms underlying GRIN2A mutations and their relevance to phenotypic expression and therapeutic responsiveness.

Emerging evidence highlights the spatiotemporal regulation of GRIN2A expression during early development, particularly in the cerebellum. Rodent models demonstrate minimal prenatal GluN2A expression, followed by a sharp postnatal increase during the second postnatal week, coinciding with critical periods of synaptogenesis and gait acquisition (11,12). Similarly, human single-cell and bulk transcriptomic analyses reveal a progressive rise in GRIN2A mRNA levels within cerebellar Purkinje and granule cells from late gestation through the first year of life (13,14). In contrast, forebrain regions such as the cortex exhibit a more gradual GRIN2A expression trajectory, extending into later childhood, correlating with the delayed onset of cortical features such as seizures and ESES (10). Positron emission tomography (PET) studies further support this pattern, revealing relatively higher cerebellar NMDAR density during infancy compared to other brain regions (15). These spatiotemporal dynamics likely account for the early cerebellar vulnerability observed in our patient and suggest that subsequent cortical involvement underlies later electrophysiological and developmental abnormalities.

This report describes, for the first time, a patient harboring a potentially pathogenic GRIN2A frameshift mutation, c.1717delG (p.Val573Phefs*16) [NM_001134407.3]. The c.1717delG (p.Val573Phefs*16) frameshift mutation introduces a premature stop codon, predicted to trigger nonsense-mediated mRNA decay (NMD). NMD is a cellular quality control mechanism that degrades mRNAs containing premature termination codons, thereby preventing the synthesis of truncated and potentially deleterious proteins. In this context, NMD likely results in haploinsufficiency of GRIN2A, contributing to the patient’s clinical manifestations. Sanger sequencing confirmed that the patient and his mother carried the same GRIN2A mutation. Although a detailed seizure history from the mother was unavailable, she had experienced childhood seizures and later exhibited only mild speech dysfluency without ataxia. This phenotypic variability is consistent with previous reports demonstrating that GRIN2A mutations can manifest with a wide range of clinical severities, even among individuals within the same family.

GRIN2A mutations are most commonly associated with speech impairment and epilepsy (9,15-17). However, the unique presentation in our patient—initial ataxia without seizures—necessitated extensive diagnostic workup before confirming the genetic etiology. The GRIN2A gene encodes the GluN2A subunit of the NMDAR (18), and gait abnormalities in this context may resemble those observed in anti-NMDAR encephalitis, where NMDAR hypofunction is implicated (19). Pathogenic missense variants in the transmembrane and linker structural domains (misTMD + linker) are associated with more severe developmental phenotypes (7,13), while mutations affecting the amino-terminal or ligand-binding domains (misATD + LBD) or loss-of-function mutations typically result in milder manifestations (9,20,21). Notably, the frameshift mutation in this case localizes to the transmembrane domain, supporting a potential severe impact on receptor assembly and function.

To further elucidate the pathogenic mechanism, structural modeling using AlphaFold2 was performed. The wild-type GRIN2A protein exhibited a well-folded structure encompassing the ligand-binding and transmembrane domains, both critical for receptor assembly and ion channel function. In contrast, the mutant protein was predicted to be truncated before the TMD and showed extensive structural disorder with low confidence scores, consistent with a loss-of-function effect (Figure 3) (10-12). Recent studies indicate that disruption of the TMD severely impairs synaptic localization and receptor gating (12,13), supporting our loss-of-function hypothesis and explaining the early cerebellar involvement prior to the appearance of cortical symptoms.

Figure 3 Comparative AlphaFold2 structural modeling of wild-type and mutant GRIN2A proteins. (A) The predicted structure of wild-type GRIN2A shows a complete domain architecture, including the ATD, LBD, TMD, and CTD. These domains are well-folded and display high per-residue confidence (pLDDT), represented by predominantly blue coloring. (B) The mutant GRIN2A (p.Val573Phefs*16) structure is truncated following residue 573, resulting in the complete loss of the TMD and CTD. The remaining structure is largely disordered and exhibits low confidence scores (orange-red coloring), consistent with impaired folding and likely degradation or mislocalization. The predicted truncation point is indicated. Domain boundaries are based on UniProt annotation: ATD (residues 1–400), LBD (401–600), TMD (601–800), CTD (801–994). ATD, amino-terminal domain; CTD, cytoplasmic tail domain; LBD, ligand-binding domain; pLDDT, predicted Local Distance Difference Test; TMD, transmembrane domain.

The mechanism by which immunotherapy ameliorated the patient’s symptoms remains incompletely understood. It is plausible that corticosteroids and IVIG exerted effects through modulation of the immune microenvironment, rather than directly restoring mutant NMDAR function. IVIG has been shown to alter Fc receptor signaling and cytokine profiles, reducing secondary immune-mediated neuronal stress (14). Similarly, corticosteroids can suppress microglial activation and proinflammatory pathways, mitigating excitotoxicity and maladaptive synaptic plasticity (11). Given that the frameshift mutation likely leads to haploinsufficiency via nonsense-mediated mRNA decay, immunotherapy would not restore receptor expression but might stabilize neuronal networks and reduce excitability. This potential convergence of genetic susceptibility and immune-mediated modulation merits further investigation.

Although treatment with corticosteroids and IVIG resulted in clinical and electrographic improvements, the persistent developmental delay and risk of relapse highlight the underlying genetic vulnerability. Interestingly, in previously reported cases of GRIN-related developmental and epileptic encephalopathies, immunotherapy led to electrographic improvement in four out of five patients and EEG normalization in three (22). There is increasing evidence that autoimmunity may overlap with genetic epilepsies, as anti-synaptic receptor antibodies, including anti-NMDAR, have been identified in various epilepsy syndromes (23,24).

Our patient was not treated with antiepileptic drugs, yet immunotherapy alone led to substantial clinical recovery and EEG improvement, supporting a role for immunomodulatory strategies in select cases of GRIN2A-related disorders. It is important to note that ESES may arise from various etiologies, including structural brain malformations, autoimmune processes, and genetic mutations. In this case, comprehensive evaluations—including CSF studies, metabolic screening, autoimmune panels, and neuroimaging—excluded infectious, inflammatory, and metabolic causes, supporting a genetic basis for the observed phenotype.

This case highlights the potential interplay between genetic predisposition and immune-mediated processes in early-onset epileptic encephalopathies. There are limitations to this study. Although immunotherapy yielded significant short-term benefits, the relatively short follow-up period prevents definitive conclusions regarding future seizure development. The patient remains under close surveillance.

In conclusion, this case expands the phenotypic spectrum associated with GRIN2A-related disorders. The identification of a novel pathogenic frameshift variant, combined with a positive clinical response to immunotherapy, underscores the importance of integrating genetic and immune considerations when evaluating atypical neurodevelopmental disorders. Future studies are warranted to elucidate the correlation between GRIN2A mutation type, clinical phenotype, and treatment response, ultimately guiding precision medicine approaches in this emerging field.

Conclusions

In this case, whole genome sequencing combined with EEG findings identified a GRIN2A frameshift mutation, providing a definitive diagnosis and shedding light on GRIN2A-related epileptic encephalopathy. Although functional assays were not performed, immunotherapy targeting presumed NMDAR dysfunction resulted in significant clinical improvement without adverse effects. This case underscores the value of early genetic testing and suggests that GRIN2A-targeted interventions may offer promising therapeutic avenues.

Acknowledgments

The authors thank the patient mentioned in the article and his family for the clinical information they provided and their contribution to scientific research.

Footnote

Reporting Checklist: The authors have completed the CARE reporting checklist. Available at https://tp.amegroups.com/article/view/10.21037/tp-2025-93/rc

Peer Review File: Available at https://tp.amegroups.com/article/view/10.21037/tp-2025-93/prf

Funding: This work was supported by the Science and Technology Development Plan of Suzhou (grant No. SKY2022007).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://tp.amegroups.com/article/view/10.21037/tp-2025-93/coif). All authors report funding received from the Science and Technology Development Plan of Suzhou (grant No. SKY2022007) for this research. The authors have no other conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. This study was approved by the Ethics Committee of Children’s Hospital of Soochow University. All procedures performed in this study were in accordance with the ethical standards of the institutional and/or national research committee(s) and with the Declaration of Helsinki and its subsequent amendments. Written informed consent was obtained from the patient’s legal guardians for the publication of this case report and any accompanying images. All data were analyzed anonymously. A copy of the written consent is available for review by the editorial office of this journal.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Santos-Gómez A, Miguez-Cabello F, Juliá-Palacios N, et al. Paradigmatic De Novo GRIN1 Variants Recapitulate Pathophysiological Mechanisms Underlying GRIN1-Related Disorder Clinical Spectrum. Int J Mol Sci 2021;22:12656. [Crossref] [PubMed]
2. Michalski K, Furukawa H. Structure and function of GluN1-3A NMDA receptor excitatory glycine receptor channel. Sci Adv 2024;10:eadl5952. [Crossref] [PubMed]
3. Chou TH, Epstein M, Fritzemeier RG, et al. Molecular mechanism of ligand gating and opening of NMDA receptor. Nature 2024;632:209-17. [Crossref] [PubMed]
4. Banke TG, Traynelis SF, Barria A. Early expression of GluN2A-containing NMDA receptors in a model of fragile X syndrome. J Neurophysiol 2024;131:768-77. [Crossref] [PubMed]
5. Lesca G, Rudolf G, Bruneau N, et al. GRIN2A mutations in acquired epileptic aphasia and related childhood focal epilepsies and encephalopathies with speech and language dysfunction. Nat Genet 2013;45:1061-6. [Crossref] [PubMed]
6. Kingwell K. Epilepsy: GRIN2A mutations identified as key genetic drivers of epilepsy-aphasia spectrum disorders. Nat Rev Neurol 2013;9:541. [Crossref] [PubMed]
7. Yuan H, Hansen KB, Zhang J, et al. Functional analysis of a de novo GRIN2A missense mutation associated with early-onset epileptic encephalopathy. Nat Commun 2014;5:3251. [Crossref] [PubMed]
8. Li X, Xie LL, Han W, et al. Clinical Forms and GRIN2A Genotype of Severe End of Epileptic-Aphasia Spectrum Disorder. Front Pediatr 2020;8:574803. [Crossref] [PubMed]
9. Strehlow V, Heyne HO, Vlaskamp DRM, et al. GRIN2A-related disorders: genotype and functional consequence predict phenotype. Brain 2019;142:80-92. [Crossref] [PubMed]
10. Samanta D. GRIN2A-related epilepsy and speech disorders: A comprehensive overview with a focus on the role of precision therapeutics. Epilepsy Res 2023;189:107065. [Crossref] [PubMed]
11. Farsi Z, Nicolella A, Simmons SK, et al. Brain-region-specific changes in neurons and glia and dysregulation of dopamine signaling in Grin2a mutant mice. Neuron 2023;111:3378-3396.e9. [Crossref] [PubMed]
12. Camp CR, Vlachos A, Klöckner C, et al. Loss of Grin2a causes a transient delay in the electrophysiological maturation of hippocampal parvalbumin interneurons. Commun Biol 2023;6:952. [Crossref] [PubMed]
13. Xu Y, Song R, Perszyk RE, et al. De novo GRIN variants in M3 helix associated with neurological disorders control channel gating of NMDA receptor. Cell Mol Life Sci 2024;81:153. [Crossref] [PubMed]
14. Juliá-Palacios N, Olivella M, Sigatullina Bondarenko M, et al. L-serine treatment in patients with GRIN-related encephalopathy: a phase 2A, non-randomized study. Brain 2024;147:1653-66. [Crossref] [PubMed]
15. Mangano GD, Riva A, Fontana A, et al. De novo GRIN2A variants associated with epilepsy and autism and literature review. Epilepsy Behav 2022;129:108604. [Crossref] [PubMed]
16. Turner SJ, Mayes AK, Verhoeven A, et al. GRIN2A: an aptly named gene for speech dysfunction. Neurology 2015;84:586-93. [Crossref] [PubMed]
17. Yang X, Qian P, Xu X, et al. GRIN2A mutations in epilepsy-aphasia spectrum disorders. Brain Dev 2018;40:205-10. [Crossref] [PubMed]
18. DeVries SP, Patel AD. Two patients with a GRIN2A mutation and childhood-onset epilepsy. Pediatr Neurol 2013;49:482-5. [Crossref] [PubMed]
19. Yeshokumar AK, Sun LR, Klein JL, et al. Gait Disturbance as the Presenting Symptom in Young Children With Anti-NMDA Receptor Encephalitis. Pediatrics 2016;138:e20160901. [Crossref] [PubMed]
20. Xie L, McDaniel MJ, Perszyk RE, et al. Functional effects of disease-associated variants reveal that the S1-M1 linker of the NMDA receptor critically controls channel opening. Cell Mol Life Sci 2023;80:110. [Crossref] [PubMed]
21. Fernández-Marmiesse A, Kusumoto H, Rekarte S, et al. A novel missense mutation in GRIN2A causes a nonepileptic neurodevelopmental disorder. Mov Disord 2018;33:992-9. [Crossref] [PubMed]
22. Hausman-Kedem M, Menascu S, Greenstein Y, et al. Immunotherapy for GRIN2A and GRIN2D-related epileptic encephalopathy. Epilepsy Res 2020;163:106325. [Crossref] [PubMed]
23. Marwick KFM, Hansen KB, Skehel PA, et al. Functional assessment of triheteromeric NMDA receptors containing a human variant associated with epilepsy. J Physiol 2019;597:1691-704. [Crossref] [PubMed]
24. Chen X, Keramidas A, Harvey RJ, et al. Effects of GluN2A and GluN2B gain-of-function epilepsy mutations on synaptic currents mediated by diheteromeric and triheteromeric NMDA receptors. Neurobiol Dis 2020;140:104850. [Crossref] [PubMed]