Whole-exome sequencing in an individual with severe global developmental delay and intractable epilepsy identifies a novel, de novo GRIN2A mutation



We present a 4-year-old girl with profound global developmental delay and refractory epilepsy characterized by multiple seizure types (partial complex with secondary generalization, tonic, myoclonic, and atypical absence). Her seizure semiology did not fit within a specific epileptic syndrome. Despite a broad metabolic and genetic workup, a diagnosis was not forthcoming. Whole-exome sequencing with a trio analysis (affected child compared to unaffected parents) was performed and identified a novel de novo missense mutation in GRIN2A, c.2449A>G, p.Met817Val, as the likely cause of the refractory epilepsy and global developmental delay. GRIN2A encodes a subunit of N-methyl-d-aspartate (NMDA) receptor that mediates excitatory transmission in the central nervous system. A significant reduction in the frequency and the duration of her seizures was observed after the addition of topiramate over a 10-month period. Further prospective studies in additional patients with mutations in GRIN2A will be required to optimize seizure management for this rare disorder. This report expands the current phenotype associated with GRIN2A mutations.

A PowerPoint slide summarizing this article is available for download in the Supporting Information section here.

Genetic testing for the epileptic encephalopathies has improved dramatically with the advent of next-generation sequencing.[1] A whole-exome approach (whole-exome sequencing [WES]) or the application of targeted panels of channelopathy or encephalopathy genes has been successful in identifying the molecular cause for many patients with rare epilepsies.[2, 3] A molecular diagnosis is a key step for a patient and their family, as it can end invasive diagnostic investigations, better inform management decisions, provide accurate recurrence risks for counseling, and may also suggest treatment options.

Mutations in GRIN2A cause epilepsy and intellectual disability (OMIM 613971).[4-6] There are many forms of epilepsy observed in those with GRIN2A mutations including epilepsy-aphasia spectrum disorders (benign epilepsy with centrotemporal spikes, Landau-Kleffner syndrome, epileptic encephalopathy with continuous spike and wave during sleep).[7-9] A recent study showed mutations in GRIN2A are responsible for 7.5% of idiopathic focal epilepsy.[9] The neurodevelopmental profile spans a broad spectrum and includes learning difficulties to moderate-to-severe intellectual impairment.[5]

Herein we present a case of severe epilepsy characterized by multiple generalized and focal seizures and severe global developmental delay with a de novo mutation in GRIN2A. Identification of the molecular defect provided insight into disease pathogenesis and, interestingly, the addition of topiramate resulted in a decrease in seizure frequency and duration and highlights the benefit of WES for diagnosis as well as potential management of rare epilepsy syndromes.

Patient and Methods

Case description

The patient initially presented at 14 months of age with a history of delayed development and low vision. Antenatal and perinatal histories were unremarkable. Parents were of French Canadian descent, nonconsanguineous, and had a healthy 4-year-old daughter. There was no family history of neurologic or metabolic diseases. At initial assessment, head circumference was at the 50th percentile; height and weight were at the 10th percentile. She had a distinctive appearance with broad forehead, plagiocephaly, epicanthal folds, a small nose with a low-hanging columella as well as malar flattening, a small mouth, and an underbite. She was nonverbal with vocalizations and interactions limited to cooing and inconsistent smiling. Audiology testing was normal. She had roving nystagmus and bilateral abduction deficits. A funduscopic examination was normal and confirmed on electroretinography, although flash visual-evoked potentials had reduced amplitudes bilaterally.

The patient was able to move limbs spontaneously but was not able to sit or roll independently. She had not developed any fine motor skills; she would inconsistently attempt to grab objects and could not transfer. In subsequent assessments over a 3-year period, her development progressed minimally without any episodes of regression.

Magnetic resonance imaging (MRI) performed at 6 months and 18 months demonstrated prominence of extraaxial cerebrospinal fluid (CSF) spaces with normal myelination and the corpus callosum appeared thin and slightly elongated. Electroencephalography (EEG) at 14 months, performed to evaluate myoclonic jerks, demonstrated diffuse slowing with background activity at 4–5 Hz with no electroclinical correlation. At 24 months of age, she developed (1) partial complex seizures with secondary generalization consisting of sudden head and eye deviation to the left followed by tonic–clonic movements of extremities for 1–2 min and (2) tonic seizures consisting of sudden tonic posturing of the arms, staring for 5–10 s followed by vocalization. Both seizure types occurred 3–4 times per day. EEG at this time demonstrated absence of the posterior dominant rhythm and diffuse background slowing at 3–4 Hz. Levetiracetam was initiated and increased to 70 mg/kg/day over the next 2 weeks. Seizure frequency increased and EEG demonstrated epileptiform discharges in the right frontal and frontocentral region exacerbated by drowsiness and sleep. Clonazepam was added at a dose of 0.03 mg/kg/day, and levetiracetam was gradually increased to 100 mg/kg/day.

Tonic seizures continued over the next year up to five times per day, each lasting up to 15 s and gradually increased in severity involving all four limbs. The partial complex seizures occurred 1–2 times per day. Clonazepam could not be increased due to sedative effects and therefore valproic acid was added to the regimen. Valproic acid did not decrease the frequency or duration of seizures and was therefore tapered after 5 months. At the age of 4 years, her epilepsy evolved. In addition to the partial complex seizures (2–3 episodes/day) and the tonic seizures (2–6 episodes/day), two new seizures types were present: nocturnal generalized seizures with shuddering and eye fluttering lasting 30 s, and brief myoclonic seizures associated with a cry. Clonazepam dose was increased to 0.1 mg/kg/day despite sedative effects.

The patient had a broad metabolic and genetic work-up that included muscle and skin biopsies but did not yield a diagnosis. She was enrolled in the FORGE (Finding of Rare Disease Genes) Canada project, a nationwide effort to identify novel disease genes. Institutional research ethics board (Children's Hospital of Eastern Ontario) approval of the FORGE project was obtained, and free and informed consent was given prior to enrollment. DNA was extracted by standard protocols for the patient and her parents.

Exome sequencing

Target enrichment for the samples was performed using the Agilent SureSelect 50 Mb (V3) All Exon Kit (Agilent Technologies, Santa Clara, CA, U.S.A) and next-generation sequencing with an Illumina HiSeq. Read alignment, variant calling, and annotation were done with a pipeline based on Burrows-Wheeler Aligner, Picard, Annovar, and custom annotation scripts. All samples had >92% of bases in the consensus coding sequences (CCDS) covered by at least 20 reads. We excluded variants with minor allele frequency >1% in either the 1,000 genomes project (2012/04 release) or the 6,500 National Heart, Lung and Blood Institute (NHLBI) exomes (Exome Variant Server, NHLBI GO Exome Sequencing Project [ESP], Seattle, WA, data downloaded 2012-10-03) or in ~800 local exomes sequenced at the McGill University and Genome Quebec Innovation Centre. Candidates were validated with Sanger sequencing.


Data were analyzed under an autosomal recessive model as well as a de novo dominant model. There were no homozygous or compound heterozygous rare (frequency of <1%) variants in an epilepsy-associated gene. Comparison of the patient's variants with parents identified three variants that were de novo in FAM189A2, CAPN12, and GRIN2A genes. Only GRIN2A was known to cause a condition characterized by epilepsy with neurodevelopmental defects. The GRIN2A de novo variant was c.2449A>G, p.Met817Val [NM_001134407.1]. The POLYPHEN2 evaluation was “pathogenic,” and the PHAST conservation score was 616 and GERP of 4.2, all in keeping with a highly conserved residue and further evidence of the variant's pathogenicity. The variant was validated by Sanger sequencing (Fig. 1). Taken together, these data provide a diagnosis of epilepsy and neurodevelopmental defects secondary to a de novo mutation in GRIN2A.

Figure 1.

De novo GRIN2A mutation. (A) This shows the pedigree structure of the patient and both parents. Square represents male and circle represents female. (B) Shows the Sanger sequencing plots of the c.2449A>G at GRIN2A, and this represents an amino substitution at position 817 of a methionine to valine.

Insight into the disease pathogenesis for this child provided options for managing the challenging epilepsy. Antiepileptic medications targeting the glutaminergic pathway were considered as primary candidates. Based on this, topiramate was added to her antiepileptic medication regime to augment γ-aminobutyric acid (GABA) receptor function.[10, 11] With the addition of topiramate at 6 mg/kg/day, her seizures decreased significantly in duration and frequency within 1 month from more than a dozen seizures a day to 1–2 seizures per day, each lasting 5–10 s. During the day she is less sedated and more interactive with her parents. This effect has been sustained for 10 months.

A follow-up 24-h ambulatory EEG demonstrated brief clinical seizures, consisting of an epileptic spasm followed by tonic posturing with corresponding generalized high amplitude sharp and slow wave activity followed by significant voltage attenuation. In addition, multiple independent bihemispheric frontotemporal and frontocentral interictal discharges were present. Background activity was slow with runs of frontal intermittent rhythmic delta activity. Recording during sleep improved significantly; overall interictal activity diminished, with evidence of normal sleep architecture, including vertex sharp waves and sleep spindles and elements of stage 3 and 4 sleep. No clinical or electrographic seizures were present during sleep (Fig. 2). This differs greatly from the abnormalities seen in the spectrum of epilepsy-aphasia syndromes, including the severe disorders such as Landau-Kleffner syndrome or continuous spike and wave during slow-wave sleep, but may be similar to that described in subject 7 in the case series by Endele et al.[5]

Figure 2.

EEG for patient with GRIN2A mutation. (A) Independent right and left frontotemporal discharges. (B) Left frontocentral and frontotemporal epileptiform discharges without clinical correlation. (C) FIRDA (frontal intermittent delta activity) with no associated clinical seizure activity. (D) Patient had myoclonus and then tonic extension of arms lasting for <5 s. Demonstrates generalized burst followed by a decremental response.


Despite extensive investigations into the cause of the child's epilepsy and significant developmental delay, a molecular diagnosis was made by a WES trio-based approach. This is not altogether surprising, as this patient's phenotype does not fit clearly into the epilepsy-aphasia spectrum disorders (Landau-Kleffner syndrome, epileptic encephalopathy with continuous spike-wave in sleep) or idiopathic focal epilepsy syndromes recently described with a mutation in the GRIN2A gene.[7-9] Her presentation is unique given the severe developmental delay from infancy with the onset of intractable focal and generalized seizures during the second year of life. Given the atypical nature of her presentation, her phenotype would pose a diagnostic challenge with use of the traditional Sanger sequencing and, as such, this case is suited for a WES trio-based approach. This strategy has proven to be a powerful method to identify the cause of rare diseases thought likely to result from de novo genetic change.[12] An alternative is to sequence only the known epilepsy-related genes, as a panel, versus the entire coding sequence. This has been a successful strategy for mutation identification,[2] although at the time of the sequencing for our patient, GRIN2A was not present on commercially available gene panels.

The GRIN2A gene encodes a subunit of the N-methyl-d-aspartate (NMDA) receptor NR2A subunit that mediates excitatory neurotransmission via glutamate. The mechanism of disease is not entirely understood, and missense, frameshift, and nonsense mutations have been reported to be pathogenic. In voltage experiments involving an amino acid substitution, the gating kinetics of the NMDA receptor has been shown to be altered although not abrogated.[5, 7, 9] NMDA receptors have a role in the normal central nervous system (CNS) development and plasticity of the synapse.[13, 14] A disruption or alteration in this receptor would be expected to have a significant effect on neurodevelopment, cognition, and seizure presentation via the GABA pathway. Although the exact mechanisms of action for the antiseizure properties are unclear, topiramate is thought to enhance GABA-evoked currents and, in addition, to block sodium, calcium, and alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid receptor (AMPA)/kainate channels.[10] Other medications targeting the GABA pathway may also be of clinical benefit. A recent report described an individual with a substitution (p.V506A) in GRIN2A that showed clinical benefit with felbamate and clobazam that both target the glutamate pathway; however, the authors do report no benefit from a trial of topiramate.[15]

In summary, WES led to the identification of a de novo mutation in GRIN2A in a patient with an early onset severe refractory epilepsy and severe global developmental delay. The diagnosis provided accurate recurrence risk counseling for parents and initially guided the choice of antiepileptic medication for seizure management. Further study of individuals with GRIN2A mutations and their response to antiepileptic therapies that affect the GABA pathway will provide better information for both short- and long-term seizure control. Our report highlights that GRIN2A mutations should be considered in children with severe developmental delay and epilepsy, in addition to children presenting along the spectrum of epilepsy-aphasia disorders.


The authors would first like to thank the patient and her family; without their participation this work would not be possible. We also acknowledge Dr. Erick Sell (U. Ottawa) for EEG interpretation and to Drs. Johannes Lemke and Markus Wolff for helpful correspondence. This work was funded by the Government of Canada through Genome Canada, the Canadian Institutes of Health Research (CIHR), and the Ontario Genomics Institute (OGI-049). Additional funding was provided by Genome Quebec and Genome British Columbia. This work was selected for study by the FORGE Canada Steering Committee, K. Boycott (U. Ottawa), J. Friedman (U. British Columbia), J. Michaud (U. Montreal), F. Bernier (U. Calgary), M. Brudno (U. Toronto), B. Fernandez (Memorial U.), B. Knoppers (McGill U.), M. Samuels (U. de Montreal), and S. Scherer (U. Toronto). DAD and KMB are supported by the CIHR Institute of Genetics Clinical Investigatorship Award.


The authors declare no conflict of interest. We confirm that we have read the Journal's position on issues involved in ethical publication and affirm that this report is consistent with those guidelines.


  • Image of creator

    Sunita Venkateswaran is an Assistant Professor in the Division of Pediatric Neurology at the Children's Hospital of Eastern Ontario.