| Domain | Evidence-backed finding | Quantitative data | Suggested ontology terms | Evidence level |
|---|---|---:|---|---|
| Disease identity | Developmental and epileptic encephalopathy 46 (DEE46) is a rare Mendelian neurodevelopmental disorder caused by pathogenic heterozygous **GRIN2D** variants affecting the GluN2D NMDA-receptor subunit. (pqac-00000008, pqac-00000029) | Open Targets reports 5 disease–target evidence items. | **MONDO:0014947**; developmental and epileptic encephalopathy 46 | Aggregated disease-resource plus human genetic evidence |
| Gene and protein | **GRIN2D** encodes glutamate ionotropic receptor NMDA-type subunit 2D. Functional NMDA receptors contain two glycine-binding GluN1 and two glutamate-binding GluN2 subunits. (pqac-00000008, pqac-00000011, pqac-00000026) | GluN2D is 1,336 amino acids. | **HGNC:4588**; **NCBI Gene:2906**; **GO:0004972**, NMDA-selective glutamate receptor activity | Established receptor biology and authoritative gene-resource evidence |
| Inheritance | The best-established mechanism is autosomal dominant, usually due to a **de novo heterozygous** missense variant. (pqac-00000029, pqac-00000030) | Recurrent p.Val667Ile was initially found in 2 unrelated children. | **HP:0000006**, autosomal dominant inheritance; **HP:0025352**, de novo variant | Human trio-sequencing and segregation evidence |
| Development | Developmental delay or intellectual disability accompanies epilepsy and is commonly severe or profound, affecting language, cognition, motor skills, and independence. (pqac-00000000, pqac-00000003) | Developmental delay/intellectual disability occurred in all 13 individuals in the early aggregated cohort. | **HP:0001263**, global developmental delay; **HP:0001249**, intellectual disability | Small human case series |
| Epilepsy | Seizures usually begin in infancy, evolve in type, and are frequently refractory to conventional antiseizure therapy. (pqac-00000005, pqac-00000027, pqac-00000030) | Mean onset **11 ± 3.6 months**; median **6.5 months**; approximately **75% (9/12)** refractory or partly responsive. | **HP:0001250**, seizure; **HP:0003593**, infantile onset; **HP:0002063**, drug-resistant epilepsy | Aggregated human cohort evidence |
| Seizure types | Epileptic spasms, focal motor/clonic, atypical absence, myoclonic, generalized tonic-clonic seizures, and status epilepticus are reported. (pqac-00000000, pqac-00000003, pqac-00000030) | Five of 8 individuals in one series had infantile spasms. | **HP:0012469**, infantile spasms; **HP:0002069**, generalized tonic-clonic seizure; **HP:0002123**, generalized myoclonic seizure; **HP:0002133**, status epilepticus | Human case-series evidence |
| EEG | Multifocal epileptiform discharges and hypsarrhythmia predominate; focal spike-and-wave and paroxysmal fast activity also occur. (pqac-00000003, pqac-00000027, pqac-00000033) | Multifocal discharges **5/12**; hypsarrhythmia **4/12**; focal spike-and-wave **2/12**; paroxysmal fast activity **1/12**. | **HP:0002353**, EEG abnormality; **HP:0002521**, hypsarrhythmia | Human EEG observations with visually reviewed cohort summary |
| Motor phenotype | Hypotonia, poor motor control, dyskinesia, and choreiform movements range from mild to profound. (pqac-00000000, pqac-00000003, pqac-00000030) | Hypotonia or poor motor control in **9/13 (69%)**. | **HP:0001252**, hypotonia; **HP:0100022**, abnormality of movement | Human case-series evidence |
| Behavior | Autistic behavior, stereotypies, reduced eye contact, and occasional ADHD-like symptoms are reported. (pqac-00000001, pqac-00000003, pqac-00000030) | Autism-spectrum features approximately **4/13 (31%)**. | **HP:0000729**, autistic behavior; **HP:0000733**, stereotypy | Observational evidence; standardized assessments usually unavailable |
| Vision | Cerebral/cortical visual impairment and oculomotor apraxia occur in a subset. (pqac-00000000, pqac-00000003, pqac-00000030) | Cerebral visual impairment **5/13 (38%)**. | **HP:0100704**, cortical visual impairment; **HP:0000657**, oculomotor apraxia | Human case-series evidence |
| MRI and anatomy | MRI may be normal or show cerebral/cortical atrophy, microcephaly, small frontal lobes, reduced white matter, or a thin corpus callosum; consistent lateralization is not established. (pqac-00000000, pqac-00000001, pqac-00000030) | Cortical/cerebral atrophy approximately **5/11 (45%)** among individuals with MRI data. | **HP:0002120**, cerebral cortical atrophy; **HP:0000252**, microcephaly; **HP:0002079**, thin corpus callosum; **UBERON:0000955**, brain | Small, heterogeneous human imaging series |
| p.Val667Ile | **c.1999G>A (p.Val667Ile)** is a recurrent de novo M3-domain gain-of-function variant that increases agonist potency, channel opening, and response duration while reducing endogenous inhibition. (pqac-00000028, pqac-00000029) | Approximately **2-fold** greater glutamate/glycine potency, **6-fold** higher open probability; Mg²⁺ IC₅₀ shifted from **220 to 346 μM**. | Sequence variant; **GO:0006816**, calcium ion transport; **GO:0007268**, chemical synaptic transmission | Human genetics plus oocyte, HEK293, and single-channel assays |
| Other variants | Reported DEE-associated missense variants include p.Asp449Asn, p.Ser573Phe, p.Leu670Phe, p.Ala675Thr, p.Ala678Asp, p.Met681Ile, p.Ser694Arg, p.Ser1271Leu/Phe, and p.Arg1313Trp. Effects vary by variant. (pqac-00000004, pqac-00000009, pqac-00000027) | Early synthesis: 12 patients with 9 unique variants across the agonist-binding, pre-M1, M3, and C-terminal regions. | Sequence variant; agonist-binding domain; protein transmembrane domain | Human genetic and in-vitro functional evidence |
| Trafficking and gating | Six tested variants reduced receptor surface expression. Some variants increase agonist potency/open probability, whereas others decrease open probability, demonstrating that missense position alone does not establish gain or loss of function. (pqac-00000009, pqac-00000013, pqac-00000032) | p.Leu670Phe and p.Ala678Asp open probabilities **0.36** and **0.20**, versus **0.0067** for wild type. | **GO:0005886**, plasma membrane; **GO:0098794**, postsynapse | In-vitro electrophysiology and surface-expression assays |
| Excitotoxicity | Excess mutant-receptor activity can cause dendritic swelling and neuronal death; contribution to human brain injury is biologically plausible but inferred. (pqac-00000010, pqac-00000028, pqac-00000029) | p.Val667Ile caused **>50% lethality** in transfected neurons; p.Ala678Asp reduced viability to **55%**, rescued to **77%** by memantine. | **GO:0008219**, cell death; **GO:0006816**, calcium ion transport | Primary rat cortical-neuron experiments; human link inferential |
| Genetic diagnosis | Trio epilepsy/DEE panels, trio exome sequencing, or genome sequencing with parental confirmation are appropriate. ACMG/AMP classification should be supplemented by functional testing when treatment depends on gain- versus loss-of-function status. (pqac-00000029, pqac-00000031) | No validated disease-specific biochemical biomarker or functional threshold exists. | Whole-exome sequencing; whole-genome sequencing; sequence-variant interpretation | Standard molecular-diagnostic practice supported by sequencing studies |
| Ancillary diagnosis | Prolonged video EEG measures subclinical seizure burden; MRI assesses atrophy and structural differentials. Metabolic, mitochondrial, karyotype, and microarray testing may be unrevealing. (pqac-00000016, pqac-00000030) | One reported patient had negative karyotype, array-CGH, metabolic, mitochondrial-gene, respiratory-chain, and ATPase testing. | **HP:0002353**, EEG abnormality; brain MRI; chromosomal microarray | Individual-patient evidence |
| Conventional treatment | Antiseizure medicines are selected by seizure type and include valproate, levetiracetam, topiramate, benzodiazepines, vigabatrin, carbamazepine/oxcarbazepine, and lamotrigine; responses vary and polytherapy is common. (pqac-00000017, pqac-00000018, pqac-00000020) | Approximately **75%** of the early cohort was refractory or only partly responsive. | Antiseizure therapy; polytherapy; supportive care | Retrospective case-series evidence; no comparative trials |
| Memantine | Off-label memantine has improved seizures, development, or behavior in some patients, particularly with demonstrated gain-of-function variants, but responses are inconsistent. (pqac-00000015, pqac-00000016, pqac-00000018) | Original p.Val667Ile case: **2 to 20 mg/day (0.85 mg/kg/day)**; other cases used approximately **0.5 mg/kg/day**. | Memantine; NMDA-receptor antagonist; precision medicine | Human n-of-1/case-series plus in-vitro pharmacology; no randomized GRIN2D trial |
| Ketamine and magnesium | Ketamine plus magnesium produced dramatic EEG and clinical improvement in one p.Val667Ile-associated refractory status case, but evidence is insufficient for routine use and ketamine may aggravate seizures at higher doses. (pqac-00000016, pqac-00000022, pqac-00000028) | ICU regimen: MgSO₄ **2 g every 4 h** plus ketamine **2 mg/kg/h**; later enteral ketamine **1 mg/kg every 6 h**. | Ketamine; magnesium sulfate; NMDA-receptor antagonist; status-epilepticus treatment | Single human case with EEG correlation; post-2024 mouse safety signal |
| Other interventions | Vagus-nerve stimulation produced partial control in one patient. Combined memantine, IV immunoglobulin, steroids, and magnesium coincided with seizure freedom in another, but the effective component is unknown. (pqac-00000017, pqac-00000030) | Individual cases only. | Vagus nerve stimulation; immunoglobulin therapy; corticosteroid therapy | Very-low-certainty case evidence |
| Models | Cellular systems include Xenopus oocytes, HEK293 cells, cultured rat cortical neurons, and a reported patient-derived iPSC line. Grin2d-null mice show behavioral abnormalities; a p.Val667Ile-ortholog knock-in model recapitulates seizures, motor impairment, cognitive deficits, and premature mortality. (pqac-00000022, pqac-00000023, pqac-00000025) | Knock-in mice showed limited correction with memantine/phenytoin; ketamine **4–10 mg/kg** provoked seizures. | **CL:0000540**, neuron; **CL:0000120**, granule cell; **CL:0000121**, Purkinje cell | In-vitro, iPSC-resource, and animal-model evidence; knock-in results post-2024 |
| Evidence gaps | Population prevalence, incidence, penetrance, carrier frequency, founder effects, robust sex ratio, validated prognostic biomarkers, disease-specific quality-of-life measures, natural-history staging, environmental modifiers, protective factors, and controlled treatment-response rates are unavailable. | Published disease cohorts remain only in the tens of patients. | Not assigned | Explicit absence of adequate epidemiologic or controlled clinical evidence |
| Therapeutic-development gaps | No approved disease-modifying therapy, GRIN2D-specific randomized trial, validated gene/RNA therapy, or established rule for converting in-vitro receptor effects into clinical treatment selection was identified. (pqac-00000021, pqac-00000025) | Clinical-trial search found no relevant GRIN2D-specific interventional trial. | Gene therapy; RNA therapy; precision medicine | Evidence gap; experimental concepts only |


*Table: Compact evidence table covering DEE46 identifiers, phenotypes, molecular mechanisms, diagnostic approaches, treatments, model systems, and major knowledge gaps. Quantitative findings are separated by evidence level to distinguish human observations from experimental and inferential results.*