DLG4-Related Synaptopathy

Key Findings

Finding 1 — Definition and core identity of the disorder

DLG4-related synaptopathy is a rare autosomal-dominant neurodevelopmental disorder caused by de novo variants in DLG4/PSD-95. It was defined by the landmark cohort of Rodríguez-Palmero and colleagues, who reported "the clinical and genetic features of 53 patients (42 previously unpublished) with DLG4 variants" and proposed "we designate this group of disorders as DLG4-related synaptopathy" [PMID: 33597769]. Of the 45 different DLG4 variants they identified, "39 were predicted to lead to loss of protein function and the majority occurred de novo" [PMID: 33597769].

The disorder maps to the following identifiers: OMIM 618793 (Intellectual developmental disorder, autosomal dominant 62 / IDD62), MONDO:0032919, MedGen/UMLS C5394083, DOID:0061035, and GARD 0025775. The causal gene DLG4 is HGNC:2903, NCBI Gene 1742, UniProt P78352, located at chromosome 17p13.1, and encodes PSD-95.

Finding 2 — Core clinical phenotype

The predominant clinical features are early-onset global developmental delay, intellectual disability, ASD, and ADHD. The original cohort found that "the clinical picture was predominated by early onset global developmental delay, intellectual disability, autism spectrum disorder, and attention deficit-hyperactivity disorder, all of which point to a brain disorder" [PMID: 33597769]. Additional features include hypotonia, sleep disturbance, movement disorders, strabismus, scoliosis, and joint hypermobility. Notably, the study refined an earlier claim: "Marfanoid habitus, which was previously suggested to be a characteristic feature of DLG4-related phenotypes, was found in only nine individuals" (9/53), and there was no distinct facial dysmorphism.

Epilepsy is a major feature. A dedicated epilepsy study noted that "even though epilepsy is present in 50% of the individuals, it has not been investigated in detail" and reported that "encephalopathy related to status epilepticus during slow-wave sleep (ESES)/developmental epileptic encephalopathy with spike-wave activation during sleep (DEE-SWAS) was diagnosed in >25% of the individuals" [PMID: 38135915]. Focal seizures were the most common type.

Suggested HPO terms: Intellectual disability (HP:0001249), Seizure (HP:0001250), Autistic behavior (HP:0000729), Attention deficit hyperactivity disorder (HP:0007018), Muscular hypotonia (HP:0001252), Sleep disturbance (HP:0002360), Strabismus (HP:0000486), Scoliosis (HP:0002650), Joint hypermobility (HP:0001382), Global developmental delay (HP:0001263), Cerebellar vermis atrophy (HP:0006855).

Finding 3 — Molecular mechanism (PSD-95 scaffolding of glutamate receptors)

PSD-95 scaffolds NMDA and AMPA receptors at the excitatory postsynaptic density; its loss impairs synaptic maturation and plasticity. As summarized in a review, "postsynaptic density protein-95 (PSD-95) is a major regulator of synaptic maturation by interacting, stabilizing and trafficking N-methyl-d-aspartic acid receptors (NMDARs) and α-amino-3-hydroxy-5-methyl-4-isoxazoleproprionic acid receptors (AMPARs) to the postsynaptic membrane" [PMID: 29169997]. The original disease paper states that "postsynaptic density protein-95 (PSD-95), encoded by DLG4, regulates excitatory synaptic function in the brain" [PMID: 33597769].

At the molecular level, protein-truncating variants and at least one deep-intronic variant act by reducing functional protein. A patient-derived DLG4 V692Wfs12 transcript illustrates the mechanism: "the mutant transcript escapes nonsense-mediated decay but results in reduced PSD-95 protein expression"* [PMID: 42565830], demonstrating that even a transcript that evades NMD ultimately yields reduced PSD-95 protein — consistent with haploinsufficiency.

Suggested GO/CL terms: postsynaptic density (GO:0014069), dendritic spine (GO:0043197), glutamatergic synapse (GO:0098978), regulation of synaptic plasticity (GO:0048167); glutamatergic neuron (CL:0000679).

Finding 4 — Extreme loss-of-function intolerance supports haploinsufficiency

DLG4 is among the most constrained genes in the human genome. gnomAD constraint metrics for DLG4 (ENSG00000132535, chr17:7,187,187–7,219,841, GRCh38) show pLI = 1.0; observed LoF variants = 7 versus expected 93.8 (oe_lof = 0.075, 90% CI 0.042–0.140, i.e. LOEUF = 0.14); LoF Z = 7.60; missense Z = 6.06 (oe_mis = 0.55). This extreme depletion of loss-of-function variation in the general population is exactly the population-genetic signature expected for a haploinsufficient, dominant neurodevelopmental gene, and it strongly corroborates the mechanistic model that a single loss-of-function allele is sufficient to cause disease.

Finding 5 — Model organisms recapitulate cognitive, behavioral, and synaptic-plasticity deficits

Multiple mouse models and patient-derived cellular models support the disease mechanism:

• PSD-95-null mice show paradoxically enhanced hippocampal LTP with severely impaired learning: "in mutant mice lacking PSD-95, the frequency function of NMDA-dependent LTP and LTD is shifted to produce strikingly enhanced LTP at different frequencies of synaptic stimulation" and "this frequency shift is accompanied by severely impaired spatial learning" [PMID: 9853749].
• PDZ1/2 ligand-binding-deficient PSD-95 knock-in mice show reduced PSD accumulation of PSD-95/PSD-93/AMPARs, abnormal anxiety, and impaired spatial, working, and remote memory [PMID: 23268962].
• Dlg4−/− mice model ASD-relevant behavior: they "showed increased repetitive behaviors, abnormal communication and social behaviors, impaired motor coordination, and increased stress reactivity and anxiety-related responses" [PMID: 20952458].
• A patient-derived Dlg4 V692Wfs*12/+ knock-in mouse faithfully models the human disorder: "Dlg4V692Wfs12/+ mice recapitulate several hallmark features of SHINE syndrome, often in a sex-specific manner"* — including reduced PSD-95, learning/cognitive-flexibility deficits (male-biased), and sleep abnormalities [PMID: 42565830].
• A patient iPSC line (AOUMEYi004-A) carrying c.2155A>T p.(Arg719*) was established for in vitro disease modeling [PMID: 42462545].

Finding 6 — Management is supportive; no disease-modifying therapy

There is no targeted or curative therapy. Care is symptom-directed: antiseizure medications for epilepsy, developmental/rehabilitative therapies, and management of sleep and behavior. Antiseizure medication response was assessed retrospectively across 35 patients with epilepsy, with variable response and frequent refractoriness in the ESES/DEE-SWAS forms [PMID: 38135915]. For treatment-resistant psychosis, an adolescent with SHINE syndrome and early-onset schizophrenia/catatonia improved markedly on clozapine: "after failing three antipsychotic drug treatments, the patient was started on clozapine, which resulted in significant improvements in positive and negative symptoms" [PMID: 37386468]. Experimental PSD-95-directed agents (e.g., nerinetide/Tat-NR2B9c, which disrupt the PSD-95/nNOS interaction) exist but are being developed for stroke and pain, not for this disorder [PMID: 40712457].

Finding 7 — Protein architecture and variant spectrum

PSD-95 (UniProt P78352, 724 aa) is a MAGUK with a characteristic modular architecture: three PDZ domains (PDZ1 aa 65–151, PDZ2 aa 160–246, PDZ3 aa 313–393), an SH3 domain (aa 428–498), and a guanylate kinase-like (GK) domain (aa 534–709), plus an N-terminal disordered region (aa 15–35). Twenty-two experimental PDB structures have been deposited. Pathogenic variants span the entire gene and include nonsense/frameshift (e.g., c.2155A>T p.Arg719; c.2074_2075 frameshift p.Val692Trpfs12 in the GK domain), splice-site, a deep-intronic pseudoexon variant (c.2105+235C>T), and six missense variants. "The six missense variants identified were suggested to lead to structural or functional changes by protein modeling studies" [PMID: 33597769], and across cohorts "the majority [are] predicted to be protein-truncating" [PMID: 37525972].

Finding 8 — Diagnosis, inheritance, and epidemiology

Diagnosis is molecular, established by clinical whole-exome or whole-genome sequencing identifying a heterozygous DLG4 variant; RNA/functional studies resolve splice and deep-intronic variants. In the landmark cohort, most cases were simplex/de novo: "the majority occurred de novo (four with unknown origin)" [PMID: 33597769]. A deep-intronic variant was "identified using whole genome sequencing" [PMID: 37525972], underscoring the value of WGS plus RNA studies when exome sequencing is unrevealing. Brain MRI and EEG (including sleep EEG/video-polygraphy for ESES/DEE-SWAS) are used for phenotyping: "data on awake and sleep electroencephalography (EEG) and/or video-polygraphy and brain magnetic resonance imaging were collected" [PMID: 38135915]. The disorder is ultra-rare (~53 patients in the defining 2021 series, with additional case reports since); no population prevalence or incidence has been established, and it is under-ascertained, including late/adolescent diagnoses [PMID: 40444229].

Finding 9 — Anatomy and temporal course

This is a brain-centered disorder affecting glutamatergic excitatory synapses. PSD-95 is "an essential scaffolding protein during synaptogenesis and neurodevelopment" [PMID: 29169997], localizing to the postsynaptic density of excitatory synapses. Primary organ: brain/nervous system; cell type: glutamatergic neurons (CL:0000679); subcellular compartment: postsynaptic density/dendritic spine (GO:0014069, GO:0043197). Some patients show cerebellar vermis atrophy on MRI (HP:0006855, ~33% in the original small series). Onset is early (infancy/early childhood) with global developmental delay; the course is chronic and lifelong and generally non-degenerative, but developmental/verbal-motor regression can occur in those who develop status epilepticus in sleep: "regression in verbal and/or motor domains was observed in all individuals who su[ffered status epilepticus]" [PMID: 38135915].

Finding 10 — Purely genetic etiology; high evolutionary conservation

DLG4-related synaptopathy is monogenic with no established environmental, infectious, lifestyle, or gene-environment contribution; de novo germline DLG4 variants arise sporadically, consistent with "the majority occurred de novo" [PMID: 33597769]. No protective alleles or modifier genes have been identified. ClinVar (accessed 2026) lists ~445 DLG4 variant records: ~207 pathogenic, ~60 likely pathogenic, and ~316 of uncertain significance. DLG4/PSD-95 is deeply conserved: orthologs include mouse Dlg4 (NCBI Gene 13385, MGI:1277959), rat Dlg4 (NCBI Gene 29495), zebrafish dlg4a/dlg4b, Drosophila dlg1 (discs large), and C. elegans dlg-1; the MAGUK/PDZ–SH3–GK architecture is conserved from invertebrates to humans.

Comprehensive Section-by-Section Report

1. Disease Information

Overview. DLG4-related synaptopathy (SHINE syndrome) is a rare autosomal-dominant neurodevelopmental "synaptopathy" — a disorder of synaptic structure and function — caused by heterozygous, predominantly de novo loss-of-function variants in DLG4 (PSD-95). It presents in infancy/early childhood with global developmental delay and evolves into a lifelong picture of intellectual disability, autism, ADHD, hypotonia, sleep disturbance, movement disorders, and (in ~50%) epilepsy [PMID: 33597769, 38135915].

Key identifiers. OMIM 618793 (IDD62); MONDO:0032919; MedGen/UMLS C5394083; DOID:0061035; GARD 0025775. Gene: DLG4 (HGNC:2903, NCBI Gene 1742, UniProt P78352, 17p13.1). A specific ICD-10/ICD-11 code is not assigned to this ultra-rare entity; it is captured under intellectual disability / developmental disorder categories. MeSH indexing is via DLG4/PSD-95 and intellectual disability terms.

Synonyms. SHINE syndrome (Sleep disturbances, Hypotonia, Intellectual disability, Neurological disorders, Epilepsy); Intellectual developmental disorder, autosomal dominant 62 (IDD62); DLG4-related synaptopathy; PSD-95-related neurodevelopmental disorder.

Information source. The knowledge base derives from aggregated disease-level resources (OMIM, ClinVar, gnomAD) plus published patient cohorts and case reports — i.e., published individual-patient data aggregated into cohorts, not routine EHR mining.

2. Etiology

Causal factors. Purely genetic: heterozygous DLG4 variants, most arising de novo [PMID: 33597769]. No environmental, infectious, or lifestyle cause is established (Finding 10).

Genetic risk factors. The causal variant itself is the sole established risk factor. DLG4 is extremely LoF-intolerant (pLI = 1.0; LOEUF = 0.14), meaning even a single loss-of-function allele confers disease (Finding 4). No susceptibility loci or modifier genes have been established.

Environmental risk factors / protective factors / gene-environment interactions. None identified. No protective alleles are known. Because most cases are de novo, advanced parental age (a general contributor to de novo mutation rates) is a plausible but unproven population-level consideration; this is not disease-specific evidence.

3. Phenotypes

Severity/progression. Severity is variable; the disorder is chronic and lifelong and generally non-degenerative. However, verbal/motor regression occurs in individuals who develop status epilepticus in sleep (ESES/DEE-SWAS) [PMID: 38135915].

Quality-of-life impact. Substantial: intellectual disability, autism, epilepsy, and sleep disturbance collectively impair communication, learning, independence, and family functioning. No disease-specific EQ-5D/SF-36/PROMIS data are available.

4. Genetic/Molecular Information

Causal gene. DLG4 (HGNC:2903; NCBI Gene 1742; OMIM gene 602887; UniProt P78352; 17p13.1) encoding PSD-95.

Pathogenic variants. Variants span the gene and are predominantly protein-truncating (nonsense, frameshift, splice-site), with a minority of missense and at least one deep-intronic pseudoexon variant. Examples: c.2155A>T p.(Arg719) [PMID: 42462545]; c.2074_2075 frameshift p.(Val692Trpfs12) in the GK domain [PMID: 42565830]; c.2105+235C>T deep-intronic [PMID: 37525972]. Of 45 variants in the defining cohort, 39 were predicted loss-of-function, and the six missense variants were modeled to disrupt structure/function [PMID: 33597769].

Variant classification (ClinVar, 2026). ~445 DLG4 records: ~207 pathogenic, ~60 likely pathogenic, ~316 uncertain significance (Finding 10). Per ACMG/AMP, truncating variants in this LoF-intolerant gene generally meet PVS1.

Allele frequency. Pathogenic variants are absent/vanishingly rare in gnomAD, consistent with de novo origin and extreme constraint (pLI = 1.0, LOEUF = 0.14).

Origin. Germline, predominantly de novo [PMID: 33597769]. No somatic disease association.

Functional consequence. Loss of function / haploinsufficiency (reduced PSD-95 protein) [PMID: 42565830]; some missense variants may act via structural disruption.

Modifier genes / epigenetics / chromosomal abnormalities. None established. DLG4 sits at 17p13.1; larger 17p deletions encompassing DLG4 could plausibly contribute but are not a defined mechanism for this entity.

5. Environmental Information

Not applicable. This is a monogenic disorder with no established environmental factors, lifestyle factors, or infectious agents (Finding 10) [PMID: 33597769].

6. Mechanism / Pathophysiology

Molecular pathway. Glutamatergic synaptic signaling. PSD-95 is the central organizer of the excitatory postsynaptic density (PSD), where it clusters and traffics NMDARs and AMPARs and couples them to downstream signaling (e.g., nNOS, SynGAP, CaMKII) [PMID: 29169997, 20554866].

Causal chain.

DLG4 LoF variant (de novo, heterozygous)
│
▼
Reduced functional PSD-95 protein (haploinsufficiency)
│  (transcript may escape NMD but yields less protein — PMID 42565830)
▼
Disorganized excitatory postsynaptic density:
  ↓ clustering/trafficking of NMDARs & AMPARs (PMID 23268962)
│
▼
Impaired synaptic maturation & aberrant plasticity
  (altered LTP/LTD; PMID 9853749, 23268962)
│
▼
Disrupted neural circuit development (glutamatergic neurons)
│
▼
Clinical manifestations: developmental delay, ID, ASD/ADHD,
hypotonia, sleep disturbance, epilepsy

Upstream vs downstream. Upstream: the DLG4 variant and reduced PSD-95. Downstream: receptor mis-clustering, altered synaptic plasticity, circuit dysfunction, and behavior.

Cellular process. Synaptogenesis, synaptic maturation, and plasticity in glutamatergic neurons. Protein dysfunction: loss of function of a scaffold (not aggregation). Immune/metabolic involvement: none established. Biochemical: receptor scaffolding defect at the PSD.

Molecular profiling. No disease-specific human transcriptomic/proteomic/metabolomic signatures are published; mechanistic evidence derives from mouse models and in vitro biochemistry [PMID: 9853749, 23268962, 20952458, 20554866]. Patient iPSC lines now enable such profiling [PMID: 42462545].

Suggested GO terms: GO:0014069 (postsynaptic density), GO:0098978 (glutamatergic synapse), GO:0048167 (regulation of synaptic plasticity), GO:0007416 (synapse assembly), GO:0035249 (synaptic transmission, glutamatergic). CL: CL:0000679 (glutamatergic neuron).

7. Anatomical Structures Affected

• Organ level: Brain / central nervous system (primary). Body system: nervous system. UBERON: brain (UBERON:0000955), cerebral cortex (UBERON:0000956), hippocampus (UBERON:0002421), cerebellar vermis (UBERON:0004720; atrophy in a subset, HP:0006855).
• Tissue/cell level: Nervous tissue; glutamatergic excitatory neurons (CL:0000679).
• Subcellular level: Postsynaptic density (GO:0014069), dendritic spine (GO:0043197), postsynaptic membrane (GO:0045211).
• Lateralization: Bilateral (diffuse CNS involvement).

8. Temporal Development

• Onset: Early — infancy/early childhood, with global developmental delay as the presenting feature; insidious/chronic onset.
• Progression: Chronic, lifelong, generally non-degenerative. Epilepsy (including ESES/DEE-SWAS) typically emerges in childhood. Regression in verbal/motor domains occurs specifically in those developing status epilepticus in sleep [PMID: 38135915].
• Critical period: The ESES/DEE-SWAS window in childhood represents a period of vulnerability (regression) and a potential opportunity for intervention (seizure control). Diagnosis may be delayed to adolescence/adulthood [PMID: 40444229, 42462545].

9. Inheritance and Population

• Inheritance: Autosomal dominant (OMIM 618793), predominantly de novo; "the majority occurred de novo (four with unknown origin)" [PMID: 33597769].
• Penetrance/expressivity: Presumed high penetrance for LoF variants; expressivity is variable (severity ranges; epilepsy in ~50%).
• Anticipation/mosaicism/founder effects/consanguinity: Not applicable/not reported (dominant, de novo, sporadic). Consanguinity is not relevant. Germline mosaicism is theoretically possible (as for any de novo disorder) but not specifically documented.
• Carrier frequency: Not applicable (dominant, de novo).
• Epidemiology: Ultra-rare; ~53 patients in the defining series with subsequent case reports. No reliable prevalence/incidence estimate exists; the disorder is under-ascertained.
• Demographics: No ethnic predilection established. Mouse models suggest possible sex-biased severity (male-biased cognitive deficits in the knock-in model [PMID: 42565830]), but human sex-ratio data are not established.

10. Diagnostics

• Genetic testing (definitive): Clinical whole-exome sequencing (WES) or whole-genome sequencing (WGS) identifying a heterozygous DLG4 variant; multi-gene neurodevelopmental/epilepsy/ID panels including DLG4; chromosomal microarray for larger 17p13.1 deletions. RNA/functional studies are required to interpret splice-site and deep-intronic variants (e.g., c.2105+235C>T identified by WGS) [PMID: 37525972, 42462545].
• Phenotyping studies: Brain MRI (may show cerebellar vermis atrophy); EEG including sleep EEG/video-polygraphy to detect ESES/DEE-SWAS [PMID: 38135915].
• Biomarkers/laboratory tests: No specific biochemical biomarker; diagnosis is molecular. No metabolic, proteomic, or metabolomic diagnostic markers.
• Clinical criteria: No formal consensus criteria; diagnosis rests on compatible neurodevelopmental phenotype plus a pathogenic DLG4 variant.
• Differential diagnosis: Other monogenic synaptopathies and neurodevelopmental disorders (e.g., SHANK-, SYNGAP1-, GRIN-, and other MAGUK/PSD-related disorders); Marfanoid habitus previously led to confusion but is not characteristic [PMID: 33597769].
• Screening: No population/newborn screening. Cascade testing is generally unnecessary given the de novo nature; parental testing informs recurrence-risk counseling.

11. Outcome/Prognosis

• Survival/mortality: No evidence of markedly reduced life expectancy; the disorder is non-degenerative. Mortality data are not established (ultra-rare).
• Morbidity/function: Substantial lifelong disability from intellectual disability, autism, epilepsy, and behavioral/sleep problems. Many individuals require lifelong support.
• Disease course/complications: Epilepsy (including refractory ESES/DEE-SWAS with associated regression), psychiatric complications (including treatment-resistant psychosis/catatonia in at least one adolescent [PMID: 37386468]), and orthopedic issues (scoliosis).
• Prognostic factors: Development of status epilepticus in sleep predicts regression [PMID: 38135915]; severity appears to correlate broadly with the degree of PSD-95 loss. No validated prognostic biomarkers.

12. Treatment

No disease-modifying therapy exists. Management is supportive and multidisciplinary.

Pharmacogenomics/gene/cell/RNA/targeted/immuno-therapies: None established for this indication. Patient iPSC lines [PMID: 42462545] and patient-derived knock-in mice [PMID: 42565830] provide platforms for future therapeutic development.

13. Prevention

• Primary prevention: Not possible (de novo genetic origin).
• Secondary/tertiary prevention: Early developmental intervention; sleep-EEG surveillance to detect and treat ESES/DEE-SWAS early (to mitigate regression); management of complications (scoliosis, psychiatric symptoms).
• Genetic counseling: Recurrence risk is low for parents of a child with a confirmed de novo variant (bounded by the small possibility of parental germline mosaicism); affected individuals who reproduce carry a 50% transmission risk. Prenatal/preimplantation testing is technically possible for a known familial variant.

14. Other Species / Natural Disease

• Taxonomy/orthologs: DLG4/PSD-95 is deeply conserved. Mouse Dlg4 (NCBI Gene 13385, MGI:1277959); rat Dlg4 (NCBI Gene 29495); zebrafish dlg4a/dlg4b; Drosophila dlg1 (discs large); C. elegans dlg-1. The MAGUK PDZ–SH3–GK architecture is conserved from invertebrates to humans (Finding 10).
• Natural disease in other species: No naturally occurring DLG4 disorder is documented in companion animals or wildlife (OMIA); disease knowledge derives from engineered models.
• Comparative/evolutionary: High conservation of the scaffold and its receptor-clustering function underlies the strong translational validity of rodent models.

15. Model Organisms

Model strengths: Rodent models robustly reproduce cognitive, behavioral, synaptic-plasticity, and (in the newest knock-in) sleep phenotypes, with a patient-specific variant. Limitations: Species differences in cognition/epilepsy; sex-specific effects require careful design; iPSC models capture cellular but not circuit-level phenotypes. Resources: MGI (Dlg4, MGI:1277959), IMPC, Cellosaurus (iPSC line).

Mechanistic Model / Interpretation

DLG4-related synaptopathy is a textbook haploinsufficiency synaptopathy. The convergent evidence — extreme population-genetic constraint (pLI = 1.0, LOEUF = 0.14), a predominance of de novo protein-truncating variants, direct demonstration of reduced PSD-95 protein from a patient variant, and faithful recapitulation of cognitive/behavioral/sleep phenotypes in a patient-derived knock-in mouse — locks together into a single coherent causal chain: one lost DLG4 allele → less PSD-95 → a disorganized glutamatergic postsynaptic density → impaired synaptic maturation and aberrant plasticity → abnormal circuit development → the SHINE clinical phenotype.

Two clinically important nuances emerge. First, the relationship between synaptic plasticity and cognition is not simply "less plasticity = worse learning": PSD-95-null mice show enhanced LTP yet impaired learning [PMID: 9853749], indicating that PSD-95 sets the correct dynamic range and metaplasticity of synapses, not merely their strength. Second, epilepsy — specifically the ESES/DEE-SWAS subtype — is a modifiable driver of the worst outcomes (regression), making its early detection via sleep EEG a high-yield clinical priority [PMID: 38135915].

Population genetics        Molecular biology          Model systems              Clinic
─────────────────────      ─────────────────────      ─────────────────────      ─────────────────────
pLI=1.0, LOEUF=0.14   →    ↓PSD-95 protein       →    KO/KI mice: learning,  →   GDD, ID, ASD/ADHD,
(LoF not tolerated)        disorganized PSD           sleep, behavior deficits    epilepsy, sleep, hypotonia
                   (↓NMDAR/AMPAR clustering)  iPSC platform               (regression if ESES)

Evidence Base

Additional supporting/context papers on PSD-95-interacting partners and MAGUK biology: [PMID: 18248606], [21878521], [37928066], [19467332], [29798891], [26609151].

Limitations and Knowledge Gaps

• Epidemiology: No reliable prevalence or incidence; the entity is ultra-rare and under-ascertained. Sex ratio and age distribution in humans are undefined.
• Human molecular profiling: No published patient transcriptomic/proteomic/metabolomic datasets; mechanistic inference relies on rodent and in vitro models.
• Genotype–phenotype correlation: Not yet quantified — whether missense/hypomorphic vs truncating variants, or variant location within domains, predicts severity or epilepsy risk remains open.
• Penetrance/expressivity: Assumed high penetrance but not formally quantified; drivers of variable expressivity (including the ~50% epilepsy split) are unknown.
• Therapeutics: No disease-modifying therapy; supportive-care evidence is limited to retrospective series and single case reports.
• Missense mechanism: Whether some missense variants act by dominant-negative rather than simple loss-of-function is not fully resolved.

Proposed Follow-up Experiments / Actions

1. Establish an international patient registry to derive prevalence, natural-history, sex-ratio, and genotype–phenotype data.
2. Systematic sleep-EEG surveillance study to define the incidence, timing, and treatment responsiveness of ESES/DEE-SWAS and its link to regression — a directly actionable clinical priority.
3. Patient-iPSC-derived neuron/organoid profiling (transcriptomics, proteomics, electrophysiology) to define human cell-autonomous consequences of PSD-95 loss and to build a drug-screening platform [building on PMID: 42462545].
4. Genotype–phenotype analysis across the expanding ClinVar/cohort variant set, stratified by variant type and PSD-95 domain, to test whether GK-domain or PDZ-domain variants confer distinct risks.
5. Preclinical therapeutic testing in the patient-derived knock-in mouse [PMID: 42565830] — e.g., ASO-mediated upregulation of the wild-type allele, or agents that stabilize the residual PSD, and rigorous evaluation of sex-specific responses.
6. Prospective ASM comparative-effectiveness study for the ESES/DEE-SWAS subtype, given its refractoriness and outsized impact on outcomes.

Evidence types: human clinical [33597769, 38135915, 37386468, 40444229, 42462545, 37525972]; model organism [9853749, 23268962, 20952458, 42565830, 20554866]; in vitro / review [29169997, 40712457]; computational/population-genetic [gnomAD constraint, ClinVar].

Artifacts

• OpenScientist final report
• OpenScientist final report

Reference Validation

Checked with linkml-reference-validator 0.2.1.

All extracted references resolved successfully.