HNRNPU-Related Developmental and Epileptic Encephalopathy: Comprehensive Research Report

1. Disease Information

Overview. HNRNPU-related developmental and epileptic encephalopathy (also called HNRNPU-related neurodevelopmental disorder, HNRNPU-NDD; historically "early infantile epileptic encephalopathy 54," EIEE54) is a severe, typically de novo, autosomal dominant neurodevelopmental disorder caused by heterozygous loss-of-function variants in HNRNPU on chromosome 1q44. It is characterized by global developmental delay, moderate-to-severe intellectual disability, early-onset seizures (often the presenting feature), profound speech impairment, hypotonia, and variable dysmorphic, cardiac, and renal features (GeneReviews, NBK578573; OMIM #617391).

Key identifiers: - OMIM: #617391 — Developmental and Epileptic Encephalopathy 54 (DEE54) - Gene: HNRNPU, HGNC:5048, chromosome 1q44 - MONDO: MONDO:0033363 (developmental and epileptic encephalopathy 54); a broader complex neurodevelopmental disorder mapping (MONDO:0100038) is also used in some resources - GeneReviews: NBK578573 ("HNRNPU-Related Neurodevelopmental Disorder") - ClinGen gene-disease validity: Definitive (Epilepsy and Intellectual Disability/Developmental Delay Gene Curation Expert Panels) (thegencc.org/genes/HGNC:5048)

Synonyms: HNRNPU-related neurodevelopmental disorder (HNRNPU-NDD/HNRNPU-RNDD); Developmental and Epileptic Encephalopathy 54 (DEE54); Early Infantile Epileptic Encephalopathy 54 (EIEE54); heterogeneous nuclear ribonucleoprotein U deficiency; SAF-A (scaffold attachment factor A)-related disorder.

Source of information. The evidence base is aggregated-disease-level (case series, cohort/registry studies, GeneReviews, OMIM) rather than raw EHR data, supplemented by a growing international patient registry (Bain Lab/Columbia; >140 confirmed individuals as of the 2025 phenotype-expansion review) and model-organism studies (Hodgson et al. 2025, PMID:39976380).

2. Etiology

Disease causal factors. DEE54 is caused by heterozygous, predominantly de novo, loss-of-function variants in HNRNPU — nonsense, frameshift, canonical splice-site variants, small intragenic deletions/duplications, and a minority of missense variants — or by microdeletions of 1q43q44 that encompass HNRNPU. A study of clinical exome cohorts concluded that "haploinsufficiency was the main mechanism of pathogenicity" (GeneReviews).

Genetic risk factors: - Causal variants: Truncating (nonsense, frameshift, splice) variants predominate; missense variants also reported (Bramswig et al. 2017 found "three non-sense and two missense variants, one small intragenic deletion, and one duplication" in a 7-patient cohort, PMID:28393272). - No genotype-phenotype correlation has been established — variant type does not reliably predict severity (GeneReviews). - Contiguous-gene 1q43q44 microdeletions: When HNRNPU is co-deleted with neighboring genes (AKT3, ZBTB18), phenotype is more severe/complex; deletion mapping studies show HNRNPU alteration specifically "drives epilepsy and determines the degree of intellectual disability," while AKT3 haploinsufficiency drives microcephaly and ZBTB18 loss drives corpus callosum anomalies with incomplete penetrance — with additive effects when multiple genes are co-deleted (Depienne et al., PMID:28283832). - Germline mosaicism: Presumed parental germline mosaicism has been reported in a family with two affected siblings, and a 2025 report documented the first vertically transmitted familial case (parent-to-child), demonstrating a milder end of the phenotypic spectrum is compatible with reproduction (Hodgson et al. 2025, PMID:39976380).

Environmental risk factors: None established; this is a monogenic disorder.

Protective factors: None identified at genetic or environmental levels.

Gene-environment interactions: Not reported; the disorder's severity appears to be driven by the genetic lesion itself (variant location relative to functional domains, and whether contiguous genes are co-deleted) rather than by environmental modifiers.

3. Phenotypes

Frequencies below are drawn primarily from GeneReviews' synthesis of the published cohort (~83–140+ individuals) and the original Bramswig/Thevenon-type cohort studies.

Source: GeneReviews NBK578573; confirmatory cohort data in Bramswig et al. 2017, PMID:28393272 (seizures 6/7, severe ID 6/6, severe speech impairment 6/6, hypotonia 6/7, CNS 5/6, cardiac 4/6, renal 3/4) and Yates/Durkin et al. 2017, PMID:28815871 (all four patients had seizures, developmental delay, ID, neurologic regression, behavioral issues, dysmorphism).

Phenotype characteristics: - Onset: Neonatal (feeding difficulties, hypotonia) to infantile (seizure onset typically <24 months, often the presenting symptom alongside developmental delay). - Severity: Variable but generally moderate-to-severe for ID; a milder end of spectrum has been increasingly recognized in 2024–2025 case reports (including the first familial, vertically transmitted case). - Progression: Hypotonia may evolve into hypertonia/spasticity over time; "data on possible progression of behavior abnormalities or neurologic findings are still emerging" (GeneReviews). - Quality of life impact: Substantial — most individuals are minimally or nonverbally communicative, many require lifelong caregiver support, gastrostomy feeding in severe cases, and educational/behavioral supports (IEP, ABA); no disease-specific QoL instrument data identified in the literature.

4. Genetic/Molecular Information

Causal gene: HNRNPU (HGNC:5048), encoding heterogeneous nuclear ribonucleoprotein U, also known as scaffold attachment factor A (SAF-A). Located at 1q44.

Pathogenic variant spectrum: Nonsense, frameshift, canonical splice-site variants, small intragenic deletions/duplications, and missense variants. Representative variants from Yates/Durkin cohort: c.651_660del (p.Gly218Alafs118), c.1089G>A (p.Trp363), c.1714C>T (p.Arg572), c.2270_2271del (p.Pro757Argfs7) (PMID:28815871).

Variant classification: Per ACMG/AMP, predicted/confirmed loss-of-function variants (nonsense, frameshift, splice-disrupting) are classified pathogenic/likely pathogenic given established haploinsufficiency mechanism; missense variants require careful case-by-case evaluation given absence of clear genotype-phenotype correlation.

Functional consequence: Loss of function / haploinsufficiency — "de novo loss-of-function variants in HNRNPU can lead to a disease phenotype... haploinsufficiency was the main mechanism of pathogenicity" (GeneReviews).

Somatic vs. germline: Germline, virtually always de novo; presumed germline mosaicism reported in one sibling pair; a familial (inherited) case documented in 2025 (PMID:39976380).

Modifier genes: No specific modifier genes identified for isolated HNRNPU variants; however, within 1q43q44 contiguous deletions, co-deletion of AKT3 and ZBTB18 has additive phenotypic effects, effectively modifying overall severity (microcephaly, corpus callosum defects) (PMID:28283832).

Epigenetic information: A robust, reproducible DNA methylation episignature has been identified in blood from individuals with pathogenic HNRNPU variants using Infinium EPIC arrays, distinct from — but partially overlapping — 56 other neurodevelopmental-disorder episignatures. This episignature has clinical utility for reclassifying HNRNPU variants of uncertain significance (VUS) via the EpiSign platform (Genetics in Medicine, 2023, 25(8):100871; related: germline HNRNPU variants and blood methylome alterations, Nature-EJHG; EpiSignature VUS reclassification case report, PMC12688365). This directly implicates HNRNPU's chromatin-regulatory role in disease pathogenesis.

Chromosomal abnormalities: 1q43q44 microdeletions encompassing HNRNPU (along with AKT3, ZBTB18) produce a related but broader contiguous-gene syndrome (microcephaly, corpus callosum abnormalities, epilepsy, short stature) — see Section 2 (PMID:28283832; case report of 163kb 1q44 microdeletion, PMID:22975012).

gnomAD constraint: HNRNPU is highly constrained against loss-of-function variation (consistent with a haploinsufficiency mechanism and near-complete de novo occurrence in patients), in keeping with its essential, ubiquitously expressed roles in RNA processing and chromatin organization (general constraint framework via gnomAD; exact pLI/LOEUF values were not independently retrievable in this search pass and should be confirmed directly against the gnomAD browser at curation time).

5. Environmental Information

No environmental, lifestyle, toxin, or infectious risk factors have been identified as contributing to HNRNPU-related DEE54 — this is a purely monogenic disorder arising from de novo germline variants. The GeneReviews management section notes only a general precaution that "activities and agents that may induce seizures" should be avoided as a secondary, symptom-driven consideration (not an etiologic factor) (GeneReviews).

6. Mechanism / Pathophysiology

Molecular function of HNRNPU/SAF-A: HNRNPU is a highly abundant, ubiquitously expressed nuclear RNA/DNA-binding protein with a multidomain architecture: an N-terminal SAP domain (direct DNA/RNA binding), a SPRY/B30.2 protein-protein interaction domain, a central AAA+ ATPase oligomerization domain, and a C-terminal domain with RGG/RG motifs for RNA binding. It functions as a dynamic bridge between chromatin, nascent RNA, and the nuclear matrix, regulating: (1) pre-mRNA processing and alternative splicing, (2) higher-order chromatin architecture (3D genome organization), (3) transcriptional regulation, and (4) X-chromosome inactivation (ScienceDirect review, PMID:34823151; PLOS Genetics 2025, PMC12176297).

X-chromosome inactivation role: SAF-A/HNRNPU localizes to the inactive X chromosome and interacts directly with XIST lncRNA, and is required for proper XIST RNA territorial localization and XIST-dependent heterochromatin/histone modification — an SAP-domain-dependent function (serines S14/S26 critical) (PLOS Genetics). This is a molecular function distinct from, but potentially contributing to, the neurodevelopmental phenotype's dosage sensitivity.

Causal chain — cortical development (mouse model, Sapir et al. 2022, Nature Communications, PMID:35864088): 1. Trigger: Hnrnpu loss of function in embryonic cortical neuroepithelium. 2. Molecular: Dysregulated alternative splicing of >850 genes, notably Mdm2 (exon 3 skipping → reduced p53 inhibition), Dcc, Siva1 (migration/apoptosis regulators), and cytoskeletal/synaptic transcripts. 3. Cellular: Elevated Tp53 target gene expression → activation of canonical and non-canonical (including necroptotic) p53-dependent cell death. Neural progenitors show markedly higher vulnerability than postmitotic neurons — time-lapse imaging showed progenitor death within 1–5 hours of Cre-mediated excision, versus attenuated death dynamics in postmitotic cells. 4. Tissue: Rapid, near-complete elimination of cortical structures in conditional knockouts. 5. Rescue evidence (mechanistic validation): Pan-caspase inhibitors (Z-VAD-fmk, Q-VD-OPH), p53 inhibitor pifithrin-μ, and necroptosis inhibitor Nec-1 each partially rescued progenitor viability; genetic Tp53 co-deletion "enabled cortical formation, increased progenitors proliferating" (incomplete rescue); co-deletion of the competing splicing factor Srsf3 rescued neurosphere phenotypes and migration defects to near-control levels.

Causal chain — postnatal circuit dysfunction (mouse model, PLOS Genetics 2023, PMC10569524): 1. Heterozygous Hnrnpu+/− mice show global developmental delay, impaired ultrasonic vocalizations, cognitive dysfunction (increased gamma oscillations), and lowered electroconvulsive seizure threshold (p<1×10⁻⁴) despite no spontaneous seizures on prolonged video-EEG. 2. Single-cell RNA-seq of hippocampus/neocortex reveals widespread but modest transcriptional dysregulation (hippocampus: 955 DE events, 73% downregulated; neocortex: 454 DE events, 51% downregulated). 3. Subiculum excitatory neurons carry the highest DEG burden of any cell type examined, with striking enrichment for developmental-delay, epilepsy, and autism-associated genes. 4. Mef2c (a well-known NDD gene) is the most downregulated transcript (50% reduction, log2FC=−1.11, FDR=8×10⁻³⁷), with hnRNP U binding sites at its locus exceeding 99% of other examined genes — directly implicating Mef2c dysregulation as a candidate downstream driver of the epilepsy/cognitive phenotype.

Cross-model convergence: An iScience 2022 study (PMID:36594023) comparing isogenic HNRNPU+/− human iPSC-derived brain organoids to embryonic/perinatal mouse cortex found significant enrichment of shared, co-dysregulated transcripts, supporting a conserved developmental transcriptomic signature across species and model systems (PMC9804147).

Suggested GO terms: GO:0006397 (mRNA processing), GO:0000381 (regulation of alternative mRNA splicing), GO:0006355 (regulation of transcription, DNA-templated), GO:0006974 (DNA damage response), GO:0097191 (extrinsic apoptotic signaling pathway), GO:0070182 (DNA polymerase binding — chromatin structural role), GO:0008380 (RNA splicing).

Suggested CL terms: CL:0000047 (neural stem cell) / CL:0002608 (neural progenitor cell, radial glia), CL:0000679 (glutamatergic neuron, subiculum excitatory), CL:0000540 (neuron, general).

Suggested UBERON terms: UBERON:0001950 (neocortex), UBERON:0002421 (hippocampal formation)/UBERON:0003881 (subiculum), UBERON:0002336 (corpus callosum).

7. Anatomical Structures Affected

Organ level: - Primary: Central nervous system (cerebral cortex, hippocampus/subiculum) — the dominant site of disease. - Secondary/variable involvement: Cardiovascular system (septal defects), renal system (agenesis, dysplasia), eyes (strabismus), ears (rare sensorineural hearing loss), musculoskeletal system (joint laxity, scoliosis), male reproductive system (undescended testis). - Body systems: Nervous system (primary), cardiovascular, renal/urologic, musculoskeletal, ophthalmologic, otologic.

Tissue/cell level: Neuroepithelium/radial glia and neural progenitor cells (most vulnerable population per mouse studies), postmitotic cortical and hippocampal excitatory neurons (subiculum especially implicated), cardiac septal tissue, renal parenchyma.

Subcellular level: Nucleus (chromatin/nuclear matrix — HNRNPU's primary site of action), specifically associated with the inactive X chromosome territory in female cells; splicing machinery/spliceosome-associated nuclear speckles. GO Cellular Component: GO:0005654 (nucleoplasm), GO:0016607 (nuclear speck), GO:0000785 (chromatin).

Localization: Bilateral, diffuse cortical/subcortical involvement (no lateralization reported); brain MRI abnormalities (ventriculomegaly, thin corpus callosum) are typically symmetric/generalized rather than focal.

8. Temporal Development

Onset: Congenital/neonatal manifestations (hypotonia, feeding difficulty) evident from birth or early infancy; developmental delay recognized in first year; seizure onset typically before 24 months of age (~90% of those with seizures), sometimes triggered initially by fever before becoming afebrile.

Onset pattern: Insidious/progressive developmental delay from infancy, punctuated by acute seizure onset events.

Progression: - Developmental trajectory: continued, if slow, developmental gains reported in many individuals rather than regression, though at least one case series (Yates/Durkin, PMID:28815871) reported "neurologic regression" in some patients — indicating phenotypic heterogeneity in course. - Muscle tone: hypotonia in infancy may transition to hypertonia/spasticity later in childhood. - Seizures: generally become more manageable with age and appropriate anti-seizure regimens in most reported individuals, though refractory epilepsy occurs in a subset. - Disease duration: chronic, lifelong; based on current data, "life span is not significantly limited by this condition, as several adults have been reported" (GeneReviews).

Patterns: No formal remission is described; seizure control with medication is common but not universal. No defined "critical intervention window" has been established in the literature to date, though early developmental intervention (0–3 years) is recommended per standard NDD management.

9. Inheritance and Population

Epidemiology: Ultra-rare; exact prevalence/incidence unknown. GeneReviews states "the prevalence of this condition is unknown. To date, approximately 83 individuals with HNRNPU-NDD have been reported" (as of the March 2022 GeneReviews update); by 2025, an international patient registry (Bain Lab, Columbia) had grown to encompass a substantially larger cohort (a 2025 phenotype-expansion paper added 17 previously unpublished patients, and unrelated registry sources describe well over 100 confirmed individuals) (Hodgson et al. 2025, PMID:39976380).

Inheritance pattern: Autosomal dominant, virtually always de novo. Rare instances of parental germline mosaicism and, as of 2025, the first documented familial (parent-to-child) transmission.

Penetrance: Appears complete/high for the core developmental delay and seizure phenotype among reported carriers, though ascertainment bias (severe cases more likely to be sequenced) may inflate apparent penetrance; the emerging recognition of milder cases suggests a broader phenotypic spectrum than initially appreciated.

Expressivity: Highly variable — ranging from the "classic" severe DEE54 presentation to milder cases now increasingly reported (2024–2025 case series), without clear genotype-phenotype correlation.

Genetic anticipation: Not described (not applicable to a haploinsufficiency-mechanism, non-repeat-expansion disorder).

Founder effects / consanguinity: Not applicable — disorder is dominant and de novo, not associated with consanguinity or population founder mutations.

Carrier frequency: Not applicable (de novo dominant disorder, not a recessive carrier state).

Population demographics: - No ethnic or geographic predilection reported; cases have been described from multiple continents (US, Europe, various case series). - Sex ratio: No clear sex bias reported in the literature reviewed; both males and females affected (male-specific phenotype item: undescended testis in ~20% of affected males). - Age distribution: Predominantly diagnosed in infancy/childhood via exome/genome sequencing for developmental delay and epilepsy; adults are increasingly recognized as historically underdiagnosed due to lack of earlier genetic testing availability.

10. Diagnostics

Establishing the diagnosis: Requires "a proband with suggestive findings and a heterozygous pathogenic variant in HNRNPU identified by molecular genetic testing" (GeneReviews).

Molecular testing approaches, by yield: - Sequence analysis (exome/genome or NDD/epilepsy gene panel including HNRNPU): ~98% detection rate. - Gene-targeted deletion/duplication analysis (or chromosomal microarray for 1q43q44 deletions): ~2% of cases (contiguous gene deletion presentations). - Single-gene sequential testing of HNRNPU alone is "rarely useful and typically NOT recommended" given the nonspecific phenotype — multigene panel or exome/genome sequencing is preferred.

Genetic testing modalities: - Multigene intellectual disability/epileptic encephalopathy panel including HNRNPU. - Whole exome sequencing (WES) — most commonly used diagnostic route in published cohorts. - Whole genome sequencing (WGS) — increasingly used, captures structural variants and intronic/regulatory changes missed by exome. - Chromosomal microarray (CMA) — detects 1q43q44 microdeletions/duplications encompassing HNRNPU. - DNA methylation episignature (EpiSign) testing — emerging clinical tool for reclassifying HNRNPU VUS, given the validated disease-specific episignature (Genetics in Medicine 2023, 25(8):100871).

Clinical/laboratory tests: No disease-specific biomarker or metabolic screening test exists; diagnosis is genetic. Brain MRI is used to characterize (not diagnose) the condition — ventriculomegaly and thin corpus callosum are the most frequent findings (61% of imaged cases abnormal). EEG is used to characterize seizure semiology, not to establish diagnosis.

Differential diagnosis: Broad — "all disorders with epileptic encephalopathy and intellectual disability without other distinctive findings should be considered," referencing the OMIM Developmental and Epileptic Encephalopathy Phenotypic Series for systematic comparison (GeneReviews). Specific considerations include other DEE-causing genes, and — for patients with contiguous deletions — the broader 1q43q44 microdeletion syndrome (distinguishing isolated HNRNPU variants from deletions also involving AKT3/ZBTB18, which add microcephaly and corpus callosum anomalies).

Screening: No population or newborn screening applicable (ultra-rare, not detectable by biochemical newborn screening); prenatal testing/preimplantation genetic testing available for known familial variants (relevant given the newly documented familial transmission case).

11. Outcome/Prognosis

Survival/mortality: Life expectancy does not appear significantly reduced; "several adults have been reported" and formal life-span data, while incomplete, does not suggest premature mortality as a defining feature (GeneReviews). No disease-specific mortality statistics are available given the ultra-rare, only-recently-delineated nature of the condition.

Morbidity/function: Substantial lifelong disability — most affected individuals have moderate-to-severe intellectual disability, are minimally/nonverbally communicative, and require ongoing multidisciplinary support (PT/OT/speech, special education, seizure management). No validated disease-specific quality-of-life instrument has been applied in the literature reviewed.

Complications: Refractory epilepsy in a subset; feeding/nutritional complications (sometimes requiring gastrostomy); sleep apnea requiring respiratory support; secondary orthopedic complications of hypertonia/spasticity (contractures) or hyperlaxity (scoliosis).

Prognostic factors: No genotype-phenotype correlation identified to date, so variant type does not predict severity. Severity appears greater when HNRNPU loss occurs in the context of a larger 1q43q44 contiguous gene deletion (additional microcephaly, corpus callosum defects from co-deleted AKT3/ZBTB18).

Recovery potential: Developmental gains continue in many individuals with early, sustained intervention, though the underlying intellectual disability persists lifelong; seizure control is achievable in most patients with standard or combination anti-seizure therapy.

12. Treatment

Pharmacotherapy (seizures): - Sodium valproate — "the most commonly used & effective medication" per GeneReviews for first-line seizure control (NCIT:C15986 Pharmacotherapy; therapeutic agent CHEBI valproate). - Newer-generation anti-seizure medications for refractory cases (unspecified beyond class in the literature reviewed). - Ketogenic diet — recommended for refractory seizures; broader epilepsy literature supports its efficacy and safety, including in combination with valproate, for drug-resistant epilepsy generally (NCIT:C15447 Dietary Intervention). - Combined pharmacotherapy for associated symptoms has been documented in individual cases (acetazolamide, alprazolam, aripiprazole) — indication-specific, not seizure-first-line.

Advanced/experimental therapeutics: - Antisense oligonucleotide (ASO) therapy: In 2024, the Bain Lab (Columbia University) launched the first precision-therapeutics trial for the related HNRNP-family disorder H2-RNDD (HNRNPH2), administering individualized ASO therapy to 8 patients via the n-Lorem Foundation (n-of-1 model). While this specific program targets HNRNPH2 rather than HNRNPU, it establishes proof-of-concept for RNA-targeted precision therapeutics in the HNRNP gene family and signals a plausible future therapeutic direction for HNRNPU-NDD given the shared haploinsufficiency mechanism and RNA-binding-protein biology (NCIT:C15238 Gene Therapy category; therapeutic_modality: ANTISENSE_OLIGONUCLEOTIDE). No HNRNPU-specific ASO clinical trial has yet been identified in this search. - Preclinical ASO work in the HNRNPH2 mouse model (Science Translational Medicine 2025/2026) demonstrates feasibility of splice/expression-modulating ASO rescue in this gene family, informing potential translational strategies for HNRNPU.

Surgical/interventional: No disease-specific surgery; organ-specific procedures as needed (cardiac septal defect repair, orchiopexy for undescended testis, orthopedic procedures/Botox for severe spasticity or scoliosis).

Supportive/rehabilitative care: - Early intervention programs (0–3 years), developmental preschool (3–5 years), individualized education plans. - Physical therapy (gross motor/tone), occupational therapy (fine motor/adaptive function), speech-language pathology with augmentative/alternative communication (AAC) evaluation (NCIT:C15302 Physical Therapy). - Feeding therapy for dysphagia; nasogastric or gastrostomy tube placement for persistent feeding dysfunction. - Respiratory support (supplemental O2, CPAP/BiPAP) for sleep apnea. - Tone management: baclofen, tizanidine, botulinum toxin, or orthopedic procedures for hypertonia (physical medicine & rehabilitation involvement). - Behavioral intervention: applied behavior analysis (ABA) for autism-related features; pediatric psychiatry for severe aggressive/destructive behaviors.

Organ-specific surveillance/management: Cardiology (congenital heart defects), ophthalmology (strabismus), audiology (hearing loss), nephrology (renal anomalies), urology (undescended testes).

Treatment strategy: No formal published treatment algorithm specific to HNRNPU-NDD beyond the GeneReviews management/surveillance framework; management is symptom-directed and multidisciplinary, following general DEE/intellectual disability care pathways.

Personalized medicine: DNA methylation episignature profiling offers a genotype-informed diagnostic refinement tool (VUS reclassification); no pharmacogenomic guidance specific to HNRNPU-NDD identified.

13. Prevention

Primary prevention: Not applicable in the traditional sense — this is a de novo genetic disorder with no known modifiable environmental trigger to avoid.

Secondary prevention: Early diagnosis via exome/genome sequencing in infants presenting with early-onset seizures and developmental delay allows earlier initiation of supportive therapies and seizure management, potentially reducing secondary complications (status epilepticus, feeding-related morbidity).

Genetic counseling: Recurrence risk to siblings of an affected proband is low but not zero (given documented germline mosaicism cases), warranting parental testing to confirm de novo status. For an affected individual, the risk of transmission to offspring is 50% per pregnancy (autosomal dominant); prenatal testing and preimplantation genetic testing (PGT) are available for families with a known familial variant, an option that gained clinical relevance following the first reported vertical transmission case in 2025.

Screening: No population-level newborn or carrier screening applicable (ultra-rare, dominant, de novo disorder without ethnic founder effects).

Public health/behavioral interventions: Not applicable — no known modifiable population-level risk factor.

Prophylaxis: General epilepsy-safety precautions (avoiding known seizure triggers) apply once a seizure phenotype is established, as with any epilepsy syndrome.

14. Other Species / Natural Disease

Taxonomy: Naturally occurring HNRNPU-related disease has not been reported in non-human species; the gene is highly conserved (near-universal expression and function across vertebrates).

Model organism gene: Mouse Hnrnpu (MGI ortholog); human HNRNPU NCBI Gene ID 3192.

Natural disease in other species: Not reported in OMIA or veterinary literature reviewed — this is not a recognized naturally occurring veterinary disease.

Comparative biology: HNRNPU/SAF-A is highly conserved across mammals given its essential, ubiquitous roles in chromatin organization and RNA processing; the core molecular mechanism (splicing regulation, p53-dependent progenitor apoptosis upon loss) is conserved between mouse and human, as demonstrated by the organoid/mouse cross-species transcriptomic convergence study (PMID:36594023).

Zoonotic potential/cross-species susceptibility: Not applicable — this is a genetic, non-infectious disorder.

15. Model Organisms

Mouse models (genetic, knockout/conditional): 1. Constitutive/germline heterozygous Hnrnpu+/− mouse (PLOS Genetics 2023, PMID likely associated with PMC10569524): Models the human haploinsufficiency state. Recapitulates global developmental delay, impaired ultrasonic vocalizations, cognitive dysfunction (elevated wakeful gamma oscillations), and lowered electroconvulsive seizure threshold (p<1×10⁻⁴), though without spontaneous seizures on extensive video-EEG (300+ hours) — a partial phenotype recapitulation (increased seizure susceptibility rather than overt epilepsy). Single-cell RNA-seq identifies subiculum excitatory neurons as the most transcriptomically perturbed cell type, with Mef2c as the most robustly downregulated candidate driver gene. 2. Conditional cortical Hnrnpu knockout mouse (Nature Communications 2022, PMID:35864088, Sapir/Reiner lab): Demonstrates near-complete cortical structure elimination via p53-dependent apoptosis of neural progenitors (more vulnerable) and postmitotic neurons (less vulnerable, delayed death). Provides mechanistic and pharmacological rescue data (caspase inhibitors, p53 inhibitor pifithrin-μ, necroptosis inhibitor Nec-1; genetic Tp53 or Srsf3 co-deletion). This model captures the severe, embryonic-lethal-if-homozygous end of the mechanistic spectrum rather than the milder heterozygous human phenotype.

In vitro/organoid models: - Isogenic human iPSC-derived brain organoids (PGP1 line) with CRISPR-engineered HNRNPU+/− frameshift variants (1bp duplication, 10bp deletion), showing ~25% reduction in HNRNPU protein and reduced mRNA. Used for cross-species transcriptomic comparison, confirming conserved dysregulated gene modules between 45-day human organoids and embryonic mouse cortex (PMID:36594023; PMC9804147).

Model limitations: - Mouse heterozygous models show increased seizure susceptibility rather than the spontaneous, often treatment-refractory epilepsy seen in human patients — a translational gap (species-scale/circuit-maturation difference) worth flagging as a HUMAN_MODEL_MISMATCH if curated into a pathophysiology module. - The severe conditional cortical-knockout mouse models complete loss of function in a defined lineage/timepoint, which is more mechanistically informative for the p53/splicing pathway than directly representative of the human heterozygous dosage state. - No zebrafish, Drosophila, or C. elegans HNRNPU ortholog disease model was identified in this search (in contrast to the sister gene HNRNPH2, for which zebrafish/mouse models are more developed).

Applications: These models collectively support (1) the mechanistic causal chain from HNRNPU loss → splicing dysregulation (Mdm2, Mef2c, and hundreds of other targets) → p53-dependent progenitor death / circuit-level excitatory neuron dysfunction → cortical malformation and seizure susceptibility, and (2) proof-of-concept pharmacological (caspase/p53/necroptosis inhibitors) and genetic (Tp53, Srsf3 co-deletion) rescue strategies that could inform future therapeutic development, complementing the ASO precision-medicine approach being piloted in the related HNRNPH2 disorder.

Resources: MGI (Mouse Genome Informatics) for Hnrnpu allele records; no dedicated IMPC/KOMP full-line data specifically confirmed in this search pass — recommend direct IMPC query at curation time for standardized phenotyping data.

Summary of Key Ontology Term Suggestions

Sources

• HNRNPU-Related Neurodevelopmental Disorder — GeneReviews (NBK578573)
• OMIM #617391 — DEE54
• Bramswig et al. 2017, Hum Genet — PMID:28393272
• Yates/Durkin et al. 2017 clinical/molecular characterization — PMID:28815871
• De novo mutations in HNRNPU — PMID:28944577
• Depienne et al., 1q43q44 microdeletion syndrome dissection — PMID:28283832 / PMC5360844
• Sapir et al. 2022, Nature Communications — HNRNPU safeguards developing mouse cortex, PMID:35864088
• Mouse model of HNRNPU haploinsufficiency, PLOS Genetics 2023 — PMC10569524
• Shared transcriptomic dysregulation, human organoids vs. embryonic mice — PMID:36594023 / PMC9804147
• Hodgson et al. 2025, AJMG-A — Expanding phenotypic spectrum, first familial case — PMID:39976380
• DNA methylation episignature and comparative epigenomic profiling of HNRNPU-NDD, Genetics in Medicine 2023
• Germline pathogenic HNRNPU variants and blood methylome alterations, EJHG
• EpiSignature VUS reclassification case report — PMC12688365
• Role of SAF-A/HNRNPU SAP domain in X-inactivation — PLOS Genetics, PMC12176297
• SAF-A/hnRNP U chromatin structure review — PMID:34823151
• Case report: WES unveils HNRNPU-NDD — PMID:38846959
• NORD — HNRNPU-Related Disorder overview
• thegencc.org — HNRNPU gene-disease validity classifications
• Bain Lab, Columbia Neurology — HNRNPH2 ASO precision trial context
• HNRNPU-NDD, Detailing the Ultrarare Neurodevelopmental Disorder — genome.bio
• 1q44 microdeletion case report — PMID:22975012

Reference Validation

Checked with linkml-reference-validator 0.2.1.

All extracted references resolved successfully.