VPS4A-Related Neurodevelopmental Syndrome

1. Disease Information

Overview

VPS4A-related neurodevelopmental syndrome — formally CIMDAG syndrome — is an ultra-rare multisystem Mendelian disorder caused by missense variants in VPS4A, which encodes the AAA-ATPase that disassembles and recycles the ESCRT-III membrane-remodeling machinery. The clinical core is a severe, congenital-onset neurodevelopmental phenotype (profound global developmental delay, severe primary microcephaly, dystonia, structural brain malformation) combined with a transfusion-dependent congenital dyserythropoietic/hemolytic anemia, congenital cataracts and retinal dystrophy, and severe growth retardation. It is one of the very few human diseases in which a core ESCRT component is the primary genetic lesion, and it is mechanistically distinctive in that pathogenesis is dominant-negative poisoning of a hexameric enzyme, not haploinsufficiency.

The acronym CIMDAG expands to: cerebellar hypoplasia and cataracts, intellectual disability, congenital microcephaly, dystonia and dyserythropoietic anaemia, growth retardation (Unique/Rare Chromosome Disorder Support Group, 2024).

Key identifiers

Cross-references verified via OLS4 (MONDO term record) and HGNC REST.

Synonyms and alternative names

• CIMDAG syndrome (preferred acronym)
• VPS4A-related neurodevelopmental disorder (VPS4A-ND) — the label used by the Unique family guide
• Cerebellar hypoplasia–intellectual disability–congenital microcephaly–dystonia–anemia–growth retardation syndrome (MONDO/Orphanet long form)
• Syndromic congenital dyserythropoietic anemia due to VPS4A (hematology literature framing; Seu et al.)
• Syndromic congenital hemolytic anemia with neurodevelopmental impairment (Lunati et al. framing)
• Gene aliases relevant to older literature: SKD1, SKD1A, SKD2, VPS4, VPS4-1, hVPS4

Nature of the evidence base

Entirely individual-patient derived, not aggregated. The whole disease concept rests on ~11 published probands described in four primary reports plus deep single-patient functional workups. There is no EHR-based cohort, no registry-derived prevalence, and no natural history study. The only registry involvement is indirect: three of the probands were ascertained through the Congenital Dyserythropoietic Anemia Registry (CDAR, NCT02964494, Cincinnati Children's, observational, actively recruiting), which is a CDA-wide registry rather than a VPS4A-specific one. Aggregated resources (OMIM, Orphanet, HPO annotation file) are themselves derived from the same handful of case reports — so HPO frequencies such as "6/6" and "4/5" are literally patient counts from Rodger et al., not epidemiologic estimates.

2. Etiology

Primary causal factor

Monogenic. Heterozygous de novo missense variants in the AAA-ATPase domain of VPS4A are the dominant cause (autosomal dominant, dominant-negative mechanism). A minority of cases are biallelic (homozygous) missense, arising in consanguineous families and acting through a distinct, milder loss-of-function-like route.

Rodger et al. (PMID:33186545) frame the causal claim as [abstract]:

"Here we describe six unrelated individuals with de novo missense variants affecting the ATPase domain of VPS4A, a critical enzyme regulating ESCRT function. Probands had structural brain abnormalities, severe neurodevelopmental delay, cataracts, growth impairment, and anemia."

Genetic risk factors

• Causal variants: see §4. All published pathogenic alleles are missense; no truncating, splice, or structural pathogenic alleles have been reported.
• Position-specific risk: the ATP-binding pocket arginine finger Arg284 is a mutational hotspot (5 of 11 reported probands), and pore-loop-1 residues Gly203/Glu206 form a second hotspot cluster.
• Consanguinity is the relevant risk factor for the recessive form (Seu proband 3, homozygous p.Ala28Val, both parents unaffected heterozygous carriers; Gupta 2026 Nepalese proband, homozygous p.Arg288Gln).
• Advanced paternal age — plausible but not demonstrated for this gene; no data. Treat as unknown.
• Modifier genes: none identified. VPS4B is a mechanistically obvious candidate modifier (see §4, §6): Seu et al. observed [body/derived] "significantly increased VPS4B expression in proband 3, indicating that VPS4B may partially, but not adequately, compensate for the VPS4A loss of function in the homozygous p.Ala28Val variant." No germline VPS4B modifier alleles have been tested.
• Susceptibility loci / GWAS: none. No GWAS Catalog signal is relevant to this Mendelian phenotype.

Environmental risk factors

None identified. As a de novo germline missense disorder, no toxin, infectious, occupational, nutritional, or lifestyle exposure has been implicated in causation. The Unique guide states explicitly to families: "This happens naturally and is not due to the parents' diet, environment or lifestyle." Sex is not a risk factor (see §9). This should be recorded as NO_EVIDENCE rather than left blank.

Protective factors

• Genetic: no protective alleles known. The only relevant observation is paralog buffering — an intact VPS4B is insufficient to prevent disease when VPS4A is mutated (see PMID:38687820), which argues against paralog dosage as a protective modifier in the dominant form but leaves it plausible in the recessive form.
• Environmental: none. Secondary/tertiary protective interventions (iron chelation, cataract extraction, seizure control) modify morbidity, not disease occurrence — record under §12/§13, not here.

Gene–environment interactions

No documented GxE. One clinically actionable pseudo-interaction deserves note: iron overload in this disease is disproportionate to transfusion burden because ineffective erythropoiesis itself drives hyperabsorption — Unique states that "Defective red blood cell development (dyserythropoiesis) and ineffective red blood cell production (ineffective erythropoiesis) may cause iron overload disproportionate to the number of transfusions." This is a genotype-driven amplification of a treatment exposure (transfusional iron) rather than a true environmental interaction, but it belongs in the causal graph.

3. Phenotypes

3.1 HPO annotation set (source: HPO annotation file for OMIM:619273, retrieved via ontology.jax.org API)

Frequencies are literal patient-count fractions from the source publications, not population estimates. Inheritance annotation: HP:0000006 Autosomal dominant inheritance.

Neurodevelopmental / neurological (core)

Brain structural (neuroimaging)

Note the discrepancy: the HPO annotation file records cerebellar hypoplasia as 2/6, but the Rodger full text states cerebellar hypoplasia in five of six probands. Seu et al. independently describe [body/derived] "moderate to marked deficiency of hemispheric white matter with global cerebral volume loss, thin corpus callosum, and atrophy of the cerebellum and the pons, suggestive of a neurodegenerative syndrome." For a dismech entry, curate cerebellar hypoplasia at the higher (full-text) frequency with the Rodger PMID, and flag the HPOA discrepancy in a discussions block.

Ophthalmologic

One proband had Leber congenital amaurosis and/or cortical blindness (Seu proband 1) — consider HP:0000548 (Leber congenital amaurosis is HP:0000548? verify; MONDO:0018998 is the disease-level term) or curate as retinal dystrophy + cortical visual impairment.

Hematologic

Underrepresented in HPOA (the annotation file returned no anemia term, a genuine gap, because the anemia was characterized primarily in the companion paper, Seu et al., which is annotated to the same OMIM entry only partially). Curate from primary literature:

Rodger et al. report anemia in 3/6, dyserythropoietic in 2 [body/derived]: "Three subjects had anemia, which was characterized as dyserythropoietic in two." Combining cohorts (Rodger 3/6 + Seu 3/3 + Lunati 1/1), anemia is present in roughly 7/10 published patients — i.e. FREQUENT in HPO frequency terms. Note the Lunati case demonstrates the anemia can be hemolytic without overt dyserythropoiesis, so the hematologic phenotype should be curated as a spectrum, not a single entity.

Growth and gastrointestinal

Other systems

Unique additionally reports, from family experience (not peer-reviewed case series, so lower evidence tier): dental abnormalities (late eruption, overcrowding, weak enamel, bruxism, eruption cysts), talipes/club foot, scoliosis/kyphosis, peripheral acrocyanosis and cold extremities, neonatal jaundice, and small hands and feet.

One Seu proband had chronic kidney disease stage II–III (HP:0012622) and one had macrocephaly rather than microcephaly (the homozygous p.Ala28Val proband) — an important genotype–phenotype exception.

3.2 Phenotype characteristics

• Age of onset: congenital / neonatal for essentially all core features. Microcephaly is present at birth (primary microcephaly in half). Hypotonia is "profound neonatal-onset." Cataracts are congenital in the majority. Anemia is "apparent from early infancy."
• Severity: uniformly severe to profound for the neurodevelopmental domain in the heterozygous dominant-negative genotypes. The homozygous p.Ala28Val (MIT-domain) genotype is milder — Seu et al. note [body/derived] "the phenotype is somewhat milder than the clinical picture secondary to de novo heterozygous variants in the large ATPase domain which appear to exert a more detrimental, dominant-negative effect."
• Progression: predominantly static/developmental rather than degenerative in most descriptions, but imaging language in Seu et al. is explicitly "suggestive of a neurodegenerative syndrome" (progressive cerebral volume loss, pontocerebellar atrophy). This is an unresolved question — curate as a KNOWLEDGE_GAP discussion: is CIMDAG a static malformation syndrome, a progressive neurodegeneration, or both? Anemia is chronic and lifelong; feeding difficulties often improve after early childhood per Unique.
• Frequency across individuals: given above per phenotype.

3.3 Quality-of-life impact (per phenotype)

No EQ-5D, PedsQL, SF-36, or PROMIS data exist for this disorder — this is a genuine and complete gap. Qualitative, family-reported impacts from the Unique guide:

• Motor/dystonia: dominant driver of disability — independent sitting, standing, or walking "may not be achieved"; requires orthotics, callipers, wheelchair, hydrotherapy; caregiver-dependent for all mobility and transfers.
• Speech/communication: profound; many remain non-verbal, requiring AAC (pictograms, gestures, high-tech aided communication). Families describe first successful interaction via auditory/tactile channels: "Hearing and making funny noises paired with touching was our first possibility to interact and bring a smile and later laughter to our son's face."
• Sleep disturbance (4/4): high family burden; frequently medicated.
• Feeding: NGT/PEG in a substantial subset; aspiration risk.
• Anemia: transfusion every 2 weeks to 6 months → recurrent hospital contact, venous access burden, and lifelong chelation.
• Seizures: "Seizures can cause a lot of worry for families and can be frightening to observe."
• Vision: cataract + retinal dystrophy compounds developmental deprivation; cataract is surgically correctable ("if a cataract is diagnosed it can be easily removed with a small surgery").
• Temperament: counterbalancing positive — "most have a happy disposition."

4. Genetic / Molecular Information

Causal gene

VPS4A — vacuolar protein sorting 4 homolog A. HGNC:13488 · NCBI Gene 27183 · Ensembl ENSG00000132612 · UniProt Q9UN37 · OMIM 609982 · RefSeq NM_013245.3 / NP_037377 · chromosome 16q22.1*.

Protein: 437 aa, 48,898 Da. Type I AAA+ ATPase (EC 3.6.4.6). Domain architecture (UniProt Q9UN37): - MIT domain, aa 2–80 — "Microtubule Interacting and Trafficking" domain; binds ESCRT-III MIM motifs (CHMP1A/B, CHMP2A/B, IST1). Key ESCRT-III contact residues include Val-13 and Leu-64. - AAA+ ATPase cassette (central) — Walker A P-loop aa 167–174 (ATP binding), catalytic Lys-173, Walker B glutamate Glu-228 (the residue mutated in the canonical experimental dominant negative E228Q), pore loop 1 containing Trp201/Leu202/Gly203 and Glu206, and the arginine finger Arg284. - β-domain / Vps4_C — substrate and VTA1 co-factor engagement.

Pathogenic variants (all reported cases)

Variants of uncertain significance reported and functionally excluded by Rodger et al.: c.502C>T (p.Pro168Ser) and c.1009A>G (p.Ile337Val) — associated with non-specific intellectual disability, deemed unlikely pathogenic. UniProt additionally flags a variant at position 193 as a VUS with a CIMDAG-like presentation.

Variant class: exclusively missense. No pathogenic PTVs, splice, CNV, or structural variants have been reported for this phenotype.

Population allele frequency: the pathogenic alleles are absent from gnomAD, ExAC, and TOPMed and fall "in regions highly constrained for variation in control populations" [body/derived, Rodger et al.]. I was unable to retrieve numeric gnomAD constraint metrics (pLI/LOEUF) — the browser is JS-rendered and the GraphQL endpoint requires POST; record this as a data gap rather than inventing a value. The qualitatively decisive constraint statement is the opposite of what one might expect and is central to the mechanism (below).

Somatic vs germline: all disease-causing variants are germline. VPS4A does appear in somatic cancer contexts — it is a well-known synthetic-lethal partner of VPS4B in 18q-deleted cancers, and recent work implicates it in oxaliplatin resistance in colorectal cancer (PMID:40558556), radioresistance via MYO1C in esophageal squamous carcinoma (PMID:41654990), EPHB2-driven autophagy in oral squamous carcinoma (PMID:40017157), and STING-driven anti-tumor immunity when VPS4 is pharmacologically targeted in rhabdomyosarcoma (PMID:42032367). None of this is relevant to CIMDAG pathogenesis and should not be conflated with it in a disease entry — but the VPS4A/VPS4B synthetic-lethal literature is the reason a VPS4 inhibitor chemical toolkit exists at all.

Functional consequence: dominant negative, explicitly not haploinsufficiency

This is the single most important molecular fact for the entry. Rodger et al. state [body/derived]:

"Multiple heterozygous VPS4A loss-of-function mutations are present in general population databases, indicating that a haploinsufficiency mechanism is unlikely."

and

"Our data and published evidence point to the p.Glu206Lys, p.Arg284Trp, and p.Arg284Gly mutants having a dominant-negative effect."

The structural logic is hexamer poisoning [body/derived]:

"As VPS4A protein stability is unaltered in proband cells, assuming equal expression of wild-type and mutant VPS4A we expect that a large majority of VPS4A hexamers will have impaired function as they will contain at least one mutant subunit."

with the proximate defect being "incorporation of the ATPase-defective protein into VPS4 hexamers and subsequent failure of disassembly of the ESCRT complexes."

Residue-level mechanism (Seu et al.) [body/derived]: - Arg284: "one of two arginine fingers functioning in the ATP binding pocket… interaction with the γ-phosphate of ATP, promoting hydrolysis. Therefore, alteration of Arg284 would be expected to compromise ATP hydrolysis and the function of VPS4A as an ATPase." - Gly203: "part of the pore loop 1 motif which forms the central pore of the active hexamer… is highly conserved, as it is the only amino acid that permits the packing and conformation required for the adjacent pore loop 1 residues Trp201 and Leu202 to create binding pockets for the sidechains of the ESCRT-III subunits passing through the pore." - Ala28 (MIT): "putatively disrupts VPS4A interaction with ESCRT-III proteins due to altered structure of the MIT domain, rather than directly affecting its ATPase activity" — a mechanistically distinct, recessive, hypomorphic route, consistent with its milder phenotype and requirement for biallelic dosage. - Arg288 (homozygous, 2026): "disruption of a conserved Arginine-288 residue within the ATPases associated with diverse cellular activities domain" causing loss of protein stability (PMID:42498620).

This yields a clean two-mechanism model worth curating as separate pathophysiology branches: 1. Dominant-negative branch (heterozygous ATPase-domain missense: R284W/G, E206K, G203E) → severe classic CIMDAG. 2. Recessive hypomorphic branch (biallelic A28V MIT-domain, R288Q destabilizing) → milder/atypical presentations, partial VPS4B compensation.

Modifier genes, epigenetics, chromosomal abnormalities

• Modifiers: VPS4B is the only credible candidate (compensatory upregulation observed; paralog-specific non-redundancy demonstrated in PMID:38687820). Not formally tested as a modifier.
• Epigenetics: no data. No methylation episignature has been described for VPS4A; it is not in the published EpiSign panels. This is a concrete, tractable gap worth flagging — episignatures exist for many chromatin and trafficking NDDs and would be diagnostically useful here.
• Chromosomal abnormalities: none causal. 16q22.1 deletions encompassing VPS4A would produce haploinsufficiency, which the population data indicate is tolerated — so CMA-detectable deletions should not be expected to cause CIMDAG. This is a useful negative for differential-diagnosis reasoning.

5. Environmental Information

• Environmental factors: none implicated. CTD lists chemical–VPS4A expression interactions (as for most genes), but none has any established role in this disease.
• Lifestyle factors: not applicable to causation. Relevant only downstream (nutrition/feeding, dental hygiene, mobility/physiotherapy).
• Infectious agents: not causal. There is an interesting inverse biological connection worth noting but not curating as etiology: VPS4A is the ATPase hijacked by many enveloped viruses (HIV-1, Ebola) for ESCRT-dependent budding (GO:0039702 viral budding via host ESCRT complex). Recurrent infections were reported in 1/5 patients, but there is no evidence of a systematic immunodeficiency or of altered viral susceptibility in patients.

6. Mechanism / Pathophysiology

6.1 The core causal chain

[MOLECULAR] De novo heterozygous missense in VPS4A ATPase domain (R284W/G, E206K, G203E)
     │  mutant subunit is stably expressed (protein abundance unchanged)
     ▼
[MOLECULAR] Poisoning of the hexameric VPS4A AAA-ATPase
     │  ≥1 mutant subunit per hexamer → loss of ATP hydrolysis / substrate translocation
     ▼
[MOLECULAR] Failure of ESCRT-III filament disassembly and recycling  (GO:1904903)
     │  IST1 (atypical ESCRT-III) accumulates on limiting membranes; core CHMP2B distribution preserved
     ▼
     ├──► [CELLULAR] Enlarged endo-lysosomal compartments (CD63+/LAMP1+/cathepsin D+)
     ├──► [CELLULAR] Cytokinetic abscission failure at the midbody (GO:0061952)
     ├──► [CELLULAR] Centrosome amplification → multipolar spindles → chromosome missegregation
     ├──► [CELLULAR] Nuclear envelope reformation defect (GO:0007084) → irregular nuclei, γH2AX↑
     ├──► [CELLULAR] Primary ciliogenesis failure ("dot cilium", basal body only)
     ├──► [CELLULAR] Impaired autophagosome closure / LC3B engagement; lipid droplet mishandling
     └──► [CELLULAR] Failed exosome release during reticulocyte maturation (CD71 retention)
     ▼
[TISSUE]  Neural progenitor depletion & aberrant corticogenesis │ Ineffective erythropoiesis │ Lens/retina degeneration
     ▼
[ORGANISM] Microcephaly, cerebellar hypoplasia, dystonia, ID │ Transfusion-dependent anemia, iron overload │ Cataract, retinal dystrophy │ Growth failure

6.2 Molecular pathways

• ESCRT (endosomal sorting complexes required for transport) pathway — the sole primary pathway. VPS4A is the terminal, energy-consuming step that disassembles ESCRT-III polymers so they can be reused. Reactome: Membrane Trafficking → ESCRT-dependent MVB biogenesis. KEGG: hsa04144 (Endocytosis).
• Downstream/secondary: autophagy–lysosome axis (PMID:42498620 — impaired LC3B interaction); cGAS-STING innate immune signaling is engaged when VPS4 is inhibited pharmacologically (PMID:42032367) — mechanistically interesting, unproven in patients; cell-cycle/mitotic checkpoint (Aurora B/ANCHR abscission checkpoint via ZFYVE19/VTA1).
• Not implicated: Wnt, MAPK, mTOR, PI3K-AKT as primary drivers. Do not assert these.

6.3 Suggested GO biological process terms

Molecular function: GO:0016887 ATP hydrolysis activity (DECREASED); GO:0140657 ATP-dependent activity. Cellular component: GO:0005770 late endosome, GO:0005769 early endosome, GO:0005764 lysosome, GO:0030496 midbody, GO:0090543 Flemming body, GO:0005813 centrosome, GO:0000922 spindle pole, GO:0005635 nuclear envelope, GO:0005811 lipid droplet, GO:0000815 ESCRT III complex.

6.4 Cellular processes — the experimental evidence

Endosomal morphology (patient fibroblasts and heterologous overexpression) — Rodger et al. [abstract]:

"In cultured cells, overexpression of VPS4A mutants caused enlarged endosomal vacuoles resembling those induced by expression of known dominant-negative ATPase-defective forms of VPS4A. Proband-derived fibroblasts had enlarged endosomal structures with abnormal accumulation of the ESCRT protein IST1 on the limiting membrane."

Critically, the defect is selective, not a global endosomal collapse [body/derived]: "Heterozygous expression of mutant VPS4A in proband cells does not affect the cellular distribution of a core ESCRT-III complex member, prevent formation of ILVs, or adversely affect the degradation of EGFR" — indeed EGFR degradation was increased at 180 min. This partial-penetrance-at-the-cell-biology-level is what makes the disease survivable and is worth modeling as a distinct node.

Mitosis and genome stability — proband fibroblasts show [body/derived] "an anomalous centrosome number and morphology in interphase"; "Multipolar spindles were observed during mitosis, resulting in a high frequency of aberrant chromosome alignment during metaphase"; "Aberrant chromosome segregation was documented by the presence of both lagging and bridging chromosomes during anaphase and telophase"; increased micronuclei; G2/M accumulation on BrdU flow cytometry; "increased number of γH2AX foci… indicating increased spontaneous DNA damage."

Primary cilium — "Normal cilia were absent in fibroblasts expressing the VPS4A-p.Arg284Gly mutant, which instead showed a visible basal body (dot cilium)"; R284W cells show a dot cilium or occasionally an elongated/normal cilium. Seu et al. additionally report altered "length, number, and Arl13b composition of primary cilia." This makes CIMDAG partially a ciliopathy-adjacent disorder — relevant to the retinal dystrophy, cerebellar hypoplasia, and renal involvement, and a plausible conforms_to link to ciliopathy_dysfunction#Basal Body and Transition Zone Dysfunction in the dismech module set (flag as partial/atypical, since the primary lesion is ESCRT not IFT).

Neurons (iPSC-derived i³Neurons + CRISPRi knockdown) — Rodger et al. [abstract]: "VPS4A function was also required for normal endosomal morphology and IST1 localization in iPSC-derived human neurons." Full text: "a significant increase in the percentage of neurons that had enlarged structures marked by CD63, LAMP1, or the lysosomal enzyme cathepsin D", and — a striking result — overexpression of mutant VPS4A was "incompatible with neuronal survival." CRISPRi VPS4A knockdown "largely recapitulates phenotypes that are observed in proband cells expressing dominant-negative VPS4A at physiological heterozygous levels."

Erythroid lineage — Seu et al. [abstract]:

"Bone marrow studies showed binucleated erythroblasts and erythroblasts with cytoplasmic bridges indicating abnormal cytokinesis and abscission. Circulating red blood cells were found to retain transferrin receptor (CD71) in their membrane, demonstrating that VPS4A is critical for normal reticulocyte maturation. Using proband-derived induced pluripotent stem cells (iPSCs), we have successfully modeled the hematologic aspects of this syndrome in vitro, recapitulating their dyserythropoietic phenotype."

Two mechanistically separable erythroid defects therefore coexist: (i) abscission failure → binucleate erythroblasts → ineffective erythropoiesis (dyserythropoiesis), and (ii) failure of exosome-mediated CD71/transferrin-receptor shedding → arrested reticulocyte maturation → shortened RBC survival (hemolysis). The Lunati case, in which hemolysis dominated without obvious dyserythropoiesis, is the clinical proof that these two arms can dissociate.

Paralog non-redundancy — Dvilansky et al., PLoS Biol 2024 (PMID:38687820) [abstract]:

"Mutations in the human AAA-ATPase VPS4 isoform, VPS4A, cause severe neurodevelopmental defects and congenital dyserythropoietic anemia (CDA). VPS4 is a crucial component of the endosomal sorting complex required for transport (ESCRT) system… while most organisms encode for a single VPS4 gene, human cells have 2 VPS4 paralogs, namely VPS4A and VPS4B, but the functional differences between these paralogs is mostly unknown."

Findings: "VPS4A depletion resulted in a more severe abscission delay than VPS4B and was found to be involved in earlier stages of abscission"; "Depletion of VTA1, a co-factor of VPS4, disrupted VPS4A-ANCHR interactions and accelerated abscission"; STORM imaging showed "the decrease in IST1 density in late intercellular bridges is driven by the VPS4B isoform." Disease relevance stated directly: "Patients who carry normal VPS4B and mutated VPS4A genes develop pathologies due to the disruption of VPS4A-mediated cellular regulation, which is essential for neurodevelopment and cannot be compensated by VPS4B." This is the mechanistic answer to "why doesn't the paralog rescue?" and belongs in the entry.

6.5 Metabolic and lipid changes (newest mechanism, 2026)

Gupta et al., J Clin Lipidol 2026 (PMID:42498620) add a lipotoxicity arm from the homozygous p.Arg288Gln case: loss of protein stability, impaired LC3B interaction, reduced lipid-droplet localization, and cellular "increased lipid accumulation, elevated free fatty acids, and higher reactive oxygen species levels." This provides a candidate unifying explanation for the otherwise puzzling hepatic steatosis (1/5), hepatomegaly (4/5), and lipodystrophy (1/5) in the HPO annotation set. Treat as EMERGING (mechanistic_hypotheses status), single-patient, in-vitro-supported.

6.6 Immune involvement

Not a primary feature. Recurrent infections in 1/5. Pharmacological VPS4 inhibition activates cGAS-STING-TBK1-IRF3 (PMID:42032367, rhabdomyosarcoma models) — mechanistically this predicts that ESCRT dysfunction could produce cytosolic-DNA-driven interferon signaling (consistent with the micronuclei observed in patient fibroblasts). This is an untested hypothesis in patients and is the single most interesting unexplored mechanism: micronuclei + cGAS-STING is the canonical route to a chronic interferonopathy, and no one has looked for an interferon signature in CIMDAG blood. Worth curating as a KNOWLEDGE_GAP with a proposed experiment.

6.7 Tissue damage mechanisms

• CNS: proliferative failure of neural progenitors (abscission defect → the classic microcephaly mechanism shared with other cytokinesis/centrosome microcephaly genes) plus post-mitotic neuronal endolysosomal dysfunction and reduced neuronal survival. Zebrafish data (below) localize the hypotonia to a central sensorimotor transformation defect, not to motor neuron or muscle failure.
• Erythron: ineffective erythropoiesis + peripheral hemolysis + secondary iron loading → hepatosplenomegaly, cholelithiasis, liver iron accumulation.
• Lens/retina: presumed proteostasis/trafficking failure in lens fiber cells and photoreceptor outer-segment turnover (an ESCRT/exosome-dependent process); mechanism not directly demonstrated.
• Oxidative stress: elevated ROS documented in the R288Q model.

6.8 Molecular profiling status

7. Anatomical Structures Affected

Organ level

Primary: - Brain / CNS (UBERON:0000955) — cerebrum (UBERON:0000956 cerebral cortex), cerebellum (UBERON:0002037), cerebellar vermis (UBERON:0004720), corpus callosum (UBERON:0002336), pons (UBERON:0000988), basal ganglia (UBERON:0002420), cerebral white matter (UBERON:0002316) - Bone marrow / erythron (UBERON:0002371) - Eye (UBERON:0000970) — lens (UBERON:0000965), retina (UBERON:0000966), sclera (UBERON:0001773)

Secondary: - Liver (UBERON:0002107) — hepatomegaly, steatosis, iron loading - Spleen (UBERON:0002106) — splenomegaly from extravascular hemolysis - Kidney (UBERON:0002113) — CKD in 1 proband - Gallbladder / biliary (UBERON:0002110) — cholelithiasis - Inner ear / cochlea (UBERON:0001844) — sensorineural hearing loss - Skeletal muscle (UBERON:0001134) — secondary to central tone abnormality, not primary myopathy - Adipose tissue (UBERON:0001013) — lipodystrophy (1/5) - Skeleton (spine UBERON:0001130, foot UBERON:0002387) — scoliosis, talipes - Teeth (UBERON:0001091) — enamel/eruption anomalies

Body systems: nervous (central, dominant), hematopoietic, visual, hepatobiliary, musculoskeletal, growth/endocrine (hypogonadism 1/4), auditory, renal.

Tissue and cell level

Suggested Cell Ontology terms:

Tissue types: nervous tissue (dominant), hematopoietic tissue, transparent avascular lens epithelium/fiber, hepatic parenchyma.

Subcellular level (GO Cellular Component)

GO:0005770 late endosome · GO:0005769 early endosome · GO:0005768 endosome · GO:0005764 lysosome · GO:0032585 multivesicular body membrane · GO:0030496 midbody · GO:0090543 Flemming body · GO:0005813 centrosome · GO:0000922 spindle pole · GO:0005635 nuclear envelope · GO:0005811 lipid droplet · GO:0005929 cilium / GO:0036064 ciliary basal body · GO:0000815 ESCRT III complex · GO:0070062 extracellular exosome.

The subcellular level is where this disease actually "lives" — a dismech entry should anchor most MOLECULAR/CELLULAR pathophysiology nodes here rather than at organ level.

Localization and lateralization

Brain involvement is bilateral and symmetric (cerebellar hypoplasia, symmetric white matter loss, bilateral polymicrogyria in one case, bilateral cataracts). Microcephaly is global. No lateralized or asymmetric pattern reported. Cataracts bilateral.

8. Temporal Development

Onset

• Congenital / prenatal. Primary microcephaly in 3/6 indicates prenatal onset of the brain growth failure; congenital cataracts confirm in-utero lens involvement.
• Neonatal: profound hypotonia, feeding difficulties.
• Early infancy: anemia becomes apparent; developmental delay evident by failure to establish head control.
• Onset pattern: insidious/congenital rather than acute. There is no asymptomatic interval and no "normal then regressed" trajectory reported.

Progression

• Stages: no formal staging system exists. Practically: (1) neonatal hypotonia/feeding/anemia; (2) infancy-early childhood — failure of motor milestones, cataract, seizure onset, transfusion dependence established; (3) later childhood — dystonia/spasticity dominate, iron overload accrues, scoliosis/contractures; (4) adolescence/adult — a minority survive; two deaths in childhood or early adult life.
• Rate: the neurodevelopmental phenotype is best described as severe and largely static with superimposed progressive elements. Dystonia is described as progressive by families; imaging in the Seu cohort was read as "suggestive of a neurodegenerative syndrome." The literature does not resolve this — curate as an explicit open question.
• Course pattern: chronic, lifelong; seizures may be episodic/isolated or recurrent; anemia is chronic with transfusion-cycle fluctuation.
• Duration: lifelong; not self-limited. Feeding difficulties are the one feature Unique describes as often improving: "Feeding issues in the newborn period are common in children with CIMDAG syndrome but usually resolve after babyhood or early childhood."

Patterns

• Remission: none spontaneous. Treatment-induced improvement is limited to specific features (cataract extraction restores optical clarity; transfusion corrects Hb; chelation reverses iron loading).
• Critical periods: (i) prenatal neurogenesis — the microcephaly window is closed before birth, so no postnatal therapy can address it (Unique: "a complete cure is unlikely, even in the future, since the brain has already formed by the time a diagnosis is made"); (ii) first year of life — the actionable window for cataract detection/extraction and for establishing hematologic monitoring, both explicitly recommended; (iii) from ~1 year — the window to begin annual iron-status monitoring before organ iron accrues.

9. Inheritance and Population

Epidemiology

• Prevalence: unknown; ultra-rare. Orphanet has not assigned a prevalence class. Unique (2024): "CIMDAG syndrome is extremely rare. Currently (2024) only 10 individuals with a VPS4A gene variant have been reported in the medical literature but more are known to have been diagnosed." With the 2026 Nepalese case, the published total is ~11. For a structured prevalence record, use prevalence_class: ULTRA_RARE, measure_type: CASES_IN_LITERATURE, count ≈ 11, population: Worldwide, with the Unique guide and the primary case series as evidence. Do not compute a rate_per_100000 — there is no denominator.
• Incidence: not estimable.

Inheritance

• Predominant: autosomal dominant, de novo (HP:0000006; HPOA inheritance annotation for OMIM:619273 is autosomal dominant). Unique: "In all individuals identified so far (2024), except one, the genetic change was a random (known as 'de novo') change."
• Also: autosomal recessive (HP:0000007) in at least two probands — Seu proband 3 (homozygous p.Ala28Val, unaffected heterozygous parents) and the 2026 Nepalese proband (homozygous p.Arg288Gln). Both inheritance modes should be curated, with the AR form tied to MIT-domain/destabilizing alleles and a milder phenotype. This dual-mode architecture (dominant-negative missense vs. recessive hypomorph in the same gene) is a genuinely notable feature of the entry.
• Penetrance: complete for the reported pathogenic alleles in the genotypes described. Heterozygous carriers of the recessive p.Ala28Val allele are unaffected, confirming that this allele is non-dominant. Heterozygous VPS4A LoF alleles in population databases are, by inference, also non-penetrant.
• Expressivity: variable within genotype. Four probands share p.Arg284Trp yet differ in seizures, hearing loss, anemia severity, and hepatic involvement — so intra-allelic variable expressivity is real and unexplained.
• Anticipation: not applicable (no repeat expansion).
• Germline mosaicism: not documented, but explicitly retained in counseling as the residual-risk mechanism — Unique: "One reason why there is some residual chance of recurrence is due to the rare phenomenon called germline mosaicism."
• Founder effects: none identified.
• Consanguinity: directly relevant to the recessive form (Nepalese consanguineous family; Seu proband 3's parents both carriers).
• Carrier frequency: not established for any pathogenic allele; all are absent from gnomAD.

Population demographics

• Affected populations: no ethnic predilection identified. Reported patients span European, North American, and South Asian (Nepalese) ancestries. With n≈11, no inference is possible.
• Geographic distribution: worldwide; reporting is concentrated in the UK/Europe (Rodger, Lunati) and USA (Seu, via CDAR), which is ascertainment bias from where exome sequencing and CDA registries operate — not a true distribution. The one Asian case (Nepal, 2026) suggests underascertainment in low-resource settings.
• Sex ratio: no sex bias reported or expected (autosomal). Individual sexes are not consistently extractable from the abstracts; treat as unknown/1:1.
• Age distribution: all reported patients ascertained in infancy or childhood; two died in childhood or early adult life; the oldest survivors are young adults.

10. Diagnostics

Genetic testing — the definitive route

Recommended approach: trio exome (WES) or genome (WGS) sequencing. Unique: "CIMDAG syndrome is caused by a VPS4A gene sequence variant that can be identified by a type of genetic test called sequencing (e.g. whole exome sequencing (WES) or whole genome sequencing (WGS))… a genetic test is needed to confirm a suspected diagnosis." All 11 reported cases were identified by exome/genome sequencing, several via GeneMatcher-style matchmaking.

Variant interpretation notes for ACMG/AMP application: pathogenic alleles are absent from gnomAD (PM2), de novo with confirmed parentage (PS2), fall in a well-established functional domain and at a mutational hotspot (PM1), have supporting functional studies (PS3 — the endosomal-vacuole and IST1 assays are essentially a validated functional assay for this gene), and R284W recurs in multiple unrelated probands (PS4_moderate). Critically, PVS1 should never be applied to VPS4A — LoF is not the mechanism. ClinVar holds 52 records classified pathogenic/likely pathogenic under a broad query and ~149 VPS4A variants overall; most non-hotspot missense will be VUS.

Clinical tests

Hematology (essential; Unique explicitly recommends): "It is recommended to check on complete blood count (CBC) and reticulocyte count at the time of the diagnosis and also monitor the iron status regularly (e.g. once a year, starting by 1 year of age)."

The CD71-retention flow assay is the closest thing to a functional biomarker for this disease and is worth curating as a biochemical/diagnostic entity — it directly reads out the exosome/ESCRT defect and distinguishes VPS4A-CDA from other CDAs.

Imaging: - Brain MRI (essential): reduced global cerebral volume, thin/hypoplastic corpus callosum, cerebellar and pontine hypoplasia/atrophy, delayed myelination, white matter deficiency, ± polymicrogyria, ± basal ganglia atrophy. Occasional ventriculomegaly ex vacuo. - Abdominal ultrasound: hepatosplenomegaly, gallstones. - Hepatic MRI R2*/FerriScan: iron quantification. - Cardiac T2* MRI: if chronically transfused.

Electrophysiology: EEG / video-telemetry for seizure characterization (Unique lists atypical absence, epileptic spasms, generalized tonic-clonic, myoclonic, tonic semiologies). ERG/VEP for retinal dystrophy vs cortical visual impairment. Auditory brainstem response / newborn hearing screen — note the useful nuance that hearing screens at birth "often give a clear response" despite later SNHL in some, so serial audiology is warranted.

Functional/ophthalmologic: slit-lamp for cataract (recommended from the first year of life and then per specialist), dilated fundoscopy for retinal dystrophy.

Biopsy/histopathology: bone marrow (above) is the only routinely informative biopsy. Skin biopsy for fibroblast culture is valuable for research-grade functional confirmation (enlarged endosomes, IST1 mislocalization, centrosome/cilium assays) — genuinely useful for VUS resolution in this gene given the robust cellular assays published.

Omics-based diagnostics

• RNA-seq: no established role (variants are missense, not splice).
• Proteomics/metabolomics/liquid biopsy: no role.
• Epigenomics: no VPS4A episignature exists — a real opportunity, since an episignature would immediately resolve the VUS burden.

Clinical criteria and differential diagnosis

No consensus diagnostic criteria exist. Diagnosis is molecular, prompted by the gestalt of severe primary microcephaly + profound DD/dystonia + cataract + otherwise-unexplained congenital anemia — a combination that should trigger VPS4A consideration specifically. The anemia is the discriminating clue: severe microcephaly + ID + dystonia has a long differential, but adding transfusion-dependent congenital dyserythropoietic anemia narrows it dramatically.

Differential diagnosis:

Screening

• Newborn screening: not included in any NBS panel; not a candidate (no treatable metabolic marker, no presymptomatic-treatment benefit).
• Carrier screening: not on expanded carrier screening panels; would only be relevant for the recessive alleles in consanguineous populations.
• Cascade screening: for de novo dominant cases, parental testing is for recurrence-risk counseling, not for finding affected relatives. For biallelic families, sibling carrier testing is appropriate.

11. Outcome / Prognosis

Survival and mortality

• Life expectancy: reduced, but not uniformly. Rodger et al.: "Two affected individuals died in childhood or early adult life" — i.e. 2/6 (33%) mortality within the reported follow-up. HPO records HP:0100613 Death in early adulthood at 2/6.
• No 5-/10-year survival estimates, no formal mortality rate, no actuarial life table exists. Do not manufacture one.
• Presumed proximate causes of death (not systematically reported): respiratory infection/aspiration in the context of severe neurodisability, seizure-related events, complications of chronic transfusion and iron overload. Treat as inference, not evidence.

Morbidity and function

• Severe lifelong disability is the rule. Non-ambulatory (no reported patient achieved independent walking), largely non-verbal, fully dependent for activities of daily living, requiring assisted feeding in a substantial fraction.
• Disability outcome (ICF framing): profound impairment across mobility, communication, self-care, and learning domains; complete dependence on caregivers.
• QoL instruments: none applied. No EQ-5D, PedsQL, CPCHILD, or PROMIS data. This is a clean, well-defined evidence gap suitable for a KNOWLEDGE_GAP discussion — and one that matters, because caregiver-reported outcome measures are the plausible endpoint for any future trial in this population.

Complications

Transfusional and disease-related iron overload (disproportionate, due to ineffective erythropoiesis) → hepatic, cardiac, endocrine iron toxicity; cholelithiasis from chronic hemolysis; splenomegaly; recurrent respiratory infection and aspiration; scoliosis and contractures; dental disease; refractory epilepsy in a subset; visual loss from cataract (reversible) and retinal dystrophy (not reversible); chronic kidney disease in at least one patient.

Recovery potential

None for the neurodevelopmental core — the microcephaly and brain malformation are established in utero. Unique is blunt about this: "There is no cure for CIMDAG syndrome since the effects of the genetic change took place during a baby's formation and development… a complete cure is unlikely, even in the future, since the brain has already formed by the time a diagnosis is made." Individual features are correctable (cataract surgery, transfusion, chelation, seizure control).

Prognostic factors

Not formally studied. Reasonable, evidence-anchored predictors to record as hypotheses: 1. Genotype class — the best-supported predictor. Heterozygous ATPase-domain dominant-negative alleles (R284W/G, E206K, G203E) → severe classic phenotype; biallelic MIT-domain/destabilizing hypomorphs (A28V, R288Q) → milder, with preserved head circumference in at least one case (macrocephaly rather than microcephaly). 2. Degree of microcephaly (Z-score) — universal severity marker. 3. Transfusion burden and iron loading — modifiable determinant of long-term organ morbidity. 4. Presence and refractoriness of epilepsy.

Prognostic biomarkers: none validated. Ferritin/LIC are established biomarkers of iron-related risk generically, not of this disease's course.

12. Treatment

There is no disease-modifying or targeted therapy. Management is entirely supportive, multisystem, and multidisciplinary. Unique's management recommendation is the closest thing to a guideline that exists:

"Children with CIMDAG syndrome should be under the care of a multidisciplinary team. The team should include a geneticist and paediatric neurologist who can oversee care so that development and behaviour can be monitored, and the best help given in the form of physiotherapy, occupational therapy, speech therapy and, if needed, behavioural therapy. For signs of anaemia or haemolysis (pallor, jaundice, decreased level of energy, or abnormal CBC/reticulocyte count), evaluation and care by a paediatric haematologist is recommended. Ophthalmology evaluation is also recommended, starting at the first year of life and then as frequently as is recommended by the specialist."

Pharmacotherapy

Pharmacogenomics: nothing VPS4A-specific. Standard PGx applies (e.g. HLA-B*15:02 / HLA-A*31:01 for carbamazepine-family anti-seizure drugs, relevant given oxcarbazepine use; CPIC guideline). Worth recording because this population is disproportionately exposed to aromatic antiepileptics.

Transfusion support

Red cell transfusion is the mainstay for the anemia. Observed intervals in the Seu cohort: every 4–6 weeks (proband 1), every 4–10 weeks (proband 2), every 2–6 months (proband 3). NCIT: NCIT:C15380 Blood Transfusion (verify with OAK) / NCIT:C15747 Supportive Care. Modality: OTHER.

Advanced therapeutics

Surgical and interventional

• Cataract extraction — high-value, low-burden. Unique quotes a family: "Check the eyes on a regular basis – if a cataract is diagnosed it can be easily removed with a small surgery." NCIT: NCIT:C15329 Surgical Procedure (or a specific lens-extraction term if OAK resolves one). Modality: SURGERY.
• Gastrostomy (PEG/G-tube) or nasogastric tube — for feeding failure/aspiration in the subset with severe difficulty. NCIT:C15329. Modality: SURGERY/DEVICE.
• Orthopedic surgery — talipes correction, scoliosis management. NCIT:C16186 Orthopedic Surgical Procedure. Modality: SURGERY.
• Splenectomy — not reported and not recommended in this disease; note explicitly, since it is a standard consideration in other congenital hemolytic anemias and its role here is unknown.
• Dental treatment under general anesthesia — frequently required.

Supportive and rehabilitative

• Physiotherapy (NCIT:C15302 Physical Therapy, BEHAVIORAL) — including hydrotherapy and hippotherapy, both named by families.
• Occupational therapy (NCIT:C121351, BEHAVIORAL).
• Speech and language therapy + AAC (NCIT:C159273, BEHAVIORAL) — Unique emphasizes AAC (pictograms, gestures, simplified sign, high-tech aided communication) given that many remain non-verbal.
• Orthotics/bracing — insoles, braces, splints, callipers, spinal bracing; wheelchair and mobility aids (DEVICE).
• Nutritional support — high-energy formula, positioning for feeds (NCIT:C15433; note the CLAUDE.md caution — do not auto-tag nutritional support as BEHAVIORAL; here the modality is genuinely dietary/behavioral for positioning and formula, so judge per item).
• Behavioral therapy and sleep hygiene (NCIT:C181743, BEHAVIORAL).
• Genetic counseling (NCIT:C15240).

Experimental treatments / clinical trials

No interventional trial for CIMDAG or VPS4A exists — a ClinicalTrials.gov API query for "VPS4A OR CIMDAG" returns zero studies (checked 2026-08-01).

The one registered study capturing these patients is observational: - NCT02964494 — The Congenital Dyserythropoietic Anemia Registry (CDAR); sponsor Cincinnati Children's Hospital Medical Center; observational patient registry; recruiting; estimated enrollment 10,000; condition: Congenital Dyserythropoietic Anemia. This registry is how the three Seu et al. probands were ascertained, and it maintains a blood/bone-marrow biorepository. It is the correct clinical_trials entry for a dismech CIMDAG page, curated with the caveat that it is CDA-wide rather than VPS4A-specific.

Treatment outcomes

No response-rate data. Transfusion reliably corrects hemoglobin; chelation reliably lowers ferritin/LIC (generic evidence, not CIMDAG-specific). Cataract surgery outcomes not reported in this cohort. Anti-seizure efficacy anecdotal — Unique: "there are currently (2024) no specific recommendations for treating CIMDAG syndrome seizures." Adverse events are those of the generic therapies (chelator nephro/hepatotoxicity and cytopenias; transfusion reactions and alloimmunization; antiepileptic AEs).

Treatment strategy

No algorithm exists. A defensible, evidence-anchored care schedule assembled from the Unique recommendations:

1. At diagnosis: CBC + reticulocyte count; brain MRI; ophthalmology (slit lamp + fundoscopy); audiology; EEG if paroxysmal events; feeding/swallow assessment; genetic counseling.
2. Annually from age 1: iron status (ferritin ± LIC by MRI); ophthalmology per specialist; growth and nutrition; audiology; spine and hip surveillance.
3. Ongoing: PT/OT/SLT; dental review; sleep review; hematology follow-up with transfusion and chelation as indicated.
4. As needed: cataract extraction; gastrostomy; orthopedic intervention; anti-seizure escalation.

Personalized medicine: genotype currently informs prognosis (dominant-negative vs. biallelic hypomorph) but not therapy. Genotype-guided therapy would become real only with allele-selective silencing.

13. Prevention

• Primary prevention: not possible for de novo dominant cases — by definition unpredictable and unpreventable. Unique makes the non-blame point explicitly: "It is important to recognize that no one should be blamed for variants in their DNA and no parent is at fault when a new DNA change occurs in their child." For the recessive form, reduction of consanguinity-associated risk at the population level is the only structural lever, and is a public-health rather than clinical intervention.
• Secondary prevention (early detection): the actionable tier. Early molecular diagnosis via trio WES/WGS enables the surveillance schedule above. Unique: "knowing the diagnosis means that appropriate monitoring and interventions can be put in place early to help each child reach their full potential." Specific high-yield items: first-year ophthalmology to catch surgically remediable cataract; CBC/retic at diagnosis to catch anemia; annual iron monitoring from age 1 to pre-empt organ iron loading.
• Tertiary prevention: iron chelation to prevent cardiac/hepatic/endocrine iron toxicity; aspiration precautions and gastrostomy to prevent recurrent pneumonia; postural management, orthotics, and spinal surveillance to prevent contracture and scoliosis progression; dental care to prevent decay and erosion from bruxism; seizure control.
• Immunization: no disease-specific vaccine. Standard schedule plus enhanced respiratory protection (influenza, pneumococcal, RSV per local policy) is appropriate for a child with severe neurodisability and aspiration risk. Transfusion-dependent patients should be hepatitis B immune.
• Genetic screening / reproductive options: prenatal diagnosis and preimplantation genetic testing (PGT-M) are available once the familial variant is known. For de novo cases, recurrence risk is "usually less than 1%" with germline mosaicism as the residual mechanism; for biallelic families, 25% per pregnancy with both parents heterozygous carriers.
• Genetic counseling: essential and should be offered at diagnosis and again at reproductive planning for parents and, later, for unaffected siblings (carrier status relevant only in the recessive families). NCIT: NCIT:C15240.
• Risk stratification / public health / environmental interventions: not applicable.
• Prophylaxis: no disease-specific prophylactic medication.

14. Other Species / Natural Disease

Taxonomy and orthologs

Evolutionary conservation is the key comparative fact and it has direct disease relevance: most organisms encode one VPS4; only vertebrates/mammals have the VPS4A/VPS4B paralog pair. Dvilansky et al. showed the paralogs are functionally non-redundant — VPS4A acts earlier in abscission and cannot be substituted by VPS4B — which is precisely why a heterozygous VPS4A missense allele causes human disease despite an intact VPS4B. Any invertebrate or yeast model necessarily collapses this paralog distinction and therefore cannot model the human genetics faithfully, only the core ESCRT biochemistry. The affected residues (Ala28, Gly203, Glu206, Arg284, Arg288) are deeply conserved across all these orthologs.

Breed

Not applicable (VBO: no entries).

Natural disease in other species

None known. An OMIA search for VPS4A returns no entries in any species — no naturally occurring VPS4A disease has been described in companion animals, livestock, or wildlife. Record this as an explicit negative rather than omitting the section.

Comparative pathology

Only engineered models exist (§15). The cross-species comparison that has been made — zebrafish vps4a T248I vs. human patients — is informative: the fish reproduce the central origin of hypotonia ("Resembling the central form of hypotonia in VPS4A patients, motor neurons and muscle cells are functional in mutant zebrafish") and the endosomal/exosome defect, but the fish inner ear is unaffected, unlike the SNHL seen in 2/5 human patients.

Transmission

Not applicable — non-infectious, non-zoonotic, no cross-species susceptibility.

15. Model Organisms

15.1 Zebrafish — the best available whole-organism neurological model

Shipman et al., J Neurosci 2024;44(50):e0680242024. PMID:39455257; PMCID:PMC11638813; DOI:10.1523/JNEUROSCI.0680-24.2024

Abstract [abstract]:

"Mutations in human VPS4A are associated with neurodevelopmental defects, including motor delays and defective muscle tone. VPS4A encodes a AAA-ATPase required for membrane scission, but how mutations in VPS4A lead to impaired control of motor function is not known. Here we identified a mutation in zebrafish vps4a, T248I, that affects sensorimotor transformation."

• Model type: forward-genetic missense allele, vps4a^T248I^, in the AAA-ATPase domain (mechanistically parallel to the human ATPase-domain alleles).
• Molecular validation: the mutation "reduces the ATPase activity of Vps4a and disassembly of ESCRT filaments, which mediate membrane scission" — the same primary lesion as in patients.
• Phenotypes recapitulated: "Optomotor responses, vestibulospinal, and acoustic startle reflexes are absent or strongly impaired"; larvae "fail to maintain an upright posture" and "swim sideways or upside down"; "decreased numbers of circulating exosomes in brain ventricles"; "enlarged endosomal compartments in the CNS"; upregulation of atf3 and jun stress-response genes.
• Key mechanistic dissociation: "Resembling the central form of hypotonia in VPS4A patients, motor neurons and muscle cells are functional in mutant zebrafish" — establishing that the hypotonia/motor phenotype is central, arising at the level of sensorimotor transformation, not neuromuscular.
• Limitations: "the function of the inner ear in vps4a^T248I^ larvae is unaffected" and vestibulo-ocular responses to head rotation are normal — so the model does not capture human SNHL. It also does not model microcephaly, cataract, or the anemia, and zebrafish lack the VPS4A/VPS4B paralog split in the mammalian configuration.
• Applications: sensorimotor circuit function, exosome biology in vivo, CNS endosomal morphology, small-molecule screening for motor-function rescue.
• Resource: ZFIN.

15.2 Mouse

MGI:1890520 (Vps4a). The MGI record lists 19 mutations/alleles across three classes — 6 endonuclease-mediated, 9 gene-trapped, 4 targeted — with 19 strains/lines available through IMSR. Existing phenotype annotations (8 phenotypes from one allele in one background, plus one multigenic phenotype) span craniofacial, digestive, embryonic development, growth/size, hematopoietic, integument, immune, and mortality/aging systems. MGI records no human disease model association for Vps4a — i.e., no mouse has yet been designated a CIMDAG model.

What exists: - Cardiomyocyte-specific Vps4a conditional knockout (PMC10341959): impaired autophagic flux — Vps4a is "mainly involved in the sealing of autophagosome in cardiomyocytes" — with cardiac hypertrophy (increased HW/BW and HW/TL), partial LV dilation, reduced ejection fraction, and partial lethality beginning at 3 months, some surviving beyond 12 months. Confirms an in vivo autophagy role for Vps4a but models no CIMDAG feature (patients have no reported cardiomyopathy). - Double Vps4a/Vps4b heterozygous null MEFs: increased endo/lysosomal organelles with large multi-membrane structures.

The critical gap: no knock-in mouse carrying a patient allele (e.g. Vps4a^R284W/+^) has been reported. This is the single most valuable missing model. Because the mechanism is dominant-negative and the paralog architecture is mammal-specific, a heterozygous knock-in mouse — not a knockout — is the design that would test whether the human genetics reproduce in vivo, and would be the required platform for any allele-selective silencing preclinical program. Note also that a Vps4a knockout mouse would likely be uninformative for CIMDAG, since human heterozygous LoF is tolerated.

Register this as a HUMAN_MODEL_MISMATCH discussion rather than a plain KNOWLEDGE_GAP: model-system evidence exists (zebrafish missense, mouse conditional KO), but its fidelity to the human dominant-negative, paralog-dependent disease is the open question.

15.3 Cellular and in vitro models — the workhorses of this field

Phenotype recapitulation summary: the cellular models are unusually strong — patient fibroblasts and patient iPSC-derived erythroblasts reproduce the disease-defining cellular lesions faithfully, and iPSC-neurons reproduce the endolysosomal lesion. This makes the assay set genuinely usable for VUS resolution (PS3-level functional evidence). What no model yet captures: microcephaly in vivo, cataract, growth failure, and the integrated multisystem phenotype.

15.4 Databases

MGI (informatics.jax.org) · IMSR (19 Vps4a lines) · IMPC · ZFIN · Alliance of Genome Resources · FlyBase · WormBase · SGD · DepMap (VPS4A/VPS4B synthetic lethality) · Cellosaurus (for patient iPSC lines, if deposited).

Appendix A — Consolidated ontology term suggestions for KB entry

Disease: MONDO:0035819 Gene: hgnc:13488 (VPS4A) — note dismech lowercase hgnc: convention Inheritance: HP:0000006 autosomal dominant (primary); HP:0010985? no — use HP:0000007 autosomal recessive for the biallelic branch

Top HPO terms (with frequency band): HP:0011344 severe global developmental delay (OBLIGATE/VERY_FREQUENT, 6/6) · HP:0010864 severe intellectual disability (6/6) · HP:0000252 microcephaly (6/6) · HP:0000505 visual impairment (6/6) · HP:0001332 dystonia (5/6) · HP:0001252 hypotonia (5/6) · HP:0001270 motor delay (5/5) · HP:0000750 delayed speech and language development (5/5) · HP:0000519 developmental cataract (4/5) · HP:0001257 spasticity (4/5) · HP:0002360 sleep disturbance (4/4) · HP:0011968 feeding difficulties (4/6) · HP:0002240 hepatomegaly (4/5) · HP:0001250 seizure (3/6) · HP:0000556 retinal dystrophy (3/5) · HP:0011451 primary microcephaly (3/6) · HP:0001321 cerebellar hypoplasia · HP:0033725 thin corpus callosum · HP:0000407 sensorineural hearing impairment (2/5) · HP:0100613 death in early adulthood (2/6) · HP:0031688 erythroid dysplasia · HP:0001878 hemolytic anemia · HP:0001744 splenomegaly · HP:0001510 growth delay · HP:0001414 microvesicular hepatic steatosis · HP:0009125 lipodystrophy · HP:0001081 cholelithiasis

GO BP: GO:1904903, GO:1904896, GO:0071985, GO:0032509, GO:0061952, GO:0007084, GO:0060271, GO:0007059, GO:0006914 GO CC: GO:0000815, GO:0005770, GO:0005769, GO:0005764, GO:0030496, GO:0090543, GO:0005813, GO:0000922, GO:0005635, GO:0005811, GO:0070062 GO MF: GO:0016887 ATP hydrolysis activity CL: CL:0000047, CL:0000540, CL:0000765, CL:0000558, CL:0000232, CL:0000057, CL:0000121 UBERON: UBERON:0000955, UBERON:0002037, UBERON:0002336, UBERON:0000988, UBERON:0002371, UBERON:0000965, UBERON:0000966, UBERON:0002107, UBERON:0002106 NCIT (treatment): NCIT:C15986, NCIT:C15747, NCIT:C15302, NCIT:C121351, NCIT:C159273, NCIT:C15329, NCIT:C16186, NCIT:C15240, NCIT:C15433

Candidate dismech module conformance: - ciliopathy_dysfunction#Basal Body and Transition Zone Dysfunction — partial/atypical (cilium defect is secondary to ESCRT failure, not an IFT/BBSome lesion). Curate with an explicit caveat. - hemolytic_anemia_erythrocyte_destruction#Premature Erythrocyte Destruction — fits the Lunati/hemolytic arm. - No existing module covers ESCRT-III recycling failure or cytokinetic abscission failure. A new escrt_membrane_scission_failure or cytokinetic_abscission_failure module is a well-motivated candidate — abscission failure is a conserved mechanism shared with KIF23-CDA III, RACGAP1-CDA, and the centrosomal MCPH microcephaly genes, which is exactly the recurrence profile that justifies a module.

Appendix B — Primary reference list

Appendix C — Explicit gaps and open questions (candidates for discussions blocks)

1. KNOWLEDGE_GAP — static vs progressive. Is CIMDAG a static congenital malformation syndrome or a progressive neurodegeneration? Imaging language in Seu et al. is explicitly neurodegenerative; no longitudinal imaging series exists. Proposed experiment: serial MRI in the ~11 known patients + the diagnosed-but-unpublished cohort known to Unique.
2. KNOWLEDGE_GAP — micronuclei → cGAS-STING interferonopathy. Patient fibroblasts make micronuclei; VPS4 inhibition activates cGAS-STING in tumor models. No one has measured an interferon signature in patient blood. Proposed experiment: IFN-stimulated gene score on patient PBMCs.
3. HUMAN_MODEL_MISMATCH — no knock-in mouse. All mouse data are knockout/conditional-knockout; the human mechanism is heterozygous dominant-negative in a paralog-redundant-but-non-equivalent system. A Vps4a^R284W/+^ knock-in is the required model and does not exist.
4. KNOWLEDGE_GAP — HPOA frequency discrepancy. HPO annotation records cerebellar hypoplasia at 2/6 while Rodger full text states 5/6; anemia is absent from HPOA entirely despite being present in ~7/10 published patients. The HPOA record for OMIM:619273 under-annotates the hematologic domain.
5. KNOWLEDGE_GAP — no episignature, no QoL instrument, no natural history study, no prevalence estimate.
6. Therapeutic hypothesis worth recording: allele-selective knockdown of the mutant transcript is the mechanistically correct strategy, uniquely supported by the fact that VPS4A haploinsufficiency is tolerated in the general population. The recurrent c.850A>T hotspot (5/11 patients) is an ideal ASO/siRNA/base-editing target. No program exists.
7. Unresolved genotype–phenotype: four unrelated p.Arg284Trp probands differ markedly in seizures, hearing, anemia, and hepatic involvement — the source of intra-allelic variable expressivity is unknown.

Sources: Rodger et al. 2020, AJHG (PMC7820634) · Seu et al. 2020, AJHG (PMC7820805) · Lunati et al. 2021, Am J Hematol · Gupta et al. 2026, J Clin Lipidol (PMID:42498620) · Shipman et al. 2024, J Neurosci (PMC11638813) · Dvilansky et al. 2024, PLoS Biol (PMC11086821) · Unique: CIMDAG syndrome guide (2024) · OMIM #619273 · OMIM *609982 · HPO annotations for OMIM:619273 · MONDO:0035819 via OLS4 · UniProt Q9UN37 · HGNC:13488 · MGI:1890520 (Vps4a) · Vps4a cardiomyocyte KO (PMC10341959) · CDAR registry NCT02964494 · Genomics England PanelApp — VPS4A (Rare anaemia) · ClinVar VPS4A