CYFIP2-Related Developmental and Epileptic Encephalopathy

Mendelian MONDO:0033374 Pathograph 11 Show in embeddings browser Epilepsy Neurological Disease

CYFIP2-related developmental and epileptic encephalopathy (DEE65) is an autosomal dominant disorder caused by heterozygous, almost always de novo, missense variants in CYFIP2. CYFIP2 is a structural subunit of the WAVE regulatory complex, which holds the actin-nucleating VCA domain of WAVE in an autoinhibited state until RAC1 signalling releases it to drive Arp2/3-mediated actin polymerization. Pathogenic variants cluster at two hotspots - p.Arg87, which sits buried at the CYFIP2-WAVE1 interface and produces a consistently severe phenotype, and p.Asp724, which produces a more variable one. Affected children have early-onset drug-resistant seizures, frequently epileptic spasms in a West syndrome pattern, with severe developmental impairment and hypotonia. Brain MRI is often unremarkable despite the severity of the clinical and electrical picture.

Ask OpenScientist

Ask a research question about CYFIP2-Related Developmental and Epileptic Encephalopathy. OpenScientist will conduct autonomous deep research using the Disorder Mechanisms Knowledge Base and PubMed literature (typically 10-30 minutes).

Submitting...

Do not include personal health information in your question. Questions and results are cached in your browser's local storage.

1
Inheritance
9
Pathophys.
8
Phenotypes
2
Hypotheses
3
Gaps
11
Pathograph
1
Genes
1
Medical Actions
2
Models
1
Deep Research
👪

Inheritance

1
Autosomal dominant HP:0000006
Heterozygous and overwhelmingly de novo. No confirmed transmission from an affected parent has been reported, consistent with the severity of the phenotype.
Autosomal dominant inheritance
Show evidence (1 reference)
PMID:30664714 SUPPORT Human Clinical
"we identified 12 independent patients carrying a total of eight distinct de novo variants in CYFIP2 with a shared phenotype of intellectual disability, seizures, and muscular hypotonia"
Establishes the de novo heterozygous architecture across twelve independent patients, together with the core phenotype triad.
◈

Mechanistic Hypotheses

2
wrc_gain_of_function_model
wrc_gain_of_function_model CANONICAL
Evidence balance 1 support
Hotspot variants destabilize the CYFIP2-WAVE1 interface, weakening sequestration of the VCA domain and releasing the complex to nucleate actin outside normal RAC1 control. On this account the disorder is a toxic gain of function at the level of actin regulation, which distinguishes it from the many developmental and epileptic encephalopathies that act by haploinsufficiency.
Pathograph links
Weakened CYFIP2-WAVE1 Interface Binding -> Aberrant WAVE Regulatory Complex Activation DIRECT Loss of VCA sequestration is what releases the complex.
Aberrant WAVE Regulatory Complex Activation -> Dysregulated Actin Cytoskeleton Dynamics DIRECT An activated complex nucleates actin.
Show evidence (1 reference)
PMID:29534297 SUPPORT In Vitro
"Our findings suggest that de novo Arg87 variants in CYFIP2 have gain-of-function effects on the WAVE signaling pathway"
The originating statement of the gain-of-function model.
protein_destabilization_model
protein_destabilization_model ALTERNATIVE
Evidence balance 1 support
p.Arg87Cys enhances ubiquitination and proteasomal degradation of CYFIP2, lowering steady-state protein levels in the brain. On this account the primary consequence is less CYFIP2, not a constitutively active complex. The two accounts are not trivially compatible: one says the complex is inappropriately switched on, the other that there is less of a required subunit.
Pathograph links
De Novo CYFIP2 Missense Variant at a Structural Hotspot -> Enhanced CYFIP2 Ubiquitination and Proteasomal Degradation DIRECT A second, mechanistically separate consequence demonstrated for p.Arg87Cys. Drawn as a parallel edge from the variant rather than downstream of the interface defect, because the two are competing accounts of the same variant rather than sequential steps - see the controversy discussion.
Enhanced CYFIP2 Ubiquitination and Proteasomal Degradation -> Dysregulated Actin Cytoskeleton Dynamics INDIRECT UNKNOWN INTERMEDIATES Reduced CYFIP2 abundance is proposed to reach the same actin readout by a different route than interface destabilization.
Show evidence (1 reference)
PMID:36251395 SUPPORT Model Organism
"Beyond identifying a decrease in CYFIP2 protein levels in the Cyfip2+/R87C brains, we demonstrated that the p.Arg87Cys variant enhances ubiquitination and proteasomal degradation of CYFIP2"
The originating statement of the destabilization model, measured in brain tissue of the genotype-matched mouse.
?

Discussions and Knowledge Gaps

3
Does p.Arg87Cys act by switching the WAVE regulatory complex on, or by removing CYFIP2 protein from it?
CONTROVERSY OPEN controversy_gain_of_function_versus_protein_destabilization
Attached to
pathophysiology#Aberrant WAVE Regulatory Complex Activation pathophysiology#Enhanced CYFIP2 Ubiquitination and Proteasomal Degradation
These are not two descriptions of one finding. The gain-of-function account says the complex is inappropriately active because the mutant subunit sequesters VCA more weakly; the destabilization account says there is measurably less CYFIP2 in the brain because the variant protein is ubiquitinated and degraded. A complex missing a required subunit and a complex that is constitutively switched on are different molecular states with opposite implications for therapy - one argues for inhibiting WAVE signalling, the other for restoring CYFIP2 abundance. Both rest on direct measurement: weaker VCA binding and increased actin accumulation in one case, reduced brain protein levels and demonstrated ubiquitination in the other, the latter in a genotype-matched knock-in rather than an overexpression system. A published exchange on exactly this point exists in the literature. The entry therefore draws both edges from the variant node in parallel, tagged to competing hypothesis groups, rather than placing either upstream of the other. Note also that the two accounts are not strictly exclusive - a destabilized interface could plausibly both loosen VCA binding and expose the protein to degradation - but no source demonstrates that reconciliation, and asserting it here would invent a synthesis rather than curate one.
Proposed experiments
Separate CYFIP2 abundance from complex activity
exp_separate_abundance_from_activity
Restore CYFIP2 abundance in Cyfip2+/R87C neurons by proteasome-independent means while leaving the variant sequence intact, and ask whether actin dysregulation and network hyperexcitability normalize. If they do, reduced abundance is the operative mechanism; if actin dysregulation persists at normal protein levels, the interface gain-of-function account is doing the work.
Does the seizure-free interval in the knock-in mouse correspond to anything in the human disease, or is it a species artefact?
HUMAN MODEL MISMATCH OPEN mismatch_mouse_has_a_latent_phase_the_human_disease_does_not
Attached to
pathophysiology#Neuronal Network Hyperexcitability
The Cyfip2+/R87C mouse passes through three phases: neonatal spasms, then a seizure-free period, then spontaneous recurrent seizures in adulthood ending in premature death. Human DEE65 is described as having intractable seizures from onset, with no reported latent phase. Two readings are possible and they have opposite consequences. If the interval is a species artefact, then the model's adult seizures are not the same illness as the human one and the intervening period should not be used to time interventions. If instead the human trajectory has an under-recognized quiescent phase - plausible, since these children are heavily medicated and profoundly impaired, which makes a partial seizure remission easy to miss - then the model is revealing something about the human course rather than diverging from it. This is a tractable question because it can be settled from existing patient records rather than new experiments.
Proposed experiments
Longitudinal seizure burden in genotyped patients
exp_longitudinal_seizure_burden_in_patients
Assemble longitudinal seizure-frequency and EEG data across the lifespan for genotyped DEE65 individuals, specifically looking for a period of reduced seizure burden after the infantile spasm phase, and stratify by hotspot. A demonstrated quiescent interval would reclassify the mouse's latent phase from artefact to finding.
By what route does dysregulated actin polymerization produce a hyperexcitable network?
KNOWLEDGE GAP OPEN gap_how_actin_dysregulation_becomes_hyperexcitability
Attached to
pathophysiology#Disrupted Synaptic and Dendritic Spine Remodeling pathophysiology#Neuronal Network Hyperexcitability
The molecular and cellular end of this disorder is well characterised, and the clinical end is unambiguous, but the step between them is the weakest link in the chain. Actin dysregulation is measured in fibroblasts and transfected cells; seizures are measured in patients and mice; what is not established is the mechanism that converts abnormal spine actin dynamics into synchronous network discharge. The entry marks both edges across this gap as indirect with unknown intermediates rather than asserting a pathway. This matters practically, because it is the step at which any therapy aimed at the actin machinery would have to act on the seizures.
Proposed experiments
Spine dynamics and excitability in the same preparation
exp_spine_dynamics_to_excitability_in_same_preparation
In Cyfip2+/R87C cortical or hippocampal slices, measure dendritic spine density and actin turnover alongside excitatory and inhibitory synaptic currents and network activity in the same neurons, at each of the three phases of the model's course. Establishing whether the excitatory-inhibitory balance shifts in step with the spine changes would convert the current inference into a measured link.
⚙

Pathophysiology

9
De Novo CYFIP2 Missense Variant at a Structural Hotspot
The initiating lesion. Variants cluster at two hotspots, p.Arg87 and p.Asp724. Arg87 is buried at the interface between CYFIP2 and WAVE1 inside the assembled complex, which is why substituting it destabilises the interface rather than simply removing a surface contact. Variants scattered elsewhere in the primary sequence converge on the same interface in three dimensions.
None
Downstream
Weakened CYFIP2-WAVE1 Interface Binding The structural lesion acts first on the interface it sits in.
Enhanced CYFIP2 Ubiquitination and Proteasomal Degradation A second, mechanistically separate consequence demonstrated for p.Arg87Cys. Drawn as a parallel edge from the variant rather than downstream of the interface defect, because the two are competing accounts of the same variant rather than sequential steps - see the controversy discussion. None
Show evidence (2 references)
PMID:29534297 SUPPORT Human Clinical
"Structural analysis indicated that the Arg87 residue is buried at an interface between CYFIP2 and WAVE1, and the Arg87 variant may disrupt hydrogen bonding, leading to structural instability and aberrant activation of the WAVE regulatory complex"
Locates the hotspot residue structurally and states the predicted consequence for the complex.
PMID:33149277 SUPPORT Human Clinical
"Asp724 as second mutational hotspot (4/19 cases)"
Establishes the second hotspot and its frequency in the expanded cohort.
Weakened CYFIP2-WAVE1 Interface Binding
Mutant CYFIP2 binds the VCA domain more weakly than wild type. Because the complex works by sequestering VCA, weaker binding means less effective autoinhibition. This node is the measured binding change, kept separate from the downstream actin readout.
None
actin binding GO:0003779 Gene Ontology (GO) Relation: this pathophysiological event involves this molecular function This pathophysiological event involves actin binding (GO:0003779). GO:0003779 is a molecular function from the Gene Ontology.
WAVE regulatory complex GO:0031209 Gene Ontology (GO) Relation: this pathophysiological event involves this cellular component This pathophysiological event involves WAVE regulatory complex, annotated with SCAR complex (GO:0031209). GO:0031209 is a cellular component from the Gene Ontology.
Downstream
Aberrant WAVE Regulatory Complex Activation Loss of VCA sequestration is what releases the complex. None
Show evidence (1 reference)
PMID:29534297 SUPPORT In Vitro
"All mutant CYFIP2 showed comparatively weaker interactions to the VCA domain than wild-type CYFIP2"
The direct binding measurement, reported for every mutant tested.
Enhanced CYFIP2 Ubiquitination and Proteasomal Degradation
A second consequence of p.Arg87Cys: the variant protein is ubiquitinated and degraded more readily, lowering steady-state CYFIP2 abundance in the brain. This is the observation that sits in tension with a pure gain-of-function reading, and it is curated as its own node rather than folded into the interface account.
None
proteasomal protein catabolic process GO:0010498 Gene Ontology (GO) Relation: this pathophysiological event involves this biological process This pathophysiological event involves increased proteasomal protein catabolic process (GO:0010498). GO:0010498 is a biological process from the Gene Ontology. ↑ INCREASED
Animal models
Cyfip2 p.Arg87Cys knock-in mouse Mouse recapitulates The model in which the degradation mechanism was demonstrated, in brain tissue.
Downstream
Dysregulated Actin Cytoskeleton Dynamics Reduced CYFIP2 abundance is proposed to reach the same actin readout by a different route than interface destabilization. None
Show evidence (2 references)
PMID:36251395 SUPPORT Model Organism
"we demonstrated that the p.Arg87Cys variant enhances ubiquitination and proteasomal degradation of CYFIP2"
Direct demonstration of the degradation mechanism in the knock-in mouse and cellular models.
PMID:36251395 SUPPORT Model Organism
"Beyond identifying a decrease in CYFIP2 protein levels in the Cyfip2+/R87C brains"
Records the reduced steady-state protein abundance that the degradation mechanism produces, in brain tissue.
Aberrant WAVE Regulatory Complex Activation
Release of the WAVE VCA domain from autoinhibition, activating the complex outside normal RAC1 control. This is the node the gain-of-function account turns on.
None
Arp2/3 complex-mediated actin nucleation GO:0034314 Gene Ontology (GO) Relation: this pathophysiological event involves this biological process This pathophysiological event involves increased Arp2/3 complex-mediated actin nucleation (GO:0034314). GO:0034314 is a biological process from the Gene Ontology. ↑ INCREASED
WAVE regulatory complex GO:0031209 Gene Ontology (GO) Relation: this pathophysiological event involves this cellular component This pathophysiological event involves WAVE regulatory complex, annotated with SCAR complex (GO:0031209). GO:0031209 is a cellular component from the Gene Ontology.
Downstream
Dysregulated Actin Cytoskeleton Dynamics An activated complex nucleates actin. None
Show evidence (1 reference)
PMID:29534297 SUPPORT In Vitro
"Our findings suggest that de novo Arg87 variants in CYFIP2 have gain-of-function effects on the WAVE signaling pathway and are associated with severe neurological disorders"
States the gain-of-function interpretation of the WAVE pathway effect.
Dysregulated Actin Cytoskeleton Dynamics
The cellular convergence point of both proposed molecular routes. Abnormal actin accumulation was shown first in transfected cells and subsequently substantiated in patient-derived fibroblasts, which matters because it moves the finding out of an overexpression system into primary patient cells.
actin cytoskeleton organization GO:0030036 Gene Ontology (GO) Relation: this pathophysiological event involves this biological process This pathophysiological event involves dysregulated actin cytoskeleton organization (GO:0030036). GO:0030036 is a biological process from the Gene Ontology. ↕ DYSREGULATED
Downstream
Disrupted Synaptic and Dendritic Spine Remodeling Actin dynamics are the substrate for spine and synapse remodeling in neurons.
Show evidence (2 references)
PMID:29534297 SUPPORT In Vitro
"Immunofluorescence revealed that ectopic speckled accumulation of actin and CYFIP2 was significantly increased in cells transfected with mutant CYFIP2"
The original actin readout in a heterologous transfection system.
PMID:33149277 SUPPORT In Vitro
"Consistent with its role in WRC-mediated actin polymerization we substantiate aberrant regulation of the actin cytoskeleton in patient fibroblasts"
Confirms the same cellular abnormality in primary cells from affected individuals rather than in an overexpression model.
Disrupted Synaptic and Dendritic Spine Remodeling
In neurons the WAVE complex drives actin remodeling at dendritic spines and growth cones, so dysregulated actin dynamics translate into abnormal synaptic architecture. The knock-in mouse shows age-progressive cytoarchitectural disorganization and gliosis in the hippocampus, and longitudinal work shows the synaptic remodeling is itself time-dependent rather than a fixed lesion.
glutamatergic neuron CL:0000679 Cell Ontology (CL) Relation: this pathophysiological event involves this cell type This pathophysiological event involves glutamatergic neuron (CL:0000679). CL:0000679 is a cell type from the Cell Ontology.
dendritic spine GO:0043197 Gene Ontology (GO) Relation: this pathophysiological event involves this cellular component This pathophysiological event involves dendritic spine (GO:0043197). GO:0043197 is a cellular component from the Gene Ontology.
Downstream
Reactive Gliosis Glial activation accompanies the evolving synaptic pathology in the model.
Neuronal Network Hyperexcitability Abnormal synaptic architecture is the proposed substrate for network hyperexcitability.
Show evidence (2 references)
PMID:36251395 SUPPORT Model Organism
"Age-progressive cytoarchitectural disorganization and gliosis were also identified in the hippocampus of Cyfip2+/R87C mice"
Structural correlate in the hippocampus of the genotype-matched mouse, and records that it worsens with age.
PMID:41066510 SUPPORT Model Organism
"This progression was associated with synaptic remodeling, sequential activation of different glial cell types, lipid droplet accumulation in astrocytes, and significant proteomic and lipidomic changes in the brain"
Establishes synaptic remodeling and staged glial involvement as accompaniments of seizure evolution rather than a static finding.
Reactive Gliosis
Sequential activation of different glial populations over the disease course in the mouse model, accompanied by lipid droplet accumulation in astrocytes. Curated as its own node because it is a distinct tissue-level process with its own time course, not a restatement of the synaptic finding.
astrocyte CL:0000127 Cell Ontology (CL) Relation: this pathophysiological event involves this cell type This pathophysiological event involves astrocyte (CL:0000127). CL:0000127 is a cell type from the Cell Ontology.
Show evidence (1 reference)
PMID:41066510 SUPPORT Model Organism
"sequential activation of different glial cell types, lipid droplet accumulation in astrocytes"
Records the staged glial response and the astrocytic lipid phenotype.
Neuronal Network Hyperexcitability
The circuit-level state that produces seizures. In the mouse model this is not a fixed property but evolves through distinct stages, which is what makes the model's temporal structure informative rather than incidental.
Animal models
Xenopus laevis cyfip2 gain-of-function tadpole Xenopus laevis recapitulates Expression of either pathogenic variant produced spontaneous epileptic activity in the brain.
Downstream
Drug-Resistant Epilepsy Network hyperexcitability manifests clinically as the seizure disorder.
Show evidence (1 reference)
PMID:41066510 SUPPORT Model Organism
"after an initial period of neonatal spasms, Cyfip2+/R87C mice entered a seizure-free phase, followed by spontaneous recurrent seizures in adulthood"
Establishes that network excitability in this model changes over time in three distinct phases.
Drug-Resistant Epilepsy
Early-onset seizures that characteristically resist antiseizure medication, frequently presenting as epileptic spasms in a West syndrome pattern.
Animal models
Cyfip2 p.Arg87Cys knock-in mouse Mouse partially recapitulates Reproduces spasm-like movements and, later, spontaneous seizures.
Xenopus laevis cyfip2 gain-of-function tadpole Xenopus laevis partially recapitulates Seizure-related behaviour, variant-dependent.
Show evidence (1 reference)
PMID:29534297 SUPPORT Human Clinical
"We identified three de novo CYFIP2 variants at the Arg87 residue in 4 unrelated individuals with early-onset epileptic encephalopathy"
Establishes the early-onset epileptic encephalopathy phenotype in the founding cohort.
⬡

Pathograph

Use the checkboxes to hide or show graph categories. Hover nodes for evidence and cross-linked metadata.
Pathograph: causal mechanism network for CYFIP2-Related Developmental and Epileptic Encephalopathy Interactive directed graph showing how pathophysiology mechanisms, phenotypes, genetic factors and variants, experimental models, environmental triggers, and treatments relate through causal and linked edges.
●

Phenotypes

8
Head and Neck 1
Microcephaly HP:0000252 Human Phenotype Ontology (HP) Relation: this clinical feature is this phenotype This clinical feature is Microcephaly (HP:0000252). HP:0000252 is a phenotype from the Human Phenotype Ontology.
Show evidence (1 reference)
PMID:36251395 SUPPORT Model Organism
"The Cyfip2+/R87C mice recapitulated many neurological and neurobehavioral phenotypes of the patients, including spasmlike movements, microcephaly, and impaired social communication"
Model-organism evidence that microcephaly is among the patient phenotypes the mouse reproduces. Human-side support for microcephaly in this entry rests on this statement about patient phenotypes rather than on a primary human cohort count.
Musculoskeletal 1
Muscular Hypotonia HP:0001252 Human Phenotype Ontology (HP) Relation: this clinical feature is this phenotype This clinical feature is Hypotonia (HP:0001252). HP:0001252 is a phenotype from the Human Phenotype Ontology.
Show evidence (1 reference)
PMID:30664714 SUPPORT Human Clinical
"a shared phenotype of intellectual disability, seizures, and muscular hypotonia"
Names muscular hypotonia as the third shared feature of the cohort.
Nervous System 3
Seizures FREQUENT HP:0001250 Human Phenotype Ontology (HP) Relation: this clinical feature is this phenotype This clinical feature is Seizure (HP:0001250). HP:0001250 is a phenotype from the Human Phenotype Ontology.
Show evidence (2 references)
PMID:30664714 SUPPORT Human Clinical
"a shared phenotype of intellectual disability, seizures, and muscular hypotonia"
Names seizures as one of three shared features across the twelve-patient cohort.
DOI:10.1002/cns3.70036 SUPPORT Human Clinical
"All individuals had global developmental delay, 32/42 (76%) developed epilepsy, and 25/42 (60%) experienced seizure onset in the first year of life"
Quantifies epilepsy at 76% across a 42-patient literature review, which places it in the 30-79% FREQUENT band and supplies the denominator the frequency claim needs. Notably epilepsy is not universal in this disorder, which the earlier small series could not have shown.
Epileptic Spasms FREQUENT HP:0011097 Human Phenotype Ontology (HP) Relation: this clinical feature is this phenotype This clinical feature is Epileptic spasm (HP:0011097). HP:0011097 is a phenotype from the Human Phenotype Ontology.
Show evidence (2 references)
DOI:10.1002/cns3.70036 SUPPORT Human Clinical
"Of those with seizures, 16/32 (50%) had epileptic spasms"
Quantifies epileptic spasms at 50% of patients with seizures, in the 30-79% FREQUENT band. The denominator is patients with seizures, not all patients.
PMID:36251395 SUPPORT Human Clinical
"the hotspot p.Arg87Cys variant of the cytoplasmic FMR1-interacting protein 2 (CYFIP2) gene, which was recently identified in individuals diagnosed with West syndrome, a developmental and epileptic encephalopathy"
Ties the hotspot variant to a West syndrome diagnosis in affected individuals.
Intellectual Disability OBLIGATE HP:0001249 Human Phenotype Ontology (HP) Relation: this clinical feature is this phenotype This clinical feature is Intellectual disability (HP:0001249). HP:0001249 is a phenotype from the Human Phenotype Ontology.
Show evidence (2 references)
PMID:30664714 SUPPORT Human Clinical
"a shared phenotype of intellectual disability, seizures, and muscular hypotonia"
Names intellectual disability as a shared feature of the cohort.
DOI:10.1002/cns3.70036 SUPPORT Human Clinical
"All individuals had global developmental delay"
Developmental impairment was present in every one of the 42 patients reviewed, supporting the OBLIGATE band. This is the one feature of the disorder that is genuinely universal, in contrast to epilepsy at 76%. Note the counted phenotype is global developmental delay rather than intellectual disability as such; in a cohort of this severity the two are taken as equivalent, but the band rests on the delay count, not on formal cognitive testing.
Other 3
Drug-Resistant Epilepsy Refractory drug response HP:0020174 Human Phenotype Ontology (HP) Relation: this clinical feature is this phenotype This clinical feature is Refractory drug response (HP:0020174). HP:0020174 is a phenotype from the Human Phenotype Ontology.
Show evidence (2 references)
PMID:36968925 SUPPORT Human Clinical
"The seizures were of different types, frequent and refractory to treatment, including different anticonvulsant drugs"
Documents refractoriness across multiple anticonvulsant classes in a genotype-confirmed patient.
PMID:36968925 SUPPORT Human Clinical
"He maintained refractory epilepsy despite being medicated with vigabatrin, phenobarbital, valproic acid, and a ketogenic diet"
Names the specific agents that failed, which is stronger than a general statement of drug resistance.
Impaired Social Communication
Deliberately left unbound - needs term / NTR. HP:0000735 "Impaired social interactions" is obsolete, and the available alternatives do not fit: HP:0008763 "No social interaction" is absolute where the source says impaired, and HP:5200026 "Impaired social imitation" is narrower than the claim. Binding HP:0000729 "Autistic behavior" would be an interpretive leap beyond what the quoted source states. A precise unbound descriptor is preferred here over a term that would misstate the finding.
Show evidence (1 reference)
PMID:36251395 SUPPORT Model Organism
"including spasmlike movements, microcephaly, and impaired social communication"
Lists impaired social communication among the patient phenotypes reproduced in the model.
Normal or Nonspecific Brain MRI
Show evidence (1 reference)
PMID:36968925 SUPPORT Human Clinical
"Brain MRI showed no pathological changes"
Records a normal MRI in a genotype-confirmed infant with severe early infantile epileptic encephalopathy.
🧬

Genetic Associations

1
CYFIP2 (Heterozygous de novo missense variants in CYFIP2 cause DEE65. Variants cluster at two hotspots, p.Arg87 and p.Asp724, and the hotspot identity predicts severity: Arg87 substitutions give a consistently severe phenotype while Asp724 and other substitutions give a more variable one.)
Gene: CYFIP2 hgnc:13760 HUGO Gene Nomenclature Committee (hgnc) Relation: this disease-associated gene is this gene This disease-associated gene is CYFIP2 (hgnc:13760). hgnc:13760 is a gene from the HUGO Gene Nomenclature Committee. relationship_type: CAUSATIVE
Show evidence (4 references)
PMID:33149277 SUPPORT Human Clinical
"Genotype-phenotype correlation confirms a consistently severe phenotype in p.Arg87 patients but a more variable phenotype in p.Asp724 and other substitutions"
States the hotspot-dependent severity difference that makes the genotype prognostically informative: Arg87 consistently severe, Asp724 variable.
PMID:30664714 SUPPORT Human Clinical
"We here report 12 independent patients harboring a variety of de novo variants in CYFIP2 broadening the molecular and clinical spectrum of a novel CYFIP2-related neurodevelopmental disorder"
Establishes the breadth of the variant spectrum beyond the original hotspot.
PMID:33149277 SUPPORT In Vitro
"By structural modeling and investigation of WAVE-regulatory complex (WRC)-mediated actin polymerization in six patient fibroblast lines we assessed the impact of CYFIP2 variants on the WRC"
Records that the functional assessment was done in six primary patient fibroblast lines, which is what moves the actin finding out of heterologous overexpression.
+ 1 more reference
💊

Medical Actions

1
Antiseizure Medication
Action: PharmacotherapyNCI Thesaurus (NCIT) Relation: this treatment is this clinical intervention This treatment is Pharmacotherapy (NCIT:C15986). NCIT:C15986 is a clinical intervention from the NCI Thesaurus. NCIT:C15986
Seizures are characteristically drug-resistant and typically require multiple agents. No agent is established as preferred in this disorder.
Show evidence (1 reference)
PMID:36968925 SUPPORT Human Clinical
"When he started having infantile spasms, and there was evidence of small fragments of disorganized activity on EEG, vigabatrin was also started with transitory clinical improvement"
Records vigabatrin producing only transitory improvement. Marked PARTIAL because it documents an attempted treatment and its limited effect, not an effective one.
🔬

Diagnosis

2
Trio Exome or Genome Sequencing
The disorder is defined genotypically. Because MRI is often normal and the clinical and EEG picture is shared with many other genetic developmental and epileptic encephalopathies, trio-based broad sequencing is what makes the diagnosis, and the trio design is what establishes de novo status.
Whole Exome Sequencing NCIT:C101295 NCI Thesaurus (NCIT)
Show evidence (1 reference)
PMID:29534297 SUPPORT Human Clinical
"We performed trio-based whole-exome sequencing (WES) in 210 families and case-only WES in 489 individuals with epileptic encephalopathies"
Describes the sequencing strategy through which the gene-disease relationship was established.
Electroencephalography
EEG is central to recognising the syndrome even though MRI is often normal. The interictal patterns reported span burst-suppression and hypsarrhythmia, mapping onto Ohtahara and West syndrome respectively - so a single genotype can present under more than one electroclinical syndrome label.
Electroencephalography NCIT:C38054 NCI Thesaurus (NCIT)
Show evidence (1 reference)
PMID:36968925 SUPPORT Human Clinical
"Initial interictal electroencephalograms showed a burst-suppression pattern and/or hypsarrhythmia, which were compatible with Ohtahara syndrome and West syndrome, respectively"
Documents both EEG patterns and the two electroclinical syndromes they correspond to, in a genotype-confirmed patient.
📊

Prevalence

1
Worldwide, published cases
Cases In Literature Ultra Rare
42 individuals with pathogenic or likely pathogenic CYFIP2 variants had been reported at the time of the most recent literature review. No population prevalence estimate exists; all knowledge derives from clinical sequencing referral cohorts rather than population-based ascertainment, so the true frequency is unknown and probably underestimated.
Show evidence (2 references)
DOI:10.1002/cns3.70036 SUPPORT Human Clinical
"We identified 41 additional patients with pathogenic or likely pathogenic"
The literature review that establishes the cumulative case count, 41 additional patients alongside the index case reported in the same paper.
PMID:33149277 SUPPORT Human Clinical
"a total of only 18 individuals with CYFIP2 variants have been described to date"
The earlier cumulative count, retained to show the trajectory of ascertainment - the reported cohort more than doubled between these two reviews.
🐁

Animal Models

2
Cyfip2 p.Arg87Cys knock-in mouse
A genotype-matched knock-in carrying the human hotspot variant. It is unusually faithful for an epilepsy model, reproducing not just seizures but the specific spasm-like semiology, and it is the source of the protein-destabilization evidence.
Species
Mouse
Genotype
Cyfip2+/R87C
Publication
PMID:36251395
Pathograph links
Enhanced CYFIP2 Ubiquitination and Proteasomal Degradation recapitulates fidelity: high The model in which the degradation mechanism was demonstrated, in brain tissue.
Readouts
Brain CYFIP2 protein abundance
Direction: DECREASED
Interpretation: Steady-state CYFIP2 is reduced in the brains of knock-in mice.
Show evidence (1 reference)
PMID:36251395 SUPPORT Model Organism
"Beyond identifying a decrease in CYFIP2 protein levels in the Cyfip2+/R87C brains"
The abundance measurement behind this readout.
Show evidence (1 reference)
PMID:36251395 SUPPORT Model Organism
"we demonstrated that the p.Arg87Cys variant enhances ubiquitination and proteasomal degradation of CYFIP2"
Supports the model as the system in which this mechanism was established.
Drug-Resistant Epilepsy partially recapitulates fidelity: moderate Reproduces spasm-like movements and, later, spontaneous seizures.
Limitations: The mouse reproduces the neonatal spasm phase, but then passes through a seizure-free interval before developing spontaneous recurrent seizures in adulthood. Human DEE65 is not described as having such a latent phase, so the model's temporal structure does not match the human course even though both endpoints are seizures.
Readouts
Seizure phenotype across the lifespan
Direction: ALTERED
Interpretation: Three distinct phases - neonatal spasms, a seizure-free interval, then adult spontaneous recurrent seizures ending in premature death.
Show evidence (1 reference)
PMID:41066510 SUPPORT Model Organism
"Our findings reveal that Cyfip2+/R87C mice, following a seizure-free period after neonatal spasms, develop spontaneous recurrent seizures in adulthood, ultimately leading to premature death"
The longitudinal seizure phenotyping behind this readout.
Show evidence (1 reference)
PMID:36251395 SUPPORT Model Organism
"The Cyfip2+/R87C mice recapitulated many neurological and neurobehavioral phenotypes of the patients, including spasmlike movements, microcephaly, and impaired social communication"
Supports treating the model as informative for the clinical phenotype, including the specific spasm semiology.
Xenopus laevis cyfip2 gain-of-function tadpole
A transient overexpression model rather than a knock-in, chosen because gain-of-function variants cannot be modelled by the knockdown and knockout techniques used for loss of function. It is the only in vivo demonstration that a pathogenic CYFIP2 variant alone is sufficient to produce epileptic brain activity, which is why it matters to the canonical hypothesis.
Species
Xenopus laevis
Genotype
transient ectopic expression of Arg87Cys or Tyr108Cys cyfip2 mRNA
Publication
DOI:10.1101/2022.12.07.519540
Pathograph links
Neuronal Network Hyperexcitability recapitulates fidelity: moderate Expression of either pathogenic variant produced spontaneous epileptic activity in the brain.
Limitations: Transient ectopic mRNA overexpression rather than expression from the endogenous locus, so protein level is not physiological and the model cannot speak to the competing protein-destabilization account, which is a claim about steady-state abundance. Published as a preprint and not peer-reviewed at time of curation.
Readouts
Spontaneous epileptic brain activity
Direction: INCREASED
Interpretation: Both variants produced spontaneous epileptic activity, and it was reduced by valproate - a pharmacological control that argues the activity is genuinely seizure-like rather than an artefact of overexpression.
Show evidence (2 references)
DOI:10.1101/2022.12.07.519540 Preprint · not peer-reviewed SUPPORT Model Organism
"Expression of either pathogenic variant resulted in spontaneous epileptic activity in the brain"
The electrophysiological measurement behind this readout.
+ 1 more reference
Show evidence (1 reference)
DOI:10.1101/2022.12.07.519540 Preprint · not peer-reviewed SUPPORT Model Organism
"transient ectopic expression of the Arg87Cys pathogenic variant of"
Establishes that the model expresses the human hotspot variant, which is what makes it informative for this disorder rather than for CYFIP2 biology generally.
Drug-Resistant Epilepsy partially recapitulates fidelity: low Seizure-related behaviour, variant-dependent.
Limitations: Only one of the two variants tested produced the behavioural phenotype. Arg87Cys caused seizure-related darting and circling; Tyr108Cys did not alter behaviour at all, despite producing the same epileptic brain activity. So the model reproduces behavioural seizures for some pathogenic variants and not others, and a behavioural readout alone would have scored Tyr108Cys as unaffected. A second divergence matters specifically for anyone using this model as the pre-clinical drug-testing platform its authors propose: the tadpole hyperactivity was reduced by valproate, whereas the human phenotype curated in this entry is drug-resistant epilepsy in a patient who remained refractory despite valproic acid among four treatments. The model is valproate-responsive against a valproate-refractory human disorder, so a compound that suppresses activity here has not thereby been shown to work in patients.
Readouts
Seizure-related swimming behaviour
Direction: INCREASED
Interpretation: Increased for Arg87Cys; explicitly unchanged for Tyr108Cys.
Show evidence (1 reference)
DOI:10.1101/2022.12.07.519540 Preprint · not peer-reviewed SUPPORT Model Organism
"In contrast, expression of a second pathological variant, Tyr108Cys, did not alter tadpole behaviour"
Curated as PARTIAL because it is a negative behavioural result for one variant alongside a positive one for the other. The dissociation is the informative part: brain activity and behaviour come apart within the same model.
Show evidence (1 reference)
DOI:10.1101/2022.12.07.519540 Preprint · not peer-reviewed SUPPORT Model Organism
"resulted in increased seizure-related behaviours such as rapid darting and swimming in circles"
Supports the behavioural phenotype for the Arg87Cys variant specifically.
{ }

Source YAML

click to show 
name: CYFIP2-Related Developmental and Epileptic Encephalopathy
creation_date: "2026-08-20T00:20:00Z"
category: Mendelian
synonyms:
- Developmental and epileptic encephalopathy 65
- DEE65
- EIEE65
- CYFIP2-related neurodevelopmental disorder
description: >-
  CYFIP2-related developmental and epileptic encephalopathy (DEE65) is an autosomal
  dominant disorder caused by heterozygous, almost always de novo, missense variants in
  CYFIP2. CYFIP2 is a structural subunit of the WAVE regulatory complex, which holds the
  actin-nucleating VCA domain of WAVE in an autoinhibited state until RAC1 signalling
  releases it to drive Arp2/3-mediated actin polymerization. Pathogenic variants cluster
  at two hotspots - p.Arg87, which sits buried at the CYFIP2-WAVE1 interface and produces
  a consistently severe phenotype, and p.Asp724, which produces a more variable one.
  Affected children have early-onset drug-resistant seizures, frequently epileptic spasms
  in a West syndrome pattern, with severe developmental impairment and hypotonia. Brain
  MRI is often unremarkable despite the severity of the clinical and electrical picture.
disease_term:
  preferred_term: developmental and epileptic encephalopathy, 65
  term:
    id: MONDO:0033374
    label: developmental and epileptic encephalopathy, 65
parents:
- Epilepsy
- Neurological Disease
inheritance:
- name: Autosomal dominant
  inheritance_term:
    preferred_term: Autosomal dominant inheritance
    term:
      id: HP:0000006
      label: Autosomal dominant inheritance
  description: >-
    Heterozygous and overwhelmingly de novo. No confirmed transmission from an affected
    parent has been reported, consistent with the severity of the phenotype.
  evidence:
  - reference: PMID:30664714
    reference_title: "Spatially clustering de novo variants in CYFIP2, encoding the cytoplasmic FMRP interacting protein 2, cause intellectual disability and seizures."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: "we identified 12 independent patients carrying a total of eight distinct de
      novo variants in CYFIP2 with a shared phenotype of intellectual disability,
      seizures, and muscular hypotonia"
    explanation: >-
      Establishes the de novo heterozygous architecture across twelve independent
      patients, together with the core phenotype triad.

pathophysiology:

- name: De Novo CYFIP2 Missense Variant at a Structural Hotspot
  biological_scale: MOLECULAR
  description: >-
    The initiating lesion. Variants cluster at two hotspots, p.Arg87 and p.Asp724. Arg87
    is buried at the interface between CYFIP2 and WAVE1 inside the assembled complex,
    which is why substituting it destabilises the interface rather than simply removing a
    surface contact. Variants scattered elsewhere in the primary sequence converge on the
    same interface in three dimensions.
  evidence:
  - reference: PMID:29534297
    reference_title: "De novo hotspot variants in CYFIP2 cause early-onset epileptic encephalopathy."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: "Structural analysis indicated that the Arg87 residue is buried at an
      interface between CYFIP2 and WAVE1, and the Arg87 variant may disrupt hydrogen
      bonding, leading to structural instability and aberrant activation of the WAVE
      regulatory complex"
    explanation: >-
      Locates the hotspot residue structurally and states the predicted consequence for
      the complex.
  - reference: PMID:33149277
    reference_title: "New insights into the clinical and molecular spectrum of the novel CYFIP2-related neurodevelopmental disorder and impairment of the WRC-mediated actin dynamics."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: "Asp724 as second mutational hotspot (4/19 cases)"
    explanation: >-
      Establishes the second hotspot and its frequency in the expanded cohort.
  downstream:
  - target: Weakened CYFIP2-WAVE1 Interface Binding
    causal_link_type: DIRECT
    description: >-
      The structural lesion acts first on the interface it sits in.
  - target: Enhanced CYFIP2 Ubiquitination and Proteasomal Degradation
    causal_link_type: DIRECT
    description: >-
      A second, mechanistically separate consequence demonstrated for p.Arg87Cys. Drawn
      as a parallel edge from the variant rather than downstream of the interface defect,
      because the two are competing accounts of the same variant rather than sequential
      steps - see the controversy discussion.
    hypothesis_groups:
    - protein_destabilization_model

- name: Weakened CYFIP2-WAVE1 Interface Binding
  biological_scale: MOLECULAR
  description: >-
    Mutant CYFIP2 binds the VCA domain more weakly than wild type. Because the complex
    works by sequestering VCA, weaker binding means less effective autoinhibition. This
    node is the measured binding change, kept separate from the downstream actin readout.
  molecular_functions:
  - preferred_term: actin binding
    term:
      id: GO:0003779
      label: actin binding
  cellular_components:
  - preferred_term: WAVE regulatory complex
    term:
      id: GO:0031209
      label: SCAR complex
  evidence:
  - reference: PMID:29534297
    reference_title: "De novo hotspot variants in CYFIP2 cause early-onset epileptic encephalopathy."
    supports: SUPPORT
    evidence_source: IN_VITRO
    snippet: "All mutant CYFIP2 showed comparatively weaker interactions to the VCA domain
      than wild-type CYFIP2"
    explanation: >-
      The direct binding measurement, reported for every mutant tested.
  downstream:
  - target: Aberrant WAVE Regulatory Complex Activation
    causal_link_type: DIRECT
    description: >-
      Loss of VCA sequestration is what releases the complex.
    hypothesis_groups:
    - wrc_gain_of_function_model

- name: Enhanced CYFIP2 Ubiquitination and Proteasomal Degradation
  biological_scale: MOLECULAR
  description: >-
    A second consequence of p.Arg87Cys: the variant protein is ubiquitinated and degraded
    more readily, lowering steady-state CYFIP2 abundance in the brain. This is the
    observation that sits in tension with a pure gain-of-function reading, and it is
    curated as its own node rather than folded into the interface account.
  biological_processes:
  - preferred_term: proteasomal protein catabolic process
    modifier: INCREASED
    term:
      id: GO:0010498
      label: proteasomal protein catabolic process
  evidence:
  - reference: PMID:36251395
    reference_title: "CYFIP2 p.Arg87Cys Causes Neurological Defects and Degradation of CYFIP2."
    supports: SUPPORT
    evidence_source: MODEL_ORGANISM
    snippet: "we demonstrated that the p.Arg87Cys variant enhances ubiquitination and
      proteasomal degradation of CYFIP2"
    explanation: >-
      Direct demonstration of the degradation mechanism in the knock-in mouse and cellular
      models.
  - reference: PMID:36251395
    reference_title: "CYFIP2 p.Arg87Cys Causes Neurological Defects and Degradation of CYFIP2."
    supports: SUPPORT
    evidence_source: MODEL_ORGANISM
    snippet: "Beyond identifying a decrease in CYFIP2 protein levels in the Cyfip2+/R87C
      brains"
    explanation: >-
      Records the reduced steady-state protein abundance that the degradation mechanism
      produces, in brain tissue.
  downstream:
  - target: Dysregulated Actin Cytoskeleton Dynamics
    causal_link_type: INDIRECT_UNKNOWN_INTERMEDIATES
    description: >-
      Reduced CYFIP2 abundance is proposed to reach the same actin readout by a different
      route than interface destabilization.
    hypothesis_groups:
    - protein_destabilization_model

- name: Aberrant WAVE Regulatory Complex Activation
  biological_scale: MOLECULAR
  description: >-
    Release of the WAVE VCA domain from autoinhibition, activating the complex outside
    normal RAC1 control. This is the node the gain-of-function account turns on.
  cellular_components:
  - preferred_term: WAVE regulatory complex
    term:
      id: GO:0031209
      label: SCAR complex
  biological_processes:
  - preferred_term: Arp2/3 complex-mediated actin nucleation
    modifier: INCREASED
    term:
      id: GO:0034314
      label: Arp2/3 complex-mediated actin nucleation
  evidence:
  - reference: PMID:29534297
    reference_title: "De novo hotspot variants in CYFIP2 cause early-onset epileptic encephalopathy."
    supports: SUPPORT
    evidence_source: IN_VITRO
    snippet: "Our findings suggest that de novo Arg87 variants in CYFIP2 have
      gain-of-function effects on the WAVE signaling pathway and are associated with
      severe neurological disorders"
    explanation: >-
      States the gain-of-function interpretation of the WAVE pathway effect.
  downstream:
  - target: Dysregulated Actin Cytoskeleton Dynamics
    causal_link_type: DIRECT
    description: >-
      An activated complex nucleates actin.
    hypothesis_groups:
    - wrc_gain_of_function_model

- name: Dysregulated Actin Cytoskeleton Dynamics
  biological_scale: CELLULAR
  description: >-
    The cellular convergence point of both proposed molecular routes. Abnormal actin
    accumulation was shown first in transfected cells and subsequently substantiated in
    patient-derived fibroblasts, which matters because it moves the finding out of an
    overexpression system into primary patient cells.
  biological_processes:
  - preferred_term: actin cytoskeleton organization
    modifier: DYSREGULATED
    term:
      id: GO:0030036
      label: actin cytoskeleton organization
  evidence:
  - reference: PMID:29534297
    reference_title: "De novo hotspot variants in CYFIP2 cause early-onset epileptic encephalopathy."
    supports: SUPPORT
    evidence_source: IN_VITRO
    snippet: "Immunofluorescence revealed that ectopic speckled accumulation of actin and
      CYFIP2 was significantly increased in cells transfected with mutant CYFIP2"
    explanation: >-
      The original actin readout in a heterologous transfection system.
  - reference: PMID:33149277
    reference_title: "New insights into the clinical and molecular spectrum of the novel CYFIP2-related neurodevelopmental disorder and impairment of the WRC-mediated actin dynamics."
    supports: SUPPORT
    evidence_source: IN_VITRO
    snippet: "Consistent with its role in WRC-mediated actin polymerization we substantiate
      aberrant regulation of the actin cytoskeleton in patient fibroblasts"
    explanation: >-
      Confirms the same cellular abnormality in primary cells from affected individuals
      rather than in an overexpression model.
  downstream:
  - target: Disrupted Synaptic and Dendritic Spine Remodeling
    causal_link_type: INDIRECT_UNKNOWN_INTERMEDIATES
    description: >-
      Actin dynamics are the substrate for spine and synapse remodeling in neurons.

- name: Disrupted Synaptic and Dendritic Spine Remodeling
  biological_scale: CELLULAR
  description: >-
    In neurons the WAVE complex drives actin remodeling at dendritic spines and growth
    cones, so dysregulated actin dynamics translate into abnormal synaptic architecture.
    The knock-in mouse shows age-progressive cytoarchitectural disorganization and gliosis
    in the hippocampus, and longitudinal work shows the synaptic remodeling is itself
    time-dependent rather than a fixed lesion.
  conforms_to: "epilepsy_excitation_inhibition_imbalance#Ion Channel and Synaptic Dysfunction"
  cellular_components:
  - preferred_term: dendritic spine
    term:
      id: GO:0043197
      label: dendritic spine
  cell_types:
  - preferred_term: glutamatergic neuron
    term:
      id: CL:0000679
      label: glutamatergic neuron
  evidence:
  - reference: PMID:36251395
    reference_title: "CYFIP2 p.Arg87Cys Causes Neurological Defects and Degradation of CYFIP2."
    supports: SUPPORT
    evidence_source: MODEL_ORGANISM
    snippet: "Age-progressive cytoarchitectural disorganization and gliosis were also
      identified in the hippocampus of Cyfip2+/R87C mice"
    explanation: >-
      Structural correlate in the hippocampus of the genotype-matched mouse, and records
      that it worsens with age.
  - reference: PMID:41066510
    reference_title: "Seizure evolution in a mouse model of West syndrome involves complex and time-dependent synapse remodeling, gliosis and alterations in lipid metabolism."
    supports: SUPPORT
    evidence_source: MODEL_ORGANISM
    snippet: "This progression was associated with synaptic remodeling, sequential
      activation of different glial cell types, lipid droplet accumulation in astrocytes,
      and significant proteomic and lipidomic changes in the brain"
    explanation: >-
      Establishes synaptic remodeling and staged glial involvement as accompaniments of
      seizure evolution rather than a static finding.
  downstream:
  - target: Reactive Gliosis
    causal_link_type: INDIRECT_UNKNOWN_INTERMEDIATES
    description: >-
      Glial activation accompanies the evolving synaptic pathology in the model.
  - target: Neuronal Network Hyperexcitability
    causal_link_type: INDIRECT_UNKNOWN_INTERMEDIATES
    description: >-
      Abnormal synaptic architecture is the proposed substrate for network
      hyperexcitability.

- name: Reactive Gliosis
  biological_scale: TISSUE
  description: >-
    Sequential activation of different glial populations over the disease course in the
    mouse model, accompanied by lipid droplet accumulation in astrocytes. Curated as its
    own node because it is a distinct tissue-level process with its own time course, not
    a restatement of the synaptic finding.
  cell_types:
  - preferred_term: astrocyte
    term:
      id: CL:0000127
      label: astrocyte
  evidence:
  - reference: PMID:41066510
    reference_title: "Seizure evolution in a mouse model of West syndrome involves complex and time-dependent synapse remodeling, gliosis and alterations in lipid metabolism."
    supports: SUPPORT
    evidence_source: MODEL_ORGANISM
    snippet: "sequential activation of different glial cell types, lipid droplet
      accumulation in astrocytes"
    explanation: >-
      Records the staged glial response and the astrocytic lipid phenotype.

- name: Neuronal Network Hyperexcitability
  biological_scale: CELLULAR
  description: >-
    The circuit-level state that produces seizures. In the mouse model this is not a fixed
    property but evolves through distinct stages, which is what makes the model's temporal
    structure informative rather than incidental.
  conforms_to: "epilepsy_excitation_inhibition_imbalance#Neuronal Hyperexcitability and Hypersynchrony"
  evidence:
  - reference: PMID:41066510
    reference_title: "Seizure evolution in a mouse model of West syndrome involves complex and time-dependent synapse remodeling, gliosis and alterations in lipid metabolism."
    supports: SUPPORT
    evidence_source: MODEL_ORGANISM
    snippet: "after an initial period of neonatal spasms, Cyfip2+/R87C mice entered a
      seizure-free phase, followed by spontaneous recurrent seizures in adulthood"
    explanation: >-
      Establishes that network excitability in this model changes over time in three
      distinct phases.
  downstream:
  - target: Drug-Resistant Epilepsy
    causal_link_type: DIRECT
    description: >-
      Network hyperexcitability manifests clinically as the seizure disorder.

- name: Drug-Resistant Epilepsy
  biological_scale: ORGANISM
  description: >-
    Early-onset seizures that characteristically resist antiseizure medication, frequently
    presenting as epileptic spasms in a West syndrome pattern.
  conforms_to: "epilepsy_excitation_inhibition_imbalance#Recurrent Unprovoked Seizures"
  evidence:
  - reference: PMID:29534297
    reference_title: "De novo hotspot variants in CYFIP2 cause early-onset epileptic encephalopathy."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: "We identified three de novo CYFIP2 variants at the Arg87 residue in 4
      unrelated individuals with early-onset epileptic encephalopathy"
    explanation: >-
      Establishes the early-onset epileptic encephalopathy phenotype in the founding
      cohort.

mechanistic_hypotheses:
- hypothesis_group_id: wrc_gain_of_function_model
  status: CANONICAL
  description: >-
    Hotspot variants destabilize the CYFIP2-WAVE1 interface, weakening sequestration of
    the VCA domain and releasing the complex to nucleate actin outside normal RAC1
    control. On this account the disorder is a toxic gain of function at the level of
    actin regulation, which distinguishes it from the many developmental and epileptic
    encephalopathies that act by haploinsufficiency.
  evidence:
  - reference: PMID:29534297
    reference_title: "De novo hotspot variants in CYFIP2 cause early-onset epileptic encephalopathy."
    supports: SUPPORT
    evidence_source: IN_VITRO
    snippet: "Our findings suggest that de novo Arg87 variants in CYFIP2 have
      gain-of-function effects on the WAVE signaling pathway"
    explanation: >-
      The originating statement of the gain-of-function model.

- hypothesis_group_id: protein_destabilization_model
  status: ALTERNATIVE
  description: >-
    p.Arg87Cys enhances ubiquitination and proteasomal degradation of CYFIP2, lowering
    steady-state protein levels in the brain. On this account the primary consequence is
    less CYFIP2, not a constitutively active complex. The two accounts are not trivially
    compatible: one says the complex is inappropriately switched on, the other that there
    is less of a required subunit.
  evidence:
  - reference: PMID:36251395
    reference_title: "CYFIP2 p.Arg87Cys Causes Neurological Defects and Degradation of CYFIP2."
    supports: SUPPORT
    evidence_source: MODEL_ORGANISM
    snippet: "Beyond identifying a decrease in CYFIP2 protein levels in the Cyfip2+/R87C
      brains, we demonstrated that the p.Arg87Cys variant enhances ubiquitination and
      proteasomal degradation of CYFIP2"
    explanation: >-
      The originating statement of the destabilization model, measured in brain tissue of
      the genotype-matched mouse.

phenotypes:

- category: Neurological
  name: Seizures
  description: >-
    Early-onset seizures are the presenting feature in most affected children, but they are
    not universal: the largest pooled review puts epilepsy at 32 of 42 patients. A child
    with a pathogenic CYFIP2 variant and no seizures is therefore an expected part of the
    spectrum rather than an anomaly.
  frequency: FREQUENT
  phenotype_term:
    preferred_term: Seizure
    term:
      id: HP:0001250
      label: Seizure
  evidence:
  - reference: PMID:30664714
    reference_title: "Spatially clustering de novo variants in CYFIP2, encoding the cytoplasmic FMRP interacting protein 2, cause intellectual disability and seizures."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: "a shared phenotype of intellectual disability, seizures, and muscular
      hypotonia"
    explanation: >-
      Names seizures as one of three shared features across the twelve-patient cohort.
  - reference: DOI:10.1002/cns3.70036
    reference_title: "Expanding the Phenotype of <i>CYFIP2‐</i> Related Developmental Epileptic Encephalopathy: Case Report and Literature Review"
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: "All individuals had global developmental delay, 32/42 (76%) developed
      epilepsy, and 25/42 (60%) experienced seizure onset in the first year of life"
    explanation: >-
      Quantifies epilepsy at 76% across a 42-patient literature review, which places it in
      the 30-79% FREQUENT band and supplies the denominator the frequency claim needs.
      Notably epilepsy is not universal in this disorder, which the earlier small series
      could not have shown.

- category: Neurological
  name: Epileptic Spasms
  description: >-
    Epileptic spasms in a West syndrome pattern are the characteristic seizure type,
    particularly in Arg87 carriers. The knock-in mouse reproduces spasm-like movements,
    which is unusual - most epilepsy models do not reproduce the specific seizure
    semiology of the disorder they model.
  frequency: FREQUENT
  phenotype_term:
    preferred_term: Epileptic spasm
    term:
      id: HP:0011097
      label: Epileptic spasm
  evidence:
  - reference: DOI:10.1002/cns3.70036
    reference_title: "Expanding the Phenotype of <i>CYFIP2‐</i> Related Developmental Epileptic Encephalopathy: Case Report and Literature Review"
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: "Of those with seizures, 16/32 (50%) had epileptic spasms"
    explanation: >-
      Quantifies epileptic spasms at 50% of patients with seizures, in the 30-79%
      FREQUENT band. The denominator is patients with seizures, not all patients.
  - reference: PMID:36251395
    reference_title: "CYFIP2 p.Arg87Cys Causes Neurological Defects and Degradation of CYFIP2."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: "the hotspot p.Arg87Cys variant of the cytoplasmic FMR1-interacting protein 2
      (CYFIP2) gene, which was recently identified in individuals diagnosed with West
      syndrome, a developmental and epileptic encephalopathy"
    explanation: >-
      Ties the hotspot variant to a West syndrome diagnosis in affected individuals.

- category: Neurological
  name: Drug-Resistant Epilepsy
  description: >-
    Seizures characteristically resist antiseizure medication from onset.
  phenotype_term:
    preferred_term: Refractory drug response
    term:
      id: HP:0020174
      label: Refractory drug response
  evidence:
  - reference: PMID:36968925
    reference_title: "Clinical Role of Codon 87 of the CYFIP2 Gene in Early Infantile Epileptic Encephalopathy: A Clinical Case Description."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: "The seizures were of different types, frequent and refractory to treatment,
      including different anticonvulsant drugs"
    explanation: >-
      Documents refractoriness across multiple anticonvulsant classes in a
      genotype-confirmed patient.
  - reference: PMID:36968925
    reference_title: "Clinical Role of Codon 87 of the CYFIP2 Gene in Early Infantile Epileptic Encephalopathy: A Clinical Case Description."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: "He maintained refractory epilepsy despite being medicated with vigabatrin,
      phenobarbital, valproic acid, and a ketogenic diet"
    explanation: >-
      Names the specific agents that failed, which is stronger than a general statement of
      drug resistance.

- category: Neurological
  name: Intellectual Disability
  description: >-
    Severe to profound intellectual impairment, present in essentially all reported
    individuals. Severity tracks genotype - consistently severe in Arg87 carriers, more
    variable otherwise.
  frequency: OBLIGATE
  phenotype_term:
    preferred_term: Intellectual disability
    term:
      id: HP:0001249
      label: Intellectual disability
  evidence:
  - reference: PMID:30664714
    reference_title: "Spatially clustering de novo variants in CYFIP2, encoding the cytoplasmic FMRP interacting protein 2, cause intellectual disability and seizures."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: "a shared phenotype of intellectual disability, seizures, and muscular
      hypotonia"
    explanation: >-
      Names intellectual disability as a shared feature of the cohort.
  - reference: DOI:10.1002/cns3.70036
    reference_title: "Expanding the Phenotype of <i>CYFIP2‐</i> Related Developmental Epileptic Encephalopathy: Case Report and Literature Review"
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: "All individuals had global developmental delay"
    explanation: >-
      Developmental impairment was present in every one of the 42 patients reviewed,
      supporting the OBLIGATE band. This is the one feature of the disorder that is
      genuinely universal, in contrast to epilepsy at 76%. Note the counted phenotype is
      global developmental delay rather than intellectual disability as such; in a cohort
      of this severity the two are taken as equivalent, but the band rests on the delay
      count, not on formal cognitive testing.

- category: Neurological
  name: Muscular Hypotonia
  description: >-
    Generalized or truncal hypotonia, present from infancy.
  phenotype_term:
    preferred_term: Hypotonia
    term:
      id: HP:0001252
      label: Hypotonia
  evidence:
  - reference: PMID:30664714
    reference_title: "Spatially clustering de novo variants in CYFIP2, encoding the cytoplasmic FMRP interacting protein 2, cause intellectual disability and seizures."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: "a shared phenotype of intellectual disability, seizures, and muscular
      hypotonia"
    explanation: >-
      Names muscular hypotonia as the third shared feature of the cohort.

- category: Neurological
  name: Microcephaly
  description: >-
    Microcephaly is common and is reproduced in the knock-in mouse.
  phenotype_term:
    preferred_term: Microcephaly
    term:
      id: HP:0000252
      label: Microcephaly
  evidence:
  - reference: PMID:36251395
    reference_title: "CYFIP2 p.Arg87Cys Causes Neurological Defects and Degradation of CYFIP2."
    supports: SUPPORT
    evidence_source: MODEL_ORGANISM
    snippet: "The Cyfip2+/R87C mice recapitulated many neurological and neurobehavioral
      phenotypes of the patients, including spasmlike movements, microcephaly, and
      impaired social communication"
    explanation: >-
      Model-organism evidence that microcephaly is among the patient phenotypes the mouse
      reproduces. Human-side support for microcephaly in this entry rests on this
      statement about patient phenotypes rather than on a primary human cohort count.

- category: Neurodevelopmental
  name: Impaired Social Communication
  description: >-
    Autistic features and impaired social communication are documented in affected
    individuals and reproduced in the mouse as impaired ultrasonic vocalization and social
    interaction deficits.
  phenotype_term:
    preferred_term: Impaired social communication
  evidence:
  - reference: PMID:36251395
    reference_title: "CYFIP2 p.Arg87Cys Causes Neurological Defects and Degradation of CYFIP2."
    supports: SUPPORT
    evidence_source: MODEL_ORGANISM
    snippet: "including spasmlike movements, microcephaly, and impaired social
      communication"
    explanation: >-
      Lists impaired social communication among the patient phenotypes reproduced in the
      model.
  notes: >-
    Deliberately left unbound - needs term / NTR. HP:0000735 "Impaired social
    interactions" is obsolete, and the available alternatives do not fit: HP:0008763 "No
    social interaction" is absolute where the source says impaired, and HP:5200026
    "Impaired social imitation" is narrower than the claim. Binding HP:0000729 "Autistic
    behavior" would be an interpretive leap beyond what the quoted source states. A
    precise unbound descriptor is preferred here over a term that would misstate the
    finding.

- category: Neuroimaging
  name: Normal or Nonspecific Brain MRI
  description: >-
    Brain MRI is frequently unremarkable despite a severe clinical and electrical picture.
    This is a positive curatorial point rather than an absence of information: a normal
    MRI in a child with refractory infantile spasms is itself a discriminating feature,
    and it is what makes this a functional rather than a malformation disorder.
  evidence:
  - reference: PMID:36968925
    reference_title: "Clinical Role of Codon 87 of the CYFIP2 Gene in Early Infantile Epileptic Encephalopathy: A Clinical Case Description."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: "Brain MRI showed no pathological changes"
    explanation: >-
      Records a normal MRI in a genotype-confirmed infant with severe early infantile
      epileptic encephalopathy.

prevalence:
- population: Worldwide, published cases
  measure_type: CASES_IN_LITERATURE
  prevalence_class: ULTRA_RARE
  notes: >-
    42 individuals with pathogenic or likely pathogenic CYFIP2 variants had been reported
    at the time of the most recent literature review. No population prevalence estimate
    exists; all knowledge derives from clinical sequencing referral cohorts rather than
    population-based ascertainment, so the true frequency is unknown and probably
    underestimated.
  evidence:
  - reference: DOI:10.1002/cns3.70036
    reference_title: "Expanding the Phenotype of <i>CYFIP2‐</i> Related Developmental Epileptic Encephalopathy: Case Report and Literature Review"
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: "We identified 41 additional patients with pathogenic or likely pathogenic"
    explanation: >-
      The literature review that establishes the cumulative case count, 41 additional
      patients alongside the index case reported in the same paper.
  - reference: PMID:33149277
    reference_title: "New insights into the clinical and molecular spectrum of the novel CYFIP2-related neurodevelopmental disorder and impairment of the WRC-mediated actin dynamics."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: "a total of only 18 individuals with CYFIP2 variants have been described to
      date"
    explanation: >-
      The earlier cumulative count, retained to show the trajectory of ascertainment - the
      reported cohort more than doubled between these two reviews.

genetic:
- name: CYFIP2
  gene_term:
    preferred_term: CYFIP2
    term:
      id: hgnc:13760
      label: CYFIP2
  relationship_type: CAUSATIVE
  association: >-
    Heterozygous de novo missense variants in CYFIP2 cause DEE65. Variants cluster at two
    hotspots, p.Arg87 and p.Asp724, and the hotspot identity predicts severity: Arg87
    substitutions give a consistently severe phenotype while Asp724 and other
    substitutions give a more variable one.
  evidence:
  - reference: PMID:33149277
    reference_title: "New insights into the clinical and molecular spectrum of the novel CYFIP2-related neurodevelopmental disorder and impairment of the WRC-mediated actin dynamics."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: "Genotype-phenotype correlation confirms a consistently severe phenotype in
      p.Arg87 patients but a more variable phenotype in p.Asp724 and other substitutions"
    explanation: >-
      States the hotspot-dependent severity difference that makes the genotype
      prognostically informative: Arg87 consistently severe, Asp724 variable.
  - reference: PMID:30664714
    reference_title: "Spatially clustering de novo variants in CYFIP2, encoding the cytoplasmic FMRP interacting protein 2, cause intellectual disability and seizures."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: "We here report 12 independent patients harboring a variety of de novo
      variants in CYFIP2 broadening the molecular and clinical spectrum of a novel
      CYFIP2-related neurodevelopmental disorder"
    explanation: >-
      Establishes the breadth of the variant spectrum beyond the original hotspot.
  - reference: PMID:33149277
    reference_title: "New insights into the clinical and molecular spectrum of the novel CYFIP2-related neurodevelopmental disorder and impairment of the WRC-mediated actin dynamics."
    supports: SUPPORT
    evidence_source: IN_VITRO
    snippet: "By structural modeling and investigation of WAVE-regulatory complex
      (WRC)-mediated actin polymerization in six patient fibroblast lines we assessed the
      impact of CYFIP2 variants on the WRC"
    explanation: >-
      Records that the functional assessment was done in six primary patient fibroblast
      lines, which is what moves the actin finding out of heterologous overexpression.
  - reference: DOI:10.1002/cns3.70036
    reference_title: "Expanding the Phenotype of <i>CYFIP2‐</i> Related Developmental Epileptic Encephalopathy: Case Report and Literature Review"
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: "The Arg87Cys variant was present in 12/42 (29%) patients"
    explanation: >-
      Quantifies the share of cases attributable to the single most common variant, which
      is what makes Arg87 a hotspot rather than merely a recurrent observation.
  notes: >-
    An in-frame deletion that removes codon 87 entirely, c.258_266del p.(Trp86_Ser88del),
    has been reported and produces the severe Arg87-hotspot phenotype - consistent with
    the interface account, since deleting the residue and substituting it have the same
    structural consequence. Rare truncating variants are reported with a milder phenotype
    and are of less certain significance; whether they act by haploinsufficiency is not
    settled, and this entry does not model them as a separate mechanism.
    The paralog CYFIP1 sits at the same cytogenetic band and lies within the 15q11.2
    microdeletion region, but causes a different disorder and is not to be conflated with
    CYFIP2.

diagnosis:
- name: Trio Exome or Genome Sequencing
  description: >-
    The disorder is defined genotypically. Because MRI is often normal and the clinical
    and EEG picture is shared with many other genetic developmental and epileptic
    encephalopathies, trio-based broad sequencing is what makes the diagnosis, and the
    trio design is what establishes de novo status.
  diagnosis_term:
    preferred_term: Whole Exome Sequencing
    term:
      id: NCIT:C101295
      label: Whole Exome Sequencing
  evidence:
  - reference: PMID:29534297
    reference_title: "De novo hotspot variants in CYFIP2 cause early-onset epileptic encephalopathy."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: "We performed trio-based whole-exome sequencing (WES) in 210 families and
      case-only WES in 489 individuals with epileptic encephalopathies"
    explanation: >-
      Describes the sequencing strategy through which the gene-disease relationship was
      established.

- name: Electroencephalography
  description: >-
    EEG is central to recognising the syndrome even though MRI is often normal. The
    interictal patterns reported span burst-suppression and hypsarrhythmia, mapping onto
    Ohtahara and West syndrome respectively - so a single genotype can present under more
    than one electroclinical syndrome label.
  diagnosis_term:
    preferred_term: Electroencephalography
    term:
      id: NCIT:C38054
      label: Electroencephalography
  evidence:
  - reference: PMID:36968925
    reference_title: "Clinical Role of Codon 87 of the CYFIP2 Gene in Early Infantile Epileptic Encephalopathy: A Clinical Case Description."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: "Initial interictal electroencephalograms showed a burst-suppression pattern
      and/or hypsarrhythmia, which were compatible with Ohtahara syndrome and West
      syndrome, respectively"
    explanation: >-
      Documents both EEG patterns and the two electroclinical syndromes they correspond
      to, in a genotype-confirmed patient.

treatments:
- name: Antiseizure Medication
  description: >-
    Seizures are characteristically drug-resistant and typically require multiple agents.
    No agent is established as preferred in this disorder.
  therapeutic_modality: SMALL_MOLECULE
  treatment_term:
    preferred_term: Pharmacotherapy
    term:
      id: NCIT:C15986
      label: Pharmacotherapy
  evidence:
  - reference: PMID:36968925
    reference_title: "Clinical Role of Codon 87 of the CYFIP2 Gene in Early Infantile Epileptic Encephalopathy: A Clinical Case Description."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: "When he started having infantile spasms, and there was evidence of small
      fragments of disorganized activity on EEG, vigabatrin was also started with
      transitory clinical improvement"
    explanation: >-
      Records vigabatrin producing only transitory improvement. Marked PARTIAL because it
      documents an attempted treatment and its limited effect, not an effective one.
  notes: >-
    No antiseizure agent has disease-specific efficacy evidence in CYFIP2-related DEE. The
    evidence item above is curated as a record of what was tried and how it failed, rather
    than as support for a recommendation. Vigabatrin, phenobarbital, valproic acid and a
    ketogenic diet were all used in the cited patient without achieving control.

animal_models:
- name: Cyfip2 p.Arg87Cys knock-in mouse
  species: Mouse
  genotype: Cyfip2+/R87C
  publication: PMID:36251395
  description: >-
    A genotype-matched knock-in carrying the human hotspot variant. It is unusually
    faithful for an epilepsy model, reproducing not just seizures but the specific
    spasm-like semiology, and it is the source of the protein-destabilization evidence.
  modeled_mechanisms:
  - target: Enhanced CYFIP2 Ubiquitination and Proteasomal Degradation
    relationship: RECAPITULATES
    fidelity: HIGH
    description: >-
      The model in which the degradation mechanism was demonstrated, in brain tissue.
    readouts:
    - name: Brain CYFIP2 protein abundance
      target: Enhanced CYFIP2 Ubiquitination and Proteasomal Degradation
      direction: DECREASED
      interpretation: >-
        Steady-state CYFIP2 is reduced in the brains of knock-in mice.
      evidence:
      - reference: PMID:36251395
        reference_title: "CYFIP2 p.Arg87Cys Causes Neurological Defects and Degradation of CYFIP2."
        supports: SUPPORT
        evidence_source: MODEL_ORGANISM
        snippet: "Beyond identifying a decrease in CYFIP2 protein levels in the
          Cyfip2+/R87C brains"
        explanation: >-
          The abundance measurement behind this readout.
    evidence:
    - reference: PMID:36251395
      reference_title: "CYFIP2 p.Arg87Cys Causes Neurological Defects and Degradation of CYFIP2."
      supports: SUPPORT
      evidence_source: MODEL_ORGANISM
      snippet: "we demonstrated that the p.Arg87Cys variant enhances ubiquitination and
        proteasomal degradation of CYFIP2"
      explanation: >-
        Supports the model as the system in which this mechanism was established.
  - target: Drug-Resistant Epilepsy
    relationship: PARTIALLY_RECAPITULATES
    fidelity: MODERATE
    limitations: >-
      The mouse reproduces the neonatal spasm phase, but then passes through a seizure-free
      interval before developing spontaneous recurrent seizures in adulthood. Human
      DEE65 is not described as having such a latent phase, so the model's temporal
      structure does not match the human course even though both endpoints are seizures.
    description: >-
      Reproduces spasm-like movements and, later, spontaneous seizures.
    readouts:
    - name: Seizure phenotype across the lifespan
      target: Drug-Resistant Epilepsy
      direction: ALTERED
      interpretation: >-
        Three distinct phases - neonatal spasms, a seizure-free interval, then adult
        spontaneous recurrent seizures ending in premature death.
      evidence:
      - reference: PMID:41066510
        reference_title: "Seizure evolution in a mouse model of West syndrome involves complex and time-dependent synapse remodeling, gliosis and alterations in lipid metabolism."
        supports: SUPPORT
        evidence_source: MODEL_ORGANISM
        snippet: "Our findings reveal that Cyfip2+/R87C mice, following a seizure-free
          period after neonatal spasms, develop spontaneous recurrent seizures in
          adulthood, ultimately leading to premature death"
        explanation: >-
          The longitudinal seizure phenotyping behind this readout.
    evidence:
    - reference: PMID:36251395
      reference_title: "CYFIP2 p.Arg87Cys Causes Neurological Defects and Degradation of CYFIP2."
      supports: SUPPORT
      evidence_source: MODEL_ORGANISM
      snippet: "The Cyfip2+/R87C mice recapitulated many neurological and neurobehavioral
        phenotypes of the patients, including spasmlike movements, microcephaly, and
        impaired social communication"
      explanation: >-
        Supports treating the model as informative for the clinical phenotype, including
        the specific spasm semiology.
- name: Xenopus laevis cyfip2 gain-of-function tadpole
  species: Xenopus laevis
  genotype: transient ectopic expression of Arg87Cys or Tyr108Cys cyfip2 mRNA
  publication: DOI:10.1101/2022.12.07.519540
  description: >-
    A transient overexpression model rather than a knock-in, chosen because gain-of-function
    variants cannot be modelled by the knockdown and knockout techniques used for loss of
    function. It is the only in vivo demonstration that a pathogenic CYFIP2 variant alone
    is sufficient to produce epileptic brain activity, which is why it matters to the
    canonical hypothesis.
  modeled_mechanisms:
  - target: Neuronal Network Hyperexcitability
    relationship: RECAPITULATES
    fidelity: MODERATE
    limitations: >-
      Transient ectopic mRNA overexpression rather than expression from the endogenous
      locus, so protein level is not physiological and the model cannot speak to the
      competing protein-destabilization account, which is a claim about steady-state
      abundance. Published as a preprint and not peer-reviewed at time of curation.
    description: >-
      Expression of either pathogenic variant produced spontaneous epileptic activity in
      the brain.
    readouts:
    - name: Spontaneous epileptic brain activity
      target: Neuronal Network Hyperexcitability
      direction: INCREASED
      interpretation: >-
        Both variants produced spontaneous epileptic activity, and it was reduced by
        valproate - a pharmacological control that argues the activity is genuinely
        seizure-like rather than an artefact of overexpression.
      evidence:
      - reference: DOI:10.1101/2022.12.07.519540
        reference_title: "Expression of mRNA encoding two gain-of-function <i>cyfip2</i> variants associated with DEE65 results in spontaneous seizures in <i>Xenopus laevis</i> tadpoles"
        supports: SUPPORT
        evidence_source: MODEL_ORGANISM
        snippet: "Expression of either pathogenic variant resulted in spontaneous epileptic
          activity in the brain"
        explanation: >-
          The electrophysiological measurement behind this readout.
      - reference: DOI:10.1101/2022.12.07.519540
        reference_title: "Expression of mRNA encoding two gain-of-function <i>cyfip2</i> variants associated with DEE65 results in spontaneous seizures in <i>Xenopus laevis</i> tadpoles"
        supports: SUPPORT
        evidence_source: MODEL_ORGANISM
        snippet: "For both variants, neuronal hyperactivity was reduced by treating the
          tadpole with 5 mM of the anti-seizure drug valproate (VPA)"
        explanation: >-
          Pharmacological rescue supporting the interpretation of the activity as
          seizure-like.
    evidence:
    - reference: DOI:10.1101/2022.12.07.519540
      reference_title: "Expression of mRNA encoding two gain-of-function <i>cyfip2</i> variants associated with DEE65 results in spontaneous seizures in <i>Xenopus laevis</i> tadpoles"
      supports: SUPPORT
      evidence_source: MODEL_ORGANISM
      snippet: "transient ectopic expression of the Arg87Cys pathogenic variant of"
      explanation: >-
        Establishes that the model expresses the human hotspot variant, which is what makes
        it informative for this disorder rather than for CYFIP2 biology generally.
  - target: Drug-Resistant Epilepsy
    relationship: PARTIALLY_RECAPITULATES
    fidelity: LOW
    limitations: >-
      Only one of the two variants tested produced the behavioural phenotype. Arg87Cys
      caused seizure-related darting and circling; Tyr108Cys did not alter behaviour at
      all, despite producing the same epileptic brain activity. So the model reproduces
      behavioural seizures for some pathogenic variants and not others, and a behavioural
      readout alone would have scored Tyr108Cys as unaffected.
      A second divergence matters specifically for anyone using this model as the
      pre-clinical drug-testing platform its authors propose: the tadpole hyperactivity was
      reduced by valproate, whereas the human phenotype curated in this entry is
      drug-resistant epilepsy in a patient who remained refractory despite valproic acid
      among four treatments. The model is valproate-responsive against a
      valproate-refractory human disorder, so a compound that suppresses activity here has
      not thereby been shown to work in patients.
    description: >-
      Seizure-related behaviour, variant-dependent.
    readouts:
    - name: Seizure-related swimming behaviour
      target: Drug-Resistant Epilepsy
      direction: INCREASED
      interpretation: >-
        Increased for Arg87Cys; explicitly unchanged for Tyr108Cys.
      evidence:
      - reference: DOI:10.1101/2022.12.07.519540
        reference_title: "Expression of mRNA encoding two gain-of-function <i>cyfip2</i> variants associated with DEE65 results in spontaneous seizures in <i>Xenopus laevis</i> tadpoles"
        supports: SUPPORT
        evidence_source: MODEL_ORGANISM
        snippet: "In contrast, expression of a second pathological variant, Tyr108Cys, did
          not alter tadpole behaviour"
        explanation: >-
          Curated as PARTIAL because it is a negative behavioural result for one variant
          alongside a positive one for the other. The dissociation is the informative part:
          brain activity and behaviour come apart within the same model.
    evidence:
    - reference: DOI:10.1101/2022.12.07.519540
      reference_title: "Expression of mRNA encoding two gain-of-function <i>cyfip2</i> variants associated with DEE65 results in spontaneous seizures in <i>Xenopus laevis</i> tadpoles"
      supports: SUPPORT
      evidence_source: MODEL_ORGANISM
      snippet: "resulted in increased seizure-related behaviours such as rapid darting and
        swimming in circles"
      explanation: >-
        Supports the behavioural phenotype for the Arg87Cys variant specifically.

discussions:

- discussion_id: controversy_gain_of_function_versus_protein_destabilization
  kind: CONTROVERSY
  status: OPEN
  attaches_to:
  - pathophysiology#Aberrant WAVE Regulatory Complex Activation
  - pathophysiology#Enhanced CYFIP2 Ubiquitination and Proteasomal Degradation
  prompt: >-
    Does p.Arg87Cys act by switching the WAVE regulatory complex on, or by removing CYFIP2
    protein from it?
  rationale: >-
    These are not two descriptions of one finding. The gain-of-function account says the
    complex is inappropriately active because the mutant subunit sequesters VCA more
    weakly; the destabilization account says there is measurably less CYFIP2 in the brain
    because the variant protein is ubiquitinated and degraded. A complex missing a
    required subunit and a complex that is constitutively switched on are different
    molecular states with opposite implications for therapy - one argues for inhibiting
    WAVE signalling, the other for restoring CYFIP2 abundance. Both rest on direct
    measurement: weaker VCA binding and increased actin accumulation in one case, reduced
    brain protein levels and demonstrated ubiquitination in the other, the latter in a
    genotype-matched knock-in rather than an overexpression system. A published exchange
    on exactly this point exists in the literature. The entry therefore draws both edges
    from the variant node in parallel, tagged to competing hypothesis groups, rather than
    placing either upstream of the other. Note also that the two accounts are not strictly
    exclusive - a destabilized interface could plausibly both loosen VCA binding and
    expose the protein to degradation - but no source demonstrates that reconciliation,
    and asserting it here would invent a synthesis rather than curate one.
  proposed_experiments:
  - experiment_id: exp_separate_abundance_from_activity
    name: Separate CYFIP2 abundance from complex activity
    description: >-
      Restore CYFIP2 abundance in Cyfip2+/R87C neurons by proteasome-independent means
      while leaving the variant sequence intact, and ask whether actin dysregulation and
      network hyperexcitability normalize. If they do, reduced abundance is the operative
      mechanism; if actin dysregulation persists at normal protein levels, the interface
      gain-of-function account is doing the work.

- discussion_id: mismatch_mouse_has_a_latent_phase_the_human_disease_does_not
  kind: HUMAN_MODEL_MISMATCH
  status: OPEN
  attaches_to:
  - pathophysiology#Neuronal Network Hyperexcitability
  prompt: >-
    Does the seizure-free interval in the knock-in mouse correspond to anything in the
    human disease, or is it a species artefact?
  rationale: >-
    The Cyfip2+/R87C mouse passes through three phases: neonatal spasms, then a
    seizure-free period, then spontaneous recurrent seizures in adulthood ending in
    premature death. Human DEE65 is described as having intractable seizures from onset,
    with no reported latent phase. Two readings are possible and they have opposite
    consequences. If the interval is a species artefact, then the model's adult seizures
    are not the same illness as the human one and the intervening period should not be
    used to time interventions. If instead the human trajectory has an under-recognized
    quiescent phase - plausible, since these children are heavily medicated and profoundly
    impaired, which makes a partial seizure remission easy to miss - then the model is
    revealing something about the human course rather than diverging from it. This is a
    tractable question because it can be settled from existing patient records rather than
    new experiments.
  proposed_experiments:
  - experiment_id: exp_longitudinal_seizure_burden_in_patients
    name: Longitudinal seizure burden in genotyped patients
    description: >-
      Assemble longitudinal seizure-frequency and EEG data across the lifespan for
      genotyped DEE65 individuals, specifically looking for a period of reduced seizure
      burden after the infantile spasm phase, and stratify by hotspot. A demonstrated
      quiescent interval would reclassify the mouse's latent phase from artefact to
      finding.

- discussion_id: gap_how_actin_dysregulation_becomes_hyperexcitability
  kind: KNOWLEDGE_GAP
  status: OPEN
  attaches_to:
  - pathophysiology#Disrupted Synaptic and Dendritic Spine Remodeling
  - pathophysiology#Neuronal Network Hyperexcitability
  prompt: >-
    By what route does dysregulated actin polymerization produce a hyperexcitable network?
  rationale: >-
    The molecular and cellular end of this disorder is well characterised, and the clinical
    end is unambiguous, but the step between them is the weakest link in the chain. Actin
    dysregulation is measured in fibroblasts and transfected cells; seizures are measured
    in patients and mice; what is not established is the mechanism that converts abnormal
    spine actin dynamics into synchronous network discharge. The entry marks both edges
    across this gap as indirect with unknown intermediates rather than asserting a
    pathway. This matters practically, because it is the step at which any therapy aimed
    at the actin machinery would have to act on the seizures.
  proposed_experiments:
  - experiment_id: exp_spine_dynamics_to_excitability_in_same_preparation
    name: Spine dynamics and excitability in the same preparation
    description: >-
      In Cyfip2+/R87C cortical or hippocampal slices, measure dendritic spine density and
      actin turnover alongside excitatory and inhibitory synaptic currents and network
      activity in the same neurons, at each of the three phases of the model's course.
      Establishing whether the excitatory-inhibitory balance shifts in step with the spine
      changes would convert the current inference into a measured link.

notes: >-
  Two items raised in review are deliberately not curated, with the checks recorded so a
  later curator does not repeat them. First, hyperkinetic or dyskinetic movements at 17%:
  this figure comes from the deep-research report citing Squire et al. 2025, but the
  retrievable abstract of that paper contains no mention of hyperkinetic, dyskinetic or
  choreiform movements at all, so the percentage cannot be quoted from its stated source.
  The paper is cited here for the frequency data it does contain. The second declined item has been reversed: the Xenopus
  tadpole gain-of-function model is now curated. An earlier revision of this entry
  declined it on the stated ground that it had no identifier resolving through the
  reference cache. That was wrong and was asserted without testing -
  DOI:10.1101/2022.12.07.519540 resolves and caches normally, as bioRxiv DOIs cited
  elsewhere in this knowledge base already do. The correct caveat is preprint status, which
  is recorded on the model's limitations rather than used as a reason to omit it.
  Note also that evidence items citing a DOI rather than a PMID are not snippet-checked by
  the validators, since DOI sits in skip_prefixes; the five DOI-sourced snippets in this
  entry were verified by hand against the cached record.
  Scope: this entry curates DEE65 only. CYFIP2 has a paralog, CYFIP1, at the same
  cytogenetic band; CYFIP1 lies within the 15q11.2 microdeletion region and is associated
  with a different neurodevelopmental phenotype. The NEC preflight for the deep-research
  report used to build this entry recorded CYFIP1 at 3 mentions against CYFIP2 at 75, so
  the report is about the intended entity, but the two should not be conflated in any
  future extension of this entry.
📚

References & Deep Research

Deep Research

1
Claude Code ▸
CYFIP2-Related Developmental and Epileptic Encephalopathy — Research Report
claude-haiku-4-5-20251001, claude-sonnet-5 21 citations 2026-08-20T03:56:56.347830

CYFIP2-Related Developmental and Epileptic Encephalopathy — Research Report

1. Disease Information

Overview. CYFIP2-related developmental and epileptic encephalopathy (DEE) is an autosomal dominant neurodevelopmental disorder caused by heterozygous, almost always de novo, missense (rarely in-frame deletion or splice-site) variants in CYFIP2 (Cytoplasmic FMR1-Interacting Protein 2), a core structural component of the WAVE regulatory complex (WRC) that governs Arp2/3-mediated actin polymerization. It presents with early-onset, typically drug-resistant seizures (often infantile spasms/West syndrome), severe to profound global developmental delay/intellectual disability, hypotonia, and — in a subset — hyperkinetic/dyskinetic movements and microcephaly (OMIM #618008; Nakashima et al. 2018, PMID:29534297).

Key identifiers: - OMIM disease: #618008 — Developmental and Epileptic Encephalopathy 65 (DEE65) - OMIM gene: 606323 — CYFIP2 (Cytoplasmic FMRP-Interacting Protein 2) - Gene location: 5q33.3 - HGNC: HGNC:13760 (CYFIP2) - MONDO: MONDO:0033374 (per Monarch/MalaCards cross-reference) - Orphanet (broader umbrella): ORPHA:442835 — Non-specific early-onset epileptic encephalopathy (CYFIP2 is listed as a causal gene under this umbrella; there is not yet a CYFIP2-specific ORPHA disease code distinct from this umbrella term as of current Orphanet indexing) - NCBI Gene: CYFIP2 (Gene ID 26999) - UniProt: Q96F07 - Synonyms for the disease: DEE65; CYFIP2-related neurodevelopmental disorder; CYFIP2 encephalopathy; (historically reported under) "early infantile epileptic encephalopathy" (EIEE)-spectrum, West syndrome with CYFIP2 mutation - Gene synonyms: PIR121; p140Sra-1 (protein); the gene family member CYFIP1 (5q33 as well, MIM 606322) is distinct and causes a different phenotype spectrum

Nature of the evidence base. Essentially all published information is aggregated case-series / cohort data from clinical exome and genome sequencing referral cohorts (not registry- or claims-based epidemiology), supplemented by mechanistic functional studies (patient fibroblasts, heterologous cell expression, Xenopus tadpole and mouse knock-in models). There is no large population-based EHR cohort for this ultra-rare disorder.

Sources: OMIM #618008, OMIM *606323, GeneCards CYFIP2, Orphanet CYFIP2

2. Etiology

Disease causal factor. DEE65 is a purely monogenic disorder — heterozygous, predominantly de novo pathogenic variation in CYFIP2 is both necessary and sufficient to cause disease in essentially all reported cases. There is no established environmental, infectious, or multifactorial contribution to primary causation.

Genetic risk factors: - Mutational hotspot at Arg87 (p.Arg87Cys/Leu/Pro): the dominant recurrent hotspot, identified in ~29–35% of published cases across cohorts (Nakashima et al. 2018, PMID:29534297; Zweier et al. 2019, PMID:30664714; Begemann/Zweier et al. 2021, PMID:33149277, Genet Med). All reported Arg87 substitutions produce a consistently severe DEE phenotype. - Second hotspot at Asp724: reported in 4/19 newly ascertained individuals in the 2021 cohort expansion, with a more variable/milder phenotype than Arg87 (PMID:33149277): "We report p.Asp724 as second mutational hotspot (4/19 cases). Genotype-phenotype correlation confirms a consistently severe phenotype in p.Arg87 patients but a more variable phenotype in p.Asp724 and other substitutions." - Other missense variants are distributed at additional interface residues within the CYFIP2–WAVE1/NCKAP1 tertiary structure — spatially clustered despite being scattered across the primary sequence (Zweier et al. 2019, PMID:30664714). - Putative loss-of-function variants (truncating/frameshift) are rarer, reported in a small number of individuals with a milder phenotype, and their pathogenicity/mechanism remains less certain (haploinsufficiency vs. incidental finding) (PMID:33149277). - Somatic/germline mosaicism has not been systematically characterized but is plausible given the de novo origin pattern typical of DEE genes; no confirmed parental mosaicism case has been prominently reported in the literature surveyed.

Environmental/lifestyle risk factors: None established. This is a primary monogenic channelopathy/cytoskeletal disorder with no known toxin, infectious, or lifestyle contributor.

Protective factors: None specifically documented for CYFIP2 variants. No modifier alleles have been described in the literature to date.

Gene-environment interactions: Not applicable/not reported; the disorder's severity appears driven by variant type/location (genotype) rather than environmental modulation.

Sources: PMID:29534297, PMID:30664714, PMID:33149277

3. Phenotypes

Core/near-universal features

Phenotype HPO suggestion Onset Frequency (reported cohorts)
Intellectual disability / global developmental delay HP:0001249 (Intellectual disability) / HP:0001263 (Global developmental delay) Apparent after seizure onset in most; normal development often reported before seizure onset ~100% (universal across cohorts)
Seizures, drug-resistant, various types HP:0001250 (Seizure) / HP:0031375 (Drug-resistant epilepsy) First months–years of life, commonly <6 months ~100%
Infantile spasms / West syndrome HP:0011097 (Epileptic spasm) Typically <12 months Reported in a substantial subset, especially Arg87 carriers
Muscular hypotonia (generalized/truncal) HP:0001252 (Hypotonia) Neonatal/infantile ~100% of the original 12-patient series had generalized/truncal hypotonia; 4/12 also had appendicular hypertonia (PMID:29534297)
Microcephaly HP:0000252 (Microcephaly) Postnatal, progressive in some 8/12 in the original hotspot cohort (PMID:29534297)
Mild facial dysmorphism HP:0001999 (Abnormal facial shape) Congenital Reported as "mild" in OMIM clinical description

Additional/variable features

  • Hyperkinetic/dyskinetic movements: present in a meaningful minority — one recent case-series/literature review reports hyperkinetic movements in 17% of pooled CYFIP2 patients (Squire et al. 2025, Ann Child Neurol Soc).
  • Spasm-like/paroxysmal dystonic movements, particularly around seizure onset (mirrored closely in the Cyfip2^R87C^ mouse model).
  • EEG abnormalities: hypsarrhythmia (in infantile-spasm presentations), multifocal epileptiform discharges.
  • Brain MRI: frequently normal or nonspecific despite severe clinical phenotype — noted explicitly in a codon-87 case report: "MRI findings showed no relevant changes despite severe clinical presentation" (PMC10038648).
  • Impaired social communication / autistic features: documented in patients and recapitulated in the Cyfip2^R87C^ mouse (impaired ultrasonic vocalization, social interaction deficits) (Kang et al. 2023, PMID:36251395).
  • Poor visual fixation/pursuit, hyperreflexia, feeding difficulties (swallowing/tongue protrusion) in severely affected infants (PMC10038648).
  • Developmental regression after an initial period of relatively preserved development in some patients.

Phenotype severity/progression pattern

  • Age of onset: neonatal to early infantile (majority <6 months); OMIM states onset "usually within the first months or years of life."
  • Severity: severe to profound in Arg87-hotspot carriers; more variable (mild–moderate to severe) in Asp724 and other missense/putative LOF carriers.
  • Course: progressive/static mixture — seizures often intractable from onset; the mouse model shows a triphasic course (neonatal spasms → seizure-free interval → adult-onset spontaneous recurrent seizures), which may parallel an underrecognized human trajectory (PMC12520403).

Quality of life impact

Severe global developmental impairment, drug-resistant epilepsy, and motor/communication deficits imply substantial lifelong caregiver burden and functional impairment; no disease-specific validated QOL instrument data were identified in the literature surveyed — QOL impact is inferred from the severe DEE clinical picture rather than directly measured (EQ-5D/SF-36/PedsQL data not located for this specific gene).

Sources: OMIM #618008, PMID:29534297, PMID:33149277, PMC10038648, Squire et al. 2025

4. Genetic/Molecular Information

Causal gene: CYFIP2 (HGNC:13760; OMIM *606323; NCBI Gene 26999; UniProt Q96F07), chromosome 5q33.3. Transcript reference commonly used in ClinVar: NM_001037333.3.

Pathogenic variant classes: - Missense, gain-of-function (GOF) — the dominant mechanism. Concentrated at two hotspots: - p.Arg87Cys/Leu/Pro (c.259C>T, c.260G>T, c.260G>C) — the most severe, most recurrent hotspot. - p.Asp724 substitutions — second hotspot, milder/variable phenotype. - Additional scattered missense variants (e.g., in the original 12-patient EJHG cohort: 7 missense + 1 splice-donor variant) that cluster spatially at the CYFIP2–WAVE1/NCKAP1 interface in 3D structure despite linear sequence dispersion (PMID:30664714). - In-frame deletion: e.g., c.258_266del; p.(Trp86_Ser88del), eliminating the Arg87 hotspot codon entirely (PMC10038648). - Putative loss-of-function (truncating) variants: rare, associated with a comparatively milder phenotype; pathogenic significance still described as of "unclear pathogenicity" in the largest cohort study (PMID:33149277).

Variant classification (ACMG/AMP): Missense hotspot variants (Arg87, Asp724) are classified Pathogenic/Likely Pathogenic in ClinVar (e.g., NM_001037333.3(CYFIP2):c.259C>T (p.Arg87Cys), ClinVar RCV000656389) based on de novo occurrence, absence from population databases, and concordant functional evidence.

Allele frequency: Pathogenic CYFIP2 variants are essentially absent from gnomAD/population databases — consistent with de novo occurrence and embryonic-lethal-adjacent severity of complete loss-of-function in model systems (see below).

Origin: Overwhelmingly de novo germline; no confirmed inherited transmission from an affected parent has been widely reported (consistent with reduced reproductive fitness in a severe pediatric-onset DEE).

Functional consequences — mechanistic characterization: - CYFIP2 is an obligate structural subunit of the WAVE regulatory complex (WRC) (CYFIP1/2–NCKAP1–WASF/WAVE–ABI–BRK1), which in the resting state sequesters the VCA domain of WASF to keep the complex autoinhibited. RAC1-GTP binding to CYFIP releases VCA, triggering Arp2/3-mediated actin polymerization at the membrane. - Disease variants — especially Arg87 — weaken the CYFIP2–WAVE1 interface interaction, releasing autoinhibition and driving constitutive/enhanced WRC activation — a gain-of-function mechanism at the level of actin regulation:

"The structural analysis showed this residue is positioned at the interface between CYFIP2 and WAVE1, and the variants may disrupt hydrogen bonding, leading to structural instability... Mutant CYFIP2 demonstrated weaker binding to the VCA domain and caused increased actin accumulation in transfected cells, suggesting gain-of-function effects on WAVE signaling." (Nakashima et al. 2018, PMID:29534297) - Patient fibroblast studies substantiate aberrant regulation of the actin cytoskeleton, confirming the cellular pathomechanism in primary patient-derived cells, not just heterologous overexpression systems (PMID:33149277). - A second, independent mechanism was identified for the Arg87Cys variant: it enhances ubiquitination and proteasomal degradation of CYFIP2 protein, reducing steady-state CYFIP2 levels in the brain — described in mouse and cellular models (Kang et al. 2023, PMID:36251395; note a published exchange debating "reduced CYFIP2 stability" vs. gain-of-function interpretation exists in the literature — ResearchGate summary of the reply). A review notes: "Almost all CYFIP2-derived mutations (7 out of 8) promoted WRC activation, but to variable extent and with at least two independent mechanisms" — i.e., both interface-destabilization-driven WRC hyperactivation and reduced-protein-stability/increased-turnover mechanisms are represented across the mutational spectrum. - Not classical loss-of-function haploinsufficiency for the hotspot variants — this is an important curatorial distinction from most other severe DEE genes, since the dominant disease mechanism is toxic gain-of-function/dysregulation rather than simple dosage reduction (contrast with the rarer, milder truncating-variant subset, which may act via haploinsufficiency).

Modifier genes: None specifically established.

Epigenetics: No CYFIP2-specific DNA methylation/chromatin signature has been reported in the literature surveyed. A related RNA-level regulatory mechanism (A-to-I RNA editing of CYFIP2 transcripts affecting actin regulation, axon growth, and spinogenesis) has been described as a separate area of investigation, distinct from the DEE-causing coding variants (bioRxiv, RNA editing paper) — relevant to normal CYFIP2 biology rather than to DEE65 pathogenic variants specifically.

Chromosomal abnormalities: DEE65 is caused by single-nucleotide/small-indel variants, not large chromosomal rearrangements; no recurrent CNV mechanism has been described for this locus in the DEE65 literature (contrast: CYFIP1 sits within the 15q11.2 microdeletion region relevant to a different neurodevelopmental syndrome — not to be conflated with CYFIP2/5q33.3).

Ontology suggestions: HGNC:13760 (CYFIP2); GO:0071203 (WASP-family verprolin-homologous protein regulatory complex localization) / GO:0034314 (Arp2/3 complex-mediated actin nucleation) / GO:0030041 (actin filament polymerization); CHEBI not directly applicable to the protein itself.

Sources: OMIM *606323, PMID:29534297, PMID:30664714, PMID:33149277, PMID:36251395, ClinVar RCV000656389

5. Environmental Information

No environmental toxin, occupational exposure, radiation, or pollutant exposure has been implicated in CYFIP2-related DEE causation — it is a purely monogenic disorder. No lifestyle factors (maternal or infant) have been linked to onset or severity. No infectious trigger or agent is described in the literature; seizure precipitants in already-affected individuals (fever, illness) are not specifically documented for this gene in the sources reviewed (contrast with genes like SCN1A/CACNA1C where fever-triggered exacerbation is well characterized — no equivalent CYFIP2-specific finding was located).

6. Mechanism / Pathophysiology

Causal chain (upstream → downstream):

  1. Molecular trigger: De novo heterozygous missense variant in CYFIP2 (most commonly Arg87Cys/Leu/Pro) destabilizes the CYFIP2–WAVE1 (WASF1) interface within the pentameric WAVE Regulatory Complex, OR (Arg87Cys specifically) promotes CYFIP2 ubiquitination/proteasomal degradation.
  2. Molecular consequence: Loss of normal autoinhibitory sequestration of the WASF VCA domain → constitutive/enhanced release of VCA → aberrant, excessive activation of the Arp2/3 complex.
  3. Cellular consequence: Dysregulated actin cytoskeleton dynamics — increased/aberrant F-actin polymerization demonstrated directly in transfected cells and in patient-derived fibroblasts (PMID:29534297; PMID:33149277). In neurons, CYFIP2/WRC signaling is essential for normal dendritic spine morphogenesis, synaptic actin remodeling, axon growth, and BDNF-TrkB (NTRK2) endosomal trafficking/signaling (per GeneCards functional annotation).
  4. Circuit-level consequence: Disrupted synaptic actin dynamics and spine architecture are hypothesized to produce neuronal hyperexcitability and aberrant network connectivity, providing a plausible mechanistic bridge to the epilepsy phenotype; this is directly modeled functionally in the Xenopus tadpole system, where transient expression of GOF cyfip2 mRNA (encoding pathogenic variants) is sufficient to cause spontaneous epileptiform brain activity and seizure-associated behaviors (rapid darting, circling swimming) — a direct causal (not merely correlative) demonstration that the GOF variant alone drives network hyperexcitability in vivo (bioRxiv, DEE65 Xenopus model).
  5. Organismal/clinical consequence: Seizures (often infantile spasms/West syndrome pattern), profound developmental delay, hypotonia progressing in some to hyperkinetic movement abnormalities, and microcephaly — the last plausibly reflecting a role for CYFIP2/actin dynamics in neural progenitor proliferation and cortical growth (supported by the Cyfip2-null mouse embryonic phenotype below).

Cell types and biological processes involved: - Cortical/hippocampal excitatory neurons — actin-dependent dendritic spine and synapse remodeling (documented cytoarchitectural disorganization and gliosis in hippocampus of the Cyfip2^R87C^ mouse — PMID:36251395). - Glial cells (astrocytes/microglia): sequential/temporally staged glial activation (gliosis) accompanies seizure evolution in the mouse model, with lipid droplet accumulation in astrocytes and broad proteomic/lipidomic remodeling over the disease course (PMC12520403). - Neural progenitor cells / developing cortex: Cyfip2-null mouse embryos show smaller body size, early postnatal lethality, and cortical extracellular-matrix (ECM)-related gene expression changes, though gross E18.5 brain size/cytoarchitecture were comparable to wild-type at that stage — implicating CYFIP2 in ECM-dependent aspects of cortical development (PMC6338024). - Non-neuronal role (secondary): CYFIP2/WRC has broader roles in membraneless organelle (stress granule, P-body, nucleolus) regulation and eIF2α phosphorylation/translational control, identified via CYFIP2-knockout forebrain proteomics — a newly characterized, actin-cytoskeleton-linked mechanism for translational regulation whose disease relevance is still emerging (Cho et al. 2024, PMID:38981622).

Molecular pathways: WAVE regulatory complex / Arp2/3 actin nucleation pathway (Reactome: "RHO GTPase Effectors"; KEGG: regulation of actin cytoskeleton); RAC1-GTPase signaling upstream of WRC activation.

Protein dysfunction category: Predominantly toxic gain-of-function / dysregulated activity (constitutive WRC hyperactivation) rather than simple loss-of-function, distinguishing DEE65 mechanistically from many other DEE genes that act via straightforward haploinsufficiency; a minority of putative truncating variants may act via classical loss-of-function/haploinsufficiency with a correspondingly milder phenotype.

Suggested ontology terms: - GO (biological process): GO:0030036 (actin cytoskeleton organization), GO:0034314 (Arp2/3 complex-mediated actin nucleation), GO:0051893 (regulation of focal adhesion assembly), GO:0007015 (actin filament organization) - GO (molecular function): GO:0003779 (actin binding) - GO (cellular component): GO:0071203 (WASP-family verprolin-homologous protein regulatory complex), GO:0030426 (growth cone), GO:0043197 (dendritic spine) - CL (cell types): CL:0000679 (glutamatergic neuron) / CL:0000128 (oligodendrocyte, if relevant to white matter findings) / CL:0000127 (astrocyte)

Sources: PMID:29534297, PMID:33149277, PMID:36251395, PMC12520403, PMC6338024, PMID:38981622, bioRxiv Xenopus model

7. Anatomical Structures Affected

Organ/system level: - Primary: Central nervous system (brain) — cerebral cortex, hippocampus. - Secondary: Musculoskeletal system (hypotonia-related), craniofacial structures (mild dysmorphism, microcephaly reflecting reduced brain/skull growth). - Body systems involved: Nervous system (primary); musculoskeletal (hypotonia); growth/development (microcephaly, failure to thrive in severe cases).

Tissue/cell level: - Cerebral cortical neurons and hippocampal pyramidal neurons (cytoarchitectural disorganization documented in mouse model). - Glial populations (astrocytes, microglia) showing reactive gliosis over disease course. - Peripheral patient fibroblasts (used as an accessible surrogate tissue for demonstrating actin dysregulation in functional studies — not itself clinically affected, but mechanistically informative).

Subcellular level: - Actin cytoskeleton / cortical actin network (GO:0015629 actin cytoskeleton). - Dendritic spines / postsynaptic density (site of WRC-dependent spine actin remodeling). - Growth cone (axon guidance/pathfinding, WRC-dependent). - Proteasome (site of enhanced Arg87Cys protein degradation).

Localization / laterality: Diffuse, bilateral cerebral involvement (no lateralization reported); MRI is frequently structurally unremarkable despite severe functional/electrical involvement.

UBERON suggestions: UBERON:0000955 (brain), UBERON:0001954 (Ammon's horn/hippocampus), UBERON:0001851 (cortex).

Sources: OMIM #618008, PMID:36251395, PMC10038648

8. Temporal Development

  • Onset: Congenital predisposition with clinical onset in the neonatal-to-early-infantile period; seizures most commonly begin within the first 6 months of life (OMIM), sometimes as early as the neonatal period. Development is often described as apparently normal before seizure onset in milder cases, with delay/regression becoming evident afterward.
  • Onset pattern: Acute/subacute onset of seizures, often abrupt (spasms clusters), against a background of insidious developmental impairment.
  • Progression: Predominantly static-to-progressive — seizures are frequently intractable from onset (refractory to multiple antiseizure medications, as documented in case reports trying levetiracetam, phenobarbital, vigabatrin, zonisamide, valproic acid, and ketogenic diet without seizure control — PMC10038648). The mouse model reveals a more complex, triphasic natural history: neonatal spasm-like events → a seizure-free interval → adult-onset spontaneous recurrent seizures with progressive synaptic remodeling, gliosis, and premature death (PMC12520403) — raising the possibility of an analogous, currently underrecognized, evolving human natural history.
  • Disease duration: Chronic, lifelong; DEE65 is not known to remit.
  • Remission: No spontaneous remission pattern documented; some symptomatic improvement possible with combination antiseizure regimens in individual cases, but sustained pharmacoresistance is the norm.
  • Critical periods: Early infancy appears to be a critical window for both seizure emergence and the steepest developmental impact, paralleling the well-established DEE/West-syndrome paradigm of early intervention urgency, though CYFIP2-specific intervention-timing data are not established.

Sources: OMIM #618008, PMC10038648, PMC12520403

9. Inheritance and Population

Epidemiology: DEE65 is an ultra-rare disorder. As of the original 2018 discovery, only 4 individuals were reported; by the 2019 EJHG cohort, 12 patients; by the 2021 Genetics in Medicine cohort expansion, a cumulative total of 37 individuals (19 newly reported + 18 previously published) (PMID:33149277); a 2024–2025 systematic literature review/case report identified substantially more — ~41 additional patients, with the Arg87Cys variant present in 12/42 (29%) of one pooled analysis, and at least 27 additional distinct variants described across roughly 30 more patients (Squire et al. 2025). No formal population-based incidence/prevalence rate (cases per 100,000) has been established — case counts reflect literature ascertainment via clinical exome/genome sequencing referral, not population screening.

Inheritance pattern: Autosomal dominant, virtually always due to a de novo variant.

Penetrance: Appears to be complete for the well-characterized hotspot variants (Arg87), based on the absence of reported unaffected carrier parents; penetrance/expressivity for milder putative loss-of-function variants is less certain.

Expressivity: Variable, correlated with variant identity — Arg87 substitutions show consistently severe expressivity; Asp724 and other missense variants show more heterogeneous, sometimes milder expressivity; putative LOF variants associate with the mildest reported phenotypes.

Genetic anticipation: Not applicable — this is not a repeat-expansion disorder.

Germline mosaicism: Not specifically documented as a recurrence mechanism in this gene's literature to date, though theoretically possible for any de novo dominant disorder (genetic counseling should still consider low recurrence risk from potential parental gonadal mosaicism, per general DEE counseling practice).

Founder effects: None reported — cases described in the literature span diverse ethnic/geographic backgrounds (original hotspot cohort included Japanese, Israeli, and Malaysian families; subsequent cohort recruited from Germany, France, UK, and elsewhere), consistent with recurrent independent de novo mutation at a true mutational hotspot rather than a single ancestral founder allele.

Consanguinity: Not a relevant risk factor given the dominant de novo mechanism.

Carrier frequency: Not applicable (dominant de novo disorder, not a recessive carrier state); population database allele frequency for pathogenic variants is essentially zero.

Sex ratio: No skewed sex ratio has been specifically reported; both male and female patients are represented in all major cohorts, and the Cyfip2^R87C^ mouse model shows phenotypes in both sexes.

Geographic distribution: No endemic or regionally restricted distribution; cases have been reported globally (East Asia, Middle East, Southeast Asia, Western Europe).

Sources: PMID:29534297, PMID:30664714, PMID:33149277, Squire et al. 2025

10. Diagnostics

Clinical/laboratory tests: No CYFIP2-specific biomarker or biochemical assay exists; routine metabolic/biochemical workup in suspected DEE is typically used to exclude alternative etiologies rather than to positively diagnose CYFIP2-DEE.

Electrophysiology: EEG is central to diagnosis and phenotyping — documenting hypsarrhythmia (in spasms presentations), multifocal epileptiform discharges, and background abnormalities consistent with an encephalopathic process. LOINC-coded EEG procedures apply generically (no CYFIP2-specific EEG signature established).

Imaging: Brain MRI is typically performed to exclude structural lesions; findings are frequently normal or nonspecific even in severely affected patients (PMC10038648) — an important diagnostic point (a normal MRI does not exclude CYFIP2-DEE).

Genetic testing (primary diagnostic modality): - Whole exome sequencing (WES) or whole genome sequencing (WGS), typically as trio (proband + parents) analysis, is the standard diagnostic approach, since CYFIP2-DEE was itself discovered via de novo variant discovery in WES cohorts of unexplained early-onset epileptic encephalopathy (PMID:29534297; PMID:30664714). - Gene panels for epileptic encephalopathy / early-infantile epilepsy (e.g., commercial DEE panels, the Genomics England PanelApp "Epileptic encephalopathy" panel, which lists CYFIP2 — PanelApp CYFIP2) include CYFIP2 and are a reasonable first-tier test given the recognizable phenotype. - Single-gene (Sanger) testing is appropriate for targeted confirmation once a candidate variant is identified by broader sequencing, or for recurrence-risk testing in future pregnancies (to exclude/confirm rare parental mosaicism). - Chromosomal microarray (CMA) and karyotyping are not primary diagnostic tools for this disorder (single-nucleotide variant mechanism, not CNV-based), though they are often performed as part of standard first-tier DEE/ID workup to exclude other etiologies. - Mitochondrial DNA testing / repeat-expansion testing: not relevant to this specific gene.

Clinical diagnostic criteria: No formal consensus clinical diagnostic criteria specific to CYFIP2-DEE exist; diagnosis follows the general DEE/ILAE framework (early-onset seizures + developmental impairment disproportionate to seizures alone, when applicable) combined with molecular confirmation, since the phenotype alone is not sufficiently distinctive to establish diagnosis without genetic testing.

Differential diagnosis: Other genetic DEEs presenting with infantile spasms/West syndrome and severe ID (e.g., STXBP1, KCNQ2, SCN2A, CDKL5, ARX [causing OMIM DEE1, a distinct entity from CYFIP2's DEE65 despite similar naming], and other WAVE-complex-adjacent genes such as NCKAP1, ACTB/ACTG1) should be considered and are typically distinguished only by molecular testing given phenotypic overlap.

Screening: No population-based newborn or carrier screening applies (ultra-rare, de novo dominant disorder); prenatal testing is possible in a family with a previously identified pathogenic variant (targeted testing for future pregnancies), though the near-universal de novo mechanism limits its practical yield outside of confirmed parental mosaicism.

Sources: PMID:29534297, PMC10038648, PanelApp CYFIP2

11. Outcome/Prognosis

Survival/mortality: No formal survival statistics (5-/10-year survival) have been published for the human disorder; the condition is not classically described as directly life-limiting in early reports, though severe DEEs of this type carry recognized SUDEP (sudden unexpected death in epilepsy) risk in general, and the mouse model shows premature death associated with progressive adult-onset seizures and neuroinflammation — a signal that warrants clinical vigilance even though it has not yet been formally quantified in human cohorts (PMC12520403).

Morbidity/function: Severe to profound intellectual disability and motor impairment are the norm in Arg87-hotspot carriers; more variable functional outcomes (mild–moderate ID) occur with Asp724/other missense or putative LOF variants. No standardized disability outcome measure (e.g., ICF-based) has been systematically applied across cohorts in the literature reviewed.

Complications: Drug-resistant epilepsy with attendant risks (injury, aspiration in severe cases, medication side effects from polytherapy); failure to thrive/growth impairment in the most severely affected; secondary orthopedic complications from chronic hypotonia/hypertonia are plausible but not specifically quantified in the sources reviewed.

Recovery potential: Given the drug-resistant nature of seizures across nearly all reported cases (multiple standard and dietary therapies failing in individual case reports), and the fixed neurodevelopmental impairment pattern, recovery to a mild phenotype is not typical for hotspot (Arg87) variant carriers; the milder end of the spectrum (Asp724, other missense, putative LOF variants) offers somewhat better functional prognosis.

Prognostic factors: Variant identity is the strongest known prognostic factor — Arg87 substitutions predict a consistently severe DEE course; Asp724 and other substitutions predict a more variable course; putative loss-of-function variants predict the mildest reported phenotypes (PMID:33149277). No independent biomarker-based prognostic tool exists.

Sources: PMID:33149277, PMC10038648, PMC12520403

12. Treatment

Pharmacotherapy (symptomatic/standard-of-care only — no CYFIP2-targeted approved drug exists): - Standard antiseizure medications have been trialed empirically and are frequently ineffective individually: levetiracetam, phenobarbital, vigabatrin (a first-line agent for infantile spasms generally), zonisamide, valproic acid — documented as failing to control seizures in a codon-87 case report despite combination use (PMC10038648). ACTH/corticosteroid therapy (standard for infantile spasms broadly) is presumably trialed in spasms-presenting cases per general West syndrome practice, though CYFIP2-specific response data were not located in the sources reviewed. - Ketogenic diet has been tried as an adjunct in at least one reported case without achieving seizure freedom (PMC10038648). - NCIT suggestion: NCIT:C15986 (Pharmacotherapy) as the generic treatment_term, with specific agents captured via therapeutic_agent (e.g., CHEBI terms for levetiracetam, valproic acid, vigabatrin, phenobarbital, zonisamide). - Dietary intervention: NCIT:C15447 (Dietary Intervention) for the ketogenic diet.

Precision/targeted therapy: No CYFIP2-specific precision therapeutic has been developed or trialed to date. Because the dominant mechanism is toxic gain-of-function WRC hyperactivation (analogous conceptually to other GOF-driven DEEs where targeted pharmacology has succeeded, e.g., quinidine for KCNT1-GOF epilepsy, memantine for GRIN2D-GOF epilepsy), CYFIP2-DEE is a plausible future candidate for genotype-informed precision approaches (e.g., agents modulating Rac1/WRC/Arp2/3 signaling), but this remains speculative/preclinical — no clinical trial or case report of a targeted small molecule was identified. A general review of precision epilepsy therapeutics frames this class of gene (defined mechanism, amenable pathway) as a priority area for future targeted drug development (Byrne et al. 2021, PMID:34089185).

Gene/RNA-based therapy: No CYFIP2-specific gene therapy, ASO, or siRNA approach has reached even preclinical proof-of-concept publication in the literature surveyed, though the Xenopus GOF model system provides a tractable platform on which future suppression/knockdown strategies (e.g., allele-specific ASO knockdown of the mutant transcript, analogous to strategies used for other dominant GOF DEE genes) could in principle be tested.

Surgical/interventional: Not applicable as a primary treatment (no known epileptogenic focal lesion amenable to resection; this is a diffuse genetic encephalopathy).

Supportive/rehabilitative care: Physical therapy, occupational therapy, and speech/communication therapy are standard supportive measures for the hypotonia, motor impairment, and communication deficits characteristic of severe DEE, per general DEE management practice (NCIT:C15302 Physical Therapy; NCIT:C15315 Rehabilitation).

Experimental/clinical trials: No CYFIP2-DEE-specific registered clinical trial was identified via the sources reviewed (searches did not surface an NCT-registered CYFIP2-targeted trial).

Treatment outcomes: As above, existing case reports document refractory epilepsy despite multi-drug and dietary intervention — the treatment picture at present is one of largely unmet need for a mechanism-based therapy.

Sources: PMC10038648, Byrne et al. 2021, PMID:34089185

13. Prevention

No primary, secondary, or tertiary prevention strategy exists for CYFIP2-DEE, given its near-universal de novo origin (not preventable through risk-factor modification, vaccination, or population screening). The applicable preventive interventions are:

  • Genetic counseling: for parents of an affected child, to communicate the generally low (but not zero, due to possible germline mosaicism) empiric recurrence risk for future pregnancies, and to discuss options such as targeted prenatal or preimplantation genetic testing once the familial variant is known (NCIT:C15240 Genetic Counseling).
  • Prenatal diagnosis: feasible only in a family with an already-identified pathogenic variant (e.g., via amniocentesis/CVS with targeted variant testing), not as population-level screening.
  • Early intervention services: while not preventing the underlying disorder, early developmental/rehabilitative intervention following diagnosis is standard practice to optimize functional outcomes in severe pediatric DEEs generally.

No public-health-level, environmental, or behavioral prevention measures apply, consistent with the disorder's purely monogenic de novo etiology.

14. Other Species / Natural Disease

Taxonomy of models used: Mus musculus (NCBITaxon:10090); Xenopus laevis (NCBITaxon:8355).

Orthologous gene: Mouse Cyfip2 (MGI ortholog of human CYFIP2); highly conserved across vertebrates given the essential, ancient function of the WAVE regulatory complex in actin regulation.

Naturally occurring disease in other species: No naturally occurring (spontaneous) CYFIP2-associated disease has been reported in companion animals or wildlife in the sources reviewed (no OMIA entry surfaced) — all non-human disease models identified are engineered (knock-in/knockout or induced mRNA overexpression), not naturally arising veterinary conditions.

Comparative biology: The WAVE regulatory complex and its Arp2/3-actin-nucleation function are deeply conserved from Xenopus through mammals, which is precisely why the transient mRNA-overexpression Xenopus tadpole system and the mouse knock-in system both successfully recapitulate core aspects of the human phenotype (seizures, developmental/behavioral abnormalities) — supporting strong evolutionary conservation of the underlying disease mechanism.

Zoonotic/transmission relevance: Not applicable — this is a non-communicable monogenic disorder.

Sources: Kang et al. 2023, PMID:36251395, bioRxiv Xenopus model

15. Model Organisms

Mouse models

  1. Cyfip2^+/R87C^ knock-in mouse (heterozygous, patient-matched hotspot variant) — the flagship disease model.
  2. Recapitulation: "Cyfip2+/R87C mice recapitulated many neurological and neurobehavioral phenotypes of the patients, including spasmlike movements, microcephaly, and impaired social communication." Specific findings: neonatal spasm-like behavior, developmental regression, muscular hypotonia, microcephaly in early life; adult mice show hyperlocomotion, impaired social interaction, and defective ultrasonic vocalization (Kang et al. 2023, PMID:36251395).
  3. Molecular mechanism confirmed in vivo: enhanced ubiquitination/proteasomal degradation of CYFIP2 protein, reducing brain CYFIP2 levels.
  4. Neuropathology: age-progressive hippocampal cytoarchitectural disorganization and gliosis.
  5. Longitudinal natural history (follow-on study): triphasic seizure evolution (neonatal spasms → seizure-free interval → adult spontaneous recurrent seizures → premature death), with time-dependent synapse remodeling, sequential glial activation, astrocytic lipid droplet accumulation, and significant proteomic/lipidomic brain changes across the disease course (PMC12520403; related preprint: bioRxiv longitudinal deep phenotyping).

  6. Fidelity: High for spasm-like movements, microcephaly, and social/behavioral deficits; the mouse additionally reveals an adult-onset seizure phase not yet systematically documented in humans, raising a HUMAN_MODEL_MISMATCH-type open question about whether the human natural history includes an analogous later-life seizure evolution that is currently underrecognized due to limited longitudinal follow-up of patients.

  7. Cyfip2^-/-^ (null/knockout) mouse — models complete loss of function (distinct from the disease-causing GOF hotspot alleles; useful for understanding baseline CYFIP2 developmental biology).

  8. Phenotype: Smaller embryonic body size; embryos survive to E18.5 with grossly comparable brain size/cortical cytoarchitecture to wild-type/heterozygous littermates at that stage, but all Cyfip2−/− pups are found dead at postnatal day 0 (early postnatal, not embryonic, lethality). Cortical extracellular-matrix (ECM)-related gene expression changes are identified in the null embryonic cortex (PMC6338024).
  9. Relevance: Establishes that complete biallelic loss of CYFIP2 is not viable postnatally, reinforcing that the milder human phenotypes associated with putative (heterozygous) loss-of-function variants reflect partial haploinsufficiency rather than complete pathway loss, and underscoring that the disease-causing mechanism for the severe human phenotype (Arg87 etc.) is mechanistically distinct from simple loss-of-function.

Non-mammalian model

  1. Xenopus laevis tadpole, transient mRNA overexpression of GOF cyfip2 variants — a rapid, tractable in vivo system.
  2. Approach: Ectopic expression of pathogenic-variant-encoding mRNA (including an Arg87Cys-equivalent variant) directly in tadpoles.
  3. Findings: Sufficient on its own to cause spontaneous epileptiform brain electrical activity and seizure-associated behaviors (rapid darting, circular swimming, increased agitation) — providing strong causal (not just correlative) evidence that the GOF mutant protein, expressed transiently and ectopically, is sufficient to generate a seizure phenotype (bioRxiv, DEE65 Xenopus model).
  4. Fidelity/limitations: High construct-validity for demonstrating GOF sufficiency and rapid electrophysiological/behavioral seizure readouts; lower face-validity for modeling the full chronic developmental/cognitive phenotype given the amphibian system and transient (non-genomic, mosaic) expression paradigm — a HUMAN_MODEL_MISMATCH consideration for any curated entry using this model.

Cellular/in vitro models

  1. Patient-derived dermal fibroblasts — used to demonstrate aberrant actin cytoskeleton regulation directly in human cells carrying the endogenous heterozygous variant, substantiating that the WRC-dysregulation mechanism operates in patient tissue, not only in heterologous overexpression systems (PMID:33149277).
  2. Heterologous transfected cell lines (e.g., COS/HEK-type systems) — used in the original discovery study to demonstrate weakened CYFIP2–VCA/WAVE1 binding and increased aberrant F-actin accumulation for the Arg87 hotspot variants (PMID:29534297).
  3. iPSC-derived human neuronal model (R87C) — a more recent addition assessing the R87C variant's impact in a human neuronal context in vitro (Impact of the CYFIP2 R87C variant in a human neuronal model in vitro, Sci Rep) — relevant for MorPhiC-style cellular phenotype curation if applicable.

Model databases: MGI (mouse Cyfip2 allele records), Xenbase (for Xenopus Cyfip2), IMPC (if a systematic Cyfip2 knockout line has been phenotyped — not confirmed in sources reviewed).

Sources: PMID:36251395, PMC12520403, PMC6338024, bioRxiv Xenopus model, PMID:33149277, PMID:29534297

Summary of Key Primary Literature for Curation

PMID/ID First author, Year, Journal Contribution
29534297 Nakashima et al., 2018, Ann Neurol First description of CYFIP2 Arg87 de novo hotspot in 4 patients; establishes GOF mechanism
30664714 Zweier et al., 2019, Eur J Hum Genet 12-patient cohort; spatial clustering of variants at WRC interface
33149277 Begemann/Zweier et al., 2021, Genet Med 37-patient cumulative cohort; identifies Asp724 second hotspot; fibroblast functional data; genotype-phenotype correlation; LOF-variant subset
36251395 Kang et al., 2023, Ann Neurol Cyfip2+/R87C mouse knock-in model; degradation mechanism
PMC12520403 Guo et al. (longitudinal follow-up) Mouse model natural history: triphasic seizure evolution, gliosis, lipidomics
PMC6338024 2018, Front Mol Neurosci Cyfip2-null mouse: postnatal lethality, ECM gene expression
bioRxiv 2022.12.07.519540 — Xenopus GOF model: sufficiency for spontaneous seizures
38981622 Cho et al., 2024 CYFIP2 role in membraneless organelles/eIF2α regulation
PMC10038648 Cureus case report Codon-87 in-frame deletion case; treatment-refractory course; normal MRI
Squire et al. 2025, Ann Child Neurol Soc — Most recent systematic literature review expanding phenotype spectrum
34089185 Byrne et al., 2021, Dev Med Child Neurol General precision-therapy framework relevant to future CYFIP2-targeted treatment

Note on evidence-quality caveats for curation: Several full-text sources (OMIM clinical synopsis, Nature/Genetics in Medicine, and one bioRxiv PDF) returned HTTP 403/429/cookie-wall errors during this research session and could not be directly fetched verbatim; the corresponding content above was reconstructed from search-engine-indexed summaries and abstract text retrieved via PubMed/NCBI eutils, cross-checked across multiple independent sources where possible. Before using any PMID/snippet pair from this report in a dismech KB entry, run just fetch-reference and just count-verified-snippets per the standard SOP to confirm exact-quote fidelity against the authoritative cached abstract, since some quoted fragments above were extracted via an intermediary summarization step rather than direct verbatim page retrieval.

Reference Validation

Checked with linkml-reference-validator 0.2.1.

Outcome Count
References checked 10
Resolved 10
Unresolved (possible confabulation) 0
Unverifiable 0
References weighed for topical relevance 10
On topic 9
Off topic 0

All extracted references resolved successfully.