NSF gene variants cause developmental and epileptic encephalopathy 96: expanding genotype and phenotypic spectrum with prenatal-onset features

Abstract

Background

Developmental and epileptic encephalopathy 96 (DEE96, OMIM 619340) is a rare autosomal dominant disorder caused by heterozygous variants in the NSF gene, encoding a key AAA + ATPase involved in SNARE-mediated membrane fusion. To date, only four pathogenic variants have been reported.

Methods

Trio whole-exome-sequencing and Sanger validation were performed in a fetus presenting with multiple prenatal anomalies. Bioinformatic analyses and literature review were conducted to characterize genotype-phenotype correlations.

Results

We identified a novel de novo heterozygous missense variant [c.1055 A>G; p (Asn352Ser)] in the D1 domain of NSF, classified as likely pathogenic per ACMG/AMP criteria. Prenatal ultrasound revealed increased nuchal fold thickness, left clubfoot, severe anemia with cardiac enlargement, and hepatosplenomegaly. Increased nuchal fold thickness and left clubfoot had not been previously documented in DEE96 cases. This represents the second prenatal diagnosis and fifth case overall.

Conclusion

Our findings expand the genotypic and phenotypic spectrum of DEE96, demonstrating that DEE96 manifests as a multisystem disorder with prenatal onset. The high incidence of severe anemia suggests hematological involvement may be a characteristic feature. These observations support considering NSF variants in the differential diagnosis of complex fetal syndromes with neurological and hematological abnormalities.

Background

Next-generation sequencing (NGS) has been instrumental in the diagnosis of rare genetic diseases, by virtue of its capacity to facilitate high-throughput analysis of genomic variants with exceptional speed and accuracy. Since its first successful application for Mendelian disorder gene identification in 2010, next-generation sequencing (NGS) has become the cornerstone of molecular diagnostics, significantly improving the diagnostic yield for patients with rare and genetically heterogeneous conditions (Ng et al., 2010; Boycott et al., 2013). Whole exome sequencing (WES) and whole genome sequencing (WGS) have been particularly instrumental in reducing the duration of the diagnostic process for neurodevelopmental disorders, as they permit the simultaneous interrogation of thousands of genes and the identification of novel disease-causing variants that would be overlooked by traditional Sanger sequencing (Yang et al., 2013; Tengsujaritkul et al., 2026). Recent studies have highlighted the diagnostic utility of NGS in uncovering pathogenic variants underlying cognitive impairment, microcephaly, and neurodevelopmental disorders with ataxia, further underscoring its clinical value in precision medicine (Yavas et al., 2026; Ercoskun et al., 2026; Baris et al., 2026).

NSF deficiency syndrome, also known as Developmental and epileptic encephalopathy 96 (DEE96) (OMIM 619340), is a very rare autosomal dominant disorder characterized by global developmental delay, failure to thrive, microcephaly, hypotonia, and epileptic encephalopathy (Suzuki et al., 2019; Hayashi et al., 2023; Li et al., 2024). It is caused by heterozygous variants in the NSF gene (NM_006178.4; OMIM 601633), locate at chromosome 17q21.31, which contains 21 exons and encodes the N-ethylmaleimide sensitive factor, vesicle fusing ATPase protein involved in membrane fusion (Rizo, 2018; Kh et al., 2022; Söllner et al., 1993). As a homohexameric AAA-type ATPase that mediates membrane fusion, NSF precisely drives vesicular transport underlying neurotransmitter release and hormone secretion through dynamic SNARE (soluble NSF attachment protein receptor) complex assembly/disassembly orchestrated by calcium signaling (Ferro-Novick and Brose, 2013). The dysfunction of this machinery is increasingly recognized as the fundamental pathogenic basis for diabetes and neurological disorders. Pathogenic variants of SNARE complex components are associated with Alzheimer’s disease, Parkinsonism, autism, and psychiatric disorders (Ramakrishnan et al., 2012). However, monogenic disease phenotypes of NSF mutations had not been described until recently.

To date, nine individuals carrying pathogenic or likely pathogenic variants in the NSF gene have been identified. However, only four of these individuals have been described in detail as exhibiting the clinical features of DEE96 (Supplementary Table S1) (Suzuki et al., 2019; Hayashi et al., 2023; Li et al., 2024). All variants identified were missense variants predicted to defective autophagy through overactivation of the mammalian/mechanistic target of rapamycin (mTOR) pathway. Further reports on NSF variants and their phenotypes will be crucial for understanding this disease. Here, we investigated a previously unreported heterozygous missense variant [c.1055 A>G; p (Asn352Ser)] in a Chinese foetus with DEE96. Additionally, we have summarised the genotypic, phenotypic and clinical features of DEE96 through a review of the literature.

Materials and methods

Subjects and ethics approval

The study was approved by the Institutional Review Board and Ethics Committee of the Guangxi Maternal and Child Health Hospital. Detailed informed consent was obtained from the patient’s parents for the publication of their relevant clinical data in this paper.

Prenatal ultrasonography

Ultrasound examinations were performed using a Toshiba Aplio 500 system (Toshiba Medical Systems, Tokyo, Japan) equipped with a 3–6 MHz transabdominal transducer, with imaging parameters standardized according to manufacturer protocols for obstetric scanning. The primary scans were conducted by an experienced fetal medicine specialist (>5 years of dedicated obstetric ultrasound experience). For quantitative parameters, two consultant sonographers independently acquired measurements, with each observer obtaining three consecutive readings to ensure reliability. The final values were calculated as the mean of these replicate measurements.

Genetic analysis

Trio-based whole-exome sequencing (WES) was performed on the proband and both parents. Genomic DNA was extracted from amniotic fluid samples of the foetus using the QIAamp DNA Mini Kit (Qiagen, Hilden, Germany), while parental DNA was isolated from peripheral blood lymphocytes using the La-Aid DNA kit (Zeesan Biotech Co., Ltd., Xiamen, China). For each sample, 3 μg of DNA was sheared into fragments of 180–250 base pairs (bp), followed by end-repair, A-tailing, and adapter ligation. Exome capture was carried out using the Agilent SureSelect Human All Exon V6 Kit (Agilent Technologies, Santa Clara, CA, United States). The resulting libraries, with insert sizes ranging from 280 to 320 bp, were validated on a High-Sensitivity DNA chip using the 2,100 Bioanalyzer (Agilent Technologies). Subsequently, paired-end sequencing (150 bp) was performed on the Illumina HiSeq 2,500 platform (Illumina, San Diego, CA, United States). The raw reads were then aligned to the human reference genome GRCh38/hg38 using the Burrows-Wheeler Aligner (BWA, version 0.7.15). Variant calling and annotation were performed using TGex software (LifeMap Sciences Inc., version 5.7) and the Genome Analysis Toolkit (GATK). Prioritization of candidate variants was conducted using in silico pathogenicity prediction tools, including REVEL, CADD, and AlphaMissense. Sanger sequencing was performed to validate the candidate variant in the proband and available unaffected family members. Wild-type and mutant models were constructed using the AlphaFold server. The model with the highest Global Model Quality Estimation (GMQE) score was then selected for further analysis. The models were visualised using the PyMOL molecular graphics system (version 1.3, Schrödinger, LLC). The following primers, which were designed using Primer3, were used: 5′-CTT​CTC​ACC​AAA​ACT​CGG​C-3′ and 5′-CAT​CTT​TTT​CAC​CTG​GGT-3′ for c.1055 A>G; p (Asn352Ser). Variant interpretation followed the guidelines of the American College of Medical Genetics and Genomics/Association for Molecular Pathology (ACMG/AMP) (Richards et al., 2015).

Result

A 31-year-old woman was referred to the Department of Prenatal Diagnosis and Genetic Counselling at the Guangxi Zhuang Autonomous Region Maternal and Child Healthcare Hospital in Nanning, China, because of left clubfoot and nuchal fold thickening. She conceived naturally and had a routine first-trimester ultrasound scan at 12 weeks of pregnancy, which revealed a nuchal translucency measurement of 1.6 mm. The parents were non-consanguineous and healthy, with no family history of neurological or other hereditary disorders. The mother denied drug exposure during the periconceptional period and pregnancy, and received routine antenatal care. At 23 weeks of gestation, detailed ultrasound examination revealed an appropriately grown fetus with increased nuchal fold (NF) thickness (9.8 mm), left clubfoot, cardiomegaly, hepatosplenomegaly, and elevated middle cerebral artery peak systolic velocity (MCA-PSV, 44.5 cm/s) (Figure 1). Upon follow-up at 31 weeks, these findings persisted, with additional sonographic evidence of fetal hepatosplenomegaly. Based on these combined Doppler and morphological abnormalities, fetal anemia was suspected. Cordocentesis confirmed severe anemia in the fetus (hemoglobin 54 g/L), with a red blood cell count of 1.45 × 10¹²/L. The fetal anemia was normocytic normochromic. The fetus has blood type AB and is Rh-positive. Maternal investigations revealed blood group AB Rh-positive with normal mean corpuscular volume (MCV); irregular antibody screening and infectious workup (syphilis, cytomegalovirus, and parvovirus B19) were all negative. No other significant structural anomalies were identified. After receiving genetic counselling, the parents opted to terminate the pregnancy at 32 weeks.

FIGURE 1.

Phenotypes of the patients (A–D) Ultrasound images at 23 weeks’ gestation showing: (A) increased nuchal fold thickness (9.8 mm); (B) left clubfoot;(C) cardiomegaly with a cardiothoracic ratio of 0.35; (D) elevated middle cerebral artery peak systolic velocity (MCA-PSV: 44.5 cm/s).

Genetic analysis

The proband underwent trio-based whole exome sequencing to investigate the genetic etiology of the multiple malformations (Figure 2A). The proband underwent trio-based whole exome sequencing to investigate the genetic etiology of the multiple malformations (Figure 2A). A de novo heterozygous missense variant [c.1055 A>G, p (Asn352Ser)] was identified in the NSF gene (Supplementary Figure S1). Sanger sequencing confirmed the presence of this variant in the proband and its absence in both parents, establishing its de novo origin (Figure 2B). Detailed analysis of the trio-WES data confirmed that the exon containing the variant had a sequencing depth of 244× and a variant allele fraction of 44.67%, with reads demonstrating high mapping quality (MAPQ) and balanced strand distribution; no significant strand bias or mapping artifacts were detected, confirming that this locus was adequately and uniquely mappably covered. Although the genomic region harboring NSF may present potential challenges for short-read sequencing, the variant-bearing exon does not fall within any known low-mappability or high-GC regions in the present dataset, and an IGV screenshot is provided in the supplementary materials to display read-level supporting evidence (including allele balance and read distribution) for direct visual assessment. The variant [c.1055 A>G; p (Asn352Ser)] in the NSF gene is novel and has not been previously reported. It is absent from the public population databases, including the Exome Sequencing Project, gnomAD (v4.1.1), the 1000 Genomes Project, and dbSNP. The variant is also not documented in disease-specific repositories such as ClinVar and the Human Gene Mutation Database. The p (Asn352Ser) variant is located within the D1 domain of the NSF hexamer, a region critical for ATP binding and hydrolysis, the essential driving force for SNARE complex disassembly (Figure 2C). Three complementary in silico approaches supported the pathogenicity of this variant: (i) REVEL (v1.3), an ensemble missense predictor (score: 0.73; pathogenic threshold >0.5); (ii) CADD (v1.6), which integrates conservation and functional annotation (PHRED-like score: 25.3; pathogenic threshold >20); and (iii) AlphaMissense, a structure-based predictor leveraging AlphaFold2 protein structures (score: 0.81; pathogenic threshold >0.564) (Table 1). Multiple sequence alignment showed that the Asn352 residue is highly conserved across diverse organisms (Figure 3). The Asn352Ser mutation causes the amino acid at position 352 to change from aspartic acid to serine, altering the orientation of the side chains near the mutation site and the potential hydrogen-bond interactions. A 3D protein structure visualisation analysis using PyMOL software revealed that the missense mutation c.1055 A>G (p.Asn352Ser) alters the orientation of side chains near the mutation site and potential hydrogen-bond interactions (Figure 4). According to the ACMG/AMP standards and guidelines, this variant is classified as ‘likely pathogenic’ based on the supporting criteria PS2, PM2 and PP3 (Table 1).

FIGURE 2.

Genetic analysis of the affected proband (A) The spectrum of pathogenic variants in NSF. The red variants represent the novel variants identified in this study (B) Pedigree chart of the family of the fetus with developmental and epileptic encephalopathy 96. The proband is indicated by a black arrow (C) Sanger sequencing DNA chromatograms of NSF indicated the missense c.1055 A>G; p (Asn352Ser) variant was de novo.

TABLE 1.

Predicted pathogenicity of de novo NSF variant.

Gene	Variant (NM_145331.2)	Inheritance	Revel	AlphaMissense	CADD	ACMG/AMP
NSF (NM_006178.4)	c.1055A>G (p.Asn352Ser)	DNV	0.73	0.81	25.3	LP (PS2+PM2+PP3)

DNV, de novo variant; D, damaging; LP, Likely pathogenic.

FIGURE 3.

Multispecies alignment showing the strong conservation of NSF p. Asn352.

FIGURE 4.

Predicted molecular basis of N352 S (A) Wild type (B) Mutant type. The Asn352Ser mutation causes the amino acid at position 352 to change from aspartic acid to serine, altering the orientation of the side chains near the mutation site and the potential hydrogen-bond interactions.

Discussion

Developmental and Epileptic Encephalopathy 96 (DEE96) is an emerging example of a ‘SNAREopathy’, which is a neurological disorder that arises from dysfunctional SNARE-mediated membrane fusion machinery (Boycott et al., 2013). Since its initial description, only three pathogenic variants [p (Ala459 Thr), p (Pro563Leu) and p (Glu715Lys)] have been reported in four patients who exhibited multi-domain developmental abnormalities, including global developmental delay, epileptic encephalopathy, hypotonia, feeding difficulties and severe anaemia (Suzuki et al., 2019; Hayashi et al., 2023; Li et al., 2024). Here, We describe a novel de novo heterozygous missense variant [c.1055 A>G; p (Asn352Ser)], which was identified prenatally. This is the second reported case of DEE96 diagnosed during the fetal period.

Clinically, the core postnatal phenotypes of DEE96 established by previous cases include global developmental delay, infantile-onset epileptic encephalopathy, microcephaly, hypotonia, and failure to thrive (Table 2) (Suzuki et al., 2019; Hayashi et al., 2023; Li et al., 2024). In contrast, prenatal ultrasound in this case revealed characteristic features including increased nuchal fold thickness, clubfoot, cardiomegaly, and hepatosplenomegaly, accompanied by severe fetal anemia, suggesting that neurological and extraneurological dysfunction may manifest during fetal development. To our knowledge, increased nuchal fold thickness and clubfoot are reported for the first time in association with DEE96. The left-sided clubfoot observed in this case likely reflects neuromuscular dysfunction secondary to impaired neurotransmitter release at the neuromuscular junction, a mechanism supported by findings that NSF variants can cause neuromuscular junction impairment (Hayashi et al., 2023). Furthermore, this case represents the third instance of severe anemia, bringing the total incidence of anemia to 60% (3/5), indicating that anemia may be a characteristic manifestation of DEE96. The anemia in these patients cannot be explained by common etiologies such as alloimmunization, infection, or α-thalassemia. The severe fetal anemia may reflect impaired erythrocyte membrane dynamics or defective autophagy in erythroid progenitors, given the essential role of autophagy in red blood cell maturation and mitochondrial clearance, suggesting that NSF may participate in hematopoietic processes or erythrocyte membrane dynamics regulation (Chen et al., 2008; Grosso et al., 2017). The hepatosplenomegaly may represent extramedullary hematopoiesis secondary to anemia, or alternatively, may indicate primary involvement of NSF in these tissues. These clinical findings further expand DEE96 from a purely neurodevelopmental disorder to a potentially systemic disease with prenatal onset, suggesting that NSF-related variants should be considered in the differential diagnosis of complex fetal syndromes associated with hematological abnormalities.

TABLE 2.

Summary of reported information on clinical phenotypes associated with NSF variants in patients with DEE96.

Patients	Our patient	Suzuki et al. (Suzuki et al., 2019)		Hayashi et al. (Hayashi et al., 2023)	Li et al. (Li et al., 2024)	Total
Clinical data	Patient1	Patient2	Patient3	Patient4	Patient5	N = 5
Variants in NSF (NM_006178.4)	c.1055 A>G (p.Asn352Ser)	c.1375G>A (p.Ala459Thr)	c.1688C>T (p.Pro563Leu)	c.2143G>A (p.Glu715Lys)	c.1688C>T (p.Pro563Leu)	Missense = 4
Inherit	De novo	De novo	De novo	De novo	De novo	​
Affected exon/intron	10	13	15	19	15	​
Gender	NA	Femal	Femal	Femal	NA	​
Age at last examination	31-week-old fetus	36D (deceased)	6Y	4Y	17-week-old fetus	​
Birth week	NA	37 W	33 W	-	NA	​
Fetal hydrops	-	NA	+	-	+	2/4
Intrauterine growth restriction	-	NA	+	-	+	2/4
Anemia	Severe	-	+	-	Severe	3/5
Microcephaly	NA	-	<-3.4SD	-	NA	1/3
Respiratory problems	NA	+	+	-	NA	2/3
Intellectual disability	NA	NA	Severe	-	NA	1/2
Motor delay	NA	NA	Severe	-	NA	1/2
Vomiting	NA	+	NA	-	NA	1/2
Age at seizure onset	NA	Neonatal period	At birth	No seizure	NA	​
Seizure	NA	Tonic seizures	Myoclonic seizures	Mild spastic paraplegia and startle response (both only in infancy)	NA	3/3
EEG	NA	+	+	-	NA	2/3
Other features	Cardiomegaly, hepatosplenomegaly; increased middle cerebral artery peak systolic velocity (44.5 cm/s; >2.0 MOM),Nuchal fold thickening (14.4 mm),Left clubfoot	​	​	​	Cardiomegaly, hepatosplenomegaly; pleural effusion, ascites,increased middle cerebral artery peak systolic velocity (44.5 cm/s; >2.0 MOM),increased placental depth (2.7 cm)	​

DD96:Developmental and epileptic encephalopathy 96; Y: years; W: weeks; D:days; N: numbers; NA: not available; SD:standard deviation; EEG:Electroencephalography.

It is noteworthy that most mutations in NSF are lethal across species. The classic “comatose” (com) mutant in Drosophila is a temperature-sensitive paralytic mutant and was one of the earliest models to link NSF dysfunction to severe neurological phenotypes. At restrictive temperatures, this mutant exhibits rapid activity-dependent paralysis and synaptic transmission failure due to impaired vesicle recycling (Siddiqi and Benzer, 1976; Pallanck et al., 1995). Furthermore, biochemical analysis of the comt st53 mutation (S483L in dNSF1, corresponding to S491L in mammalian NSF) revealed a specific defect in α-SNAP-stimulated ATPase activity without impairing α-SNAP binding. When this mutation was engineered into yeast Sec18, it produced a dominant lethal phenotype in vivo, underscoring the exquisite sensitivity of the NSF hexamer to perturbations in ATPase regulation (Pallanck et al., 1995). Similarly, distinct temperature-sensitive comatose alleles that alter different amino acids within the D1 domain converge on a common phenotype of impaired SNARE complex disassembly and synaptic transmission failure, indicating that D1 domain integrity is essential for NSF function (Stewart et al., 2002). Consistent with these findings, the fetus in our study, which harbored a p. Asn352Ser mutation in the NSF D1 domain, exhibited a similarly severe phenotype. Multiple sequence alignment confirmed absolute conservation of this asparagine residue across diverse species. Analysis of the protein’s three-dimensional structure revealed that the c.1055 A>G missense mutation (p.Asn352Ser) altered the orientation of the side chains near the mutation site, affecting potential hydrogen-bond interactions. The NSF D1 domain is structurally arranged as a hexameric ATPase ring, and inter-subunit interactions are essential for coordinating ATP hydrolysis and substrate translocation (Moeller et al., 2012; Khan et al., 2025; Zhao and Brunger, 2016; Söllner et al., 1993). Recent cryo-EM studies have further demonstrated that NSF employs a sequential hydrolysis mechanism, whereby ATP binding and hydrolysis propagate around the D1 ring via specific inter-subunit interfaces, driving substrate threading and SNARE complex disassembly (Moeller et al., 2012; Khan et al., 2025; Zhao and Brunger, 2016; Söllner et al., 1993). Asn352 is located precisely at one such critical inter-subunit junction. Therefore, the p (Asn352Ser) substitution likely disrupts the structural integrity of the hexamer interface, potentially impairing either the assembly or stability of the NSF homohexamer or the allosteric communication required to coordinate ATP hydrolysis around the ring structure. Interestingly, one individual harbouring the p (Phe715Lys) mutation exhibited an unusually mild phenotype among previously reported patients (Hayashi et al., 2023). Structural analysis of NSF indicates that, in contrast to the P563L and A459 T mutants, the E715 K mutation is not located at a site that could interfere with the transmission of conformational changes in NSF monomers. Rather, it only partially obstructs the process by which NSF monomers undergo rotational ATP hydrolysis (Hayashi et al., 2023). This phenotypic heterogeneity reflects the mutation’s specific effects on different NSF functional domains, or its differential impact on the dynamics of tissue-specific SNARE complexes during development.

The pathogenic mechanisms underlying NSF deficiency syndrome remain incompletely elucidated. Current evidence indicates that disease-associated variants disrupt multiple interconnected cellular processes. As a homohexameric AAA + ATPase, NSF primarily orchestrates the disassembly of SNARE complexes following vesicle fusion, thereby facilitating SNARE protein recycling and maintaining subsequent membrane fusion events (Orr et al., 2017). This function is critical not only for regulating synaptic vesicle neurotransmitter release but also for autophagosome-lysosome fusion—a process essential for cellular homeostasis and neuronal survival (Ramakrishnan et al., 2012; Baudot et al., 2022). Moreover, beyond its canonical role in post-fusion SNARE recycling, biochemical and structural studies have revealed that NSF and its yeast orthologue Sec18 act as crucial priming factors, disassembling non-fusogenic binary SNARE complexes (e.g., syntaxin-1A–SNAP-25) prior to membrane fusion (Zhao et al., 2012). This pre-fusion disassembly prevents the formation of off-pathway SNARE assemblies and ensures the availability of fusion-competent SNAREs for trans-complex formation,a quality control function that is essential for synaptic vesicle priming (Söllner et al., 1993; Morgan, 1996; Coleman, 1974; Woodman, 1997). Recent atomic-resolution cryo-EM structures have directly observed the interaction between NSF and the binary syntaxin–SNAP-25 complex, revealing how the D1 ATPase ring drives its disassembly (White et al., 2018; White et al., 2025). In zebrafish, NSF mutations impair the disassembly of both binary and ternary SNARE complexes, leading to defects in the organizational structure and sensory function of myelinated axons. This provides physiological evidence in vertebrates that NSF-mediated preactivation is essential for normal neural development (Mo and Nicolson, 2011). Hayashi et al. (2023) further demonstrated that DEE-associated NSF variants cause defective autophagy through overactivation of the mTOR signaling pathway (Hayashi et al., 2023). In PC12 cells and patient-derived iPSC-differentiated neurons, pathogenic NSF variants led to aberrant autophagosome accumulation, impaired autophagic flux, and neurite degeneration characterized by beading morphology—these being hallmarks of neurodegeneration. Notably, treatment with the mTOR inhibitor rapamycin or overexpression of wild-type NSF significantly ameliorated these cellular phenotypes, establishing a causal link between NSF dysfunction, mTOR hyperactivation, and neurodegeneration (Hayashi et al., 2023). Further investigation suggested that aberrant mTOR activation in NSF-deficient cells may directly result from impaired autophagosome-lysosome fusion. Under physiological conditions, SNARE complexes mediate the fusion of autophagosomes with lysosomes to form autolysosomes for intracellular waste degradation, while NSF is responsible for recycling autophagy-related SNARE proteins (including STX17, SNAP29, and VAMP8) during this process (Ramakrishnan et al., 2012; Levine and Kroemer, 2019; Baudot et al., 2022). When NSF function is compromised, SNARE complex disassembly is obstructed, leading to autophagic flux arrest and potentially triggering compensatory mTORC1 activation as a cellular feedback response to lysosomal insufficiency (Saxton and Sabatini, 2017). This abnormal activation may establish a self-perpetuating vicious cycle: autophagy impairment promotes the accumulation of aberrant proteins and damaged organelles, further exacerbating cellular stress and ultimately driving neurodegenerative progression (Hayashi et al., 2023; Levine and Kroemer, 2019). The multi-system involvement observed in this case indicates that the impact of NSF dysfunction extends beyond the nervous system to affect fundamental cellular processes. Further functional studies are warranted to deepen our understanding of NSF-related disorders and their underlying mechanisms.

This study has several limitations. Firstly, the small number of reported cases and the limited range of variants mean that the phenotypic profile of NSF-related disorders is incomplete. Secondly, the functional consequences of the identified variant were not directly assessed, so further functional investigations are needed, such as protein expression analyses and cellular models. Thirdly, while the identified NSF variant offers a plausible explanation for the patient’s phenotype, other molecular mechanisms, including genomic polymorphisms, epigenetic alterations and regulatory factors that may influence clinical expressivity, were not explored (Tüysüz et al., 2023; Coskunpinar et al., 2021). Fourthly, our trio-WES pipeline was designed and validated to detect SNVs and small indels within coding exons and flanking splice-site regions (±10 bp). However, it was not optimised to detect CNVs, structural rearrangements, chromosomal aneuploidies or mosaic variants. Future studies integrating functional validation with multi-omic analyses are needed to improve our understanding of NSF-DEE96.

In summary, our study reports a novel de novo NSF variant p (Asn352Ser) in a foetus with DEE96, broadening the disorder’s phenotype to include severe prenatal systemic manifestations such as anaemia, increased nuchal fold thickness and clubfoot. This study extends the genotype-phenotype spectrum of the disorder. Furthermore, our findings and literature review confirm DEE96 as a ‘SNAREopathy’, in which pathogenic variants disrupt vesicular fusion and autophagy, potentially via mTOR overactivation. This broadens our understanding of DEE96, shifting the focus from a purely neurodevelopmental disorder to a condition with multisystem fetal onset. Anemia emerges may be a key feature, emphasising the importance of considering NSF in the prenatal diagnosis of complex syndromes.

Funding Statement

The author(s) declared that financial support was received for this work and/or its publication. This research was supported by the Guangxi Science and Technology Program (21–220–22, GuiKe LT2600640039), the Guangxi Natural Science Foundation under Grant (No.2026GXNSFAA00641253) and the Health Department of Guangxi Province (Grant No. Z-A20220256).

Data availability statement

The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found at: https://www.ncbi.nlm.nih.gov/sra/PRJNA1492294.

Ethics statement

The studies involving humans were approved by Institutional Review Board and Ethics Committee of the Guangxi Maternal and Child Health Hospital. The studies were conducted in accordance with the local legislation and institutional requirements. Written informed consent for participation in this study was provided by the participants’ legal guardians/next of kin. Written informed consent was obtained from the individual(s), and minor(s)' legal guardian/next of kin, for the publication of any potentially identifiable images or data included in this article.

Author contributions

QY: Resources, Conceptualization, Project administration, Writing – review and editing, Investigation, Validation, Funding acquisition, Methodology, Writing – original draft, Formal Analysis, Data curation, Supervision, Software, Visualization. ZQ: Methodology, Validation, Conceptualization, Software, Formal Analysis, Writing – review and editing, Data curation. JaL: Data curation, Investigation, Visualization, Formal Analysis, Software, Writing – review and editing, Resources. QZ: Project administration, Visualization, Investigation, Data curation, Supervision, Methodology, Writing – review and editing. XZ: Methodology, Software, Investigation, Resources, Formal Analysis, Writing – review and editing. SeY: Data curation, Validation, Methodology, Writing – review and editing, Project administration, Investigation, Software. SZ: Writing – review and editing, Data curation, Software, Methodology, Resources, Validation, Formal Analysis. WL: Formal Analysis, Methodology, Data curation, Validation, Investigation, Writing – review and editing, Software. SaY: Conceptualization, Validation, Writing – review and editing, Investigation, Supervision, Methodology, Software. SH: Writing – review and editing, Formal Analysis, Visualization, Project administration, Validation, Resources, Software, Funding acquisition. JnL: Resources, Funding acquisition, Project administration, Software, Writing – review and editing, Validation, Formal Analysis, Writing – original draft, Supervision, Data curation, Methodology, Investigation, Visualization, Conceptualization.

Conflict of interest

The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fgene.2026.1882449/full#supplementary-material

SUPPLEMENTARY FIGURE S1

IGV plots of SNVs. Whole exome sequencing of NSF indicated the missense c.1055A>G; p.(Asn352Ser) variant was de novo.