SUPT16H-associated neurodevelopmental disorder and neurocristopathy: genetic and phenotypic spectrum

Abstract

SUPT16H encodes a subunit of the FACT (FAcilitates Chromatin Transcription) complex, a histone chaperone essential for maintaining chromatin integrity during transcription, replication, and DNA repair. Pathogenic de novo SUPT16H missense variants have previously been linked to neurodevelopmental disorders in eight individuals. Here, we expand the genotypic and phenotypic spectrum by identifying 24 additional individuals harboring ultrarare heterozygous missense or truncating variants, who share overlapping clinical features including intellectual disability, autism spectrum disorder, hypotonia, and characteristic craniofacial dysmorphism. To elucidate the underlying mechanisms, we generated a supt16h knockout zebrafish model using CRISPR/Cas9. The supt16h loss-of-function (LOF) model recapitulated key patient phenotypes such as developmental delay, craniofacial anomalies, and hypotonia. Structural and functional analyses of selected SUPT16H variants demonstrated differential rescue of developmental defects in supt16h-deficient embryos, indicating variant-specific LOF effects in vivo. The presence of non-neural manifestations, including facial and ear anomalies, suggested a role for SUPT16H in neural crest development. Consistently, supt16h loss impaired neural crest cell migration and differentiation and triggered p53-dependent apoptosis in the central nervous system (CNS) and neural crest–derived pharyngeal arches. Notably, supt16h deficiency impaired oligodendrocyte specification in the CNS and perturbed differentiation of neural crest– derived Schwann cells in the peripheral nervous system, providing a plausible basis for hypotonia. These findings uncover a previously unrecognized role of SUPT16H in neural crest development, linking chromatin regulation to neural crest-derived lineage specification and differentiation, thereby defining SUPT16H deficiency as a neurocristopathy that broadens the clinical and mechanistic landscape of SUPT16H-associated disorders.

Introduction

SUPT16H encodes a subunit of the FACT (FAcilitates Chromatin Transcription) complex, a chromatin-specific histone chaperone that regulates DNA replication, transcription, and repair. The FACT complex is highly expressed in stem and undifferentiated cells but is downregulated during differentiation [1]. Dysregulation of the FACT complex has been frequently observed in human cancers [1–3]. De novo missense variants in SUPT16H have been implicated in neurodevelopmental disorders characterized by intellectual disability, autism spectrum disorder (ASD), dysmorphic features, and structural brain anomalies [4–8]. Functional analyses of human SUPT16H homologs in Drosophila (dre4) and mice (Supt16h) have shown that loss of function (LOF) of these genes causes neurodevelopmental defects, demonstrating the pathogenicity of human SUPT16H variants [5, 9]. Moreover, studies using supt16h mutant zebrafish revealed that supt16h is essential for embryonic viability and hematopoietic stem cell development [10].

The spectrum of SUPT16H-associated disease likely reflects the diverse developmental functions of SUPT16H. Clinical reports describing individuals with SUPT16H variants have noted craniofacial dysmorphism, hearing loss, and cardiac defects—features suggestive of neural crest abnormalities. This observation led us to hypothesize that SUPT16H plays a crucial role in neural crest development.

Neural crest cells (NCCs) are multipotent embryonic stem cells unique to vertebrates. They arise at the neural plate borders, undergo epithelial-to-mesenchymal transition (EMT) during neural tube closure, migrate extensively throughout the embryo, and differentiate into diverse derivatives, including peripheral neurons, glia, craniofacial structures, pigment cells, and endocrine cells, depending on their axial level [11]. Disruptions in NCC formation, migration or differentiation results in neurocristopathies—congenital disorders affecting neural crest–derived tissues [12].

Given the limited understanding of SUPT16H function in development and its association with the broad clinical manifestations observed in SUPT16H-related disorders, we aimed to expand the genotypic and phenotypic spectrum of SUPT16H-associated disease. Through international collaboration, we identified 24 individuals harboring heterozygous SUPT16H variants who exhibited a consistent multisystem phenotype suggestive of impaired neural crest development. Using supt16h mutant zebrafish, we further characterized its developmental role and demonstrated that supt16h is required for proper migration and differentiation of NCCs. Loss of supt16h induced p53-dependent apoptosis in the central nervous system (CNS) and neural crest derivatives, revealing a previously unrecognized role of SUPT16H in early neurocristogenesis.

Results

Clinical and genetic overview

We analyzed clinical data from 24 individuals (10 females, 14 males) from 22 unrelated families carrying ultrarare heterozygous SUPT16H variants, aged 1.66–25 years (Supplementary Table 1). Participants were identified through an international collaborative network of diagnostic laboratories, the GeneMatcher Exchange platform [13], screening of public variant databases (ClinVar and DECIPHER [14]), and the Brain Development Research Program at the University of California, San Francisco (UCSF). Identified variant types included 18 missense variants (one shared by two affected siblings), one in-frame indel in a non-repeat region, one nonsense variant, and four frameshift deletions (one shared by two siblings). Eleven variants occurred de novo. In six individuals, inheritance could not be determined or was excluded in only one parent due to limited availability of parental DNA. Among the remaining seven inherited variants, three were maternally transmitted from mildly affected mothers (p.I747R and p.K696Sfs*7 to two siblings), whereas three were maternally inherited from unaffected mothers (p.E311D in two siblings, p.A400S, and p.T828M) (Supplementary Table 1). Two recurrent missense variants (p.R734W and p.R847W) have been previously described in affected individuals with overlapping phenotypes, and p.I747R has been reported in association with neurodevelopmental disorder [15].

Gene constraint

Canonical SUPT16H (NM 007192.4) encodes a 1047–amino acid chromatin-remodeling protein (SPT16) comprising 27 exons (Fig. 1A). The gene exhibits strong selective constraint against protein-truncating variants (PTVs; probability of loss-of-function intolerance [pLI] = 1, loss-of-function observed/expected upper bound fraction [LOEUF] = 0.24) and missense variation (Z = 7.02), with the greatest constraint across exons 14–23 (amino acids 507–970). Domain organization and length are highly conserved from yeast to mammals [16]. All variants identified in this study were absent from gnomAD v4.1, except p.T828M (Individual 10), p.I172V (Individual 19), and p.N685S (Individual 23), each present at a population frequency of < 0.0005%.

Figure 1.

Constraint landscape of SUPT16H, variant distribution, and representative brain MRIs. (A) MetaDome tolerance plot and schematic of human SUPT16H (NM_007192), annotated with putative domains and locations of novel (blue) and previously reported (grey) missense variants, as well as novel truncating variants (orange). Domains: N-terminal domain (NTD; amino acids 5–411), dimerization domain (DD; 529–651), middle domain (MD; 660–895), and C-terminal domain (CTD; 927–1020). (B) Representative MRI scans. T1-weighted sagittal images show HCC predominantly involving the genu (individual 7, p.R734W; individual 11, p.V840A), and an axial T2-weighted image shows posterior periventricular white-matter hyperintensity (individual 13, p.R213*). Abbreviations: CTD, C-terminal domain; DD, dimerization domain; HCC, hypoplastic corpus callosum; MD, middle domain; MRI, magnetic resonance imaging; NTD, N-terminal domain; SUPT16H, suppressor of ty 16 homolog.

Variant classification

Eighteen variants fulfilled the American College of Medical Genetics and Genomics (ACMG) criteria for pathogenic or likely pathogenic (P/LP) status and were predicted to exert structural or transcriptional effects (13 missense or in-frame variants and 5 PTVs; Supplementary Table 1). Individual 23 harbored two P/LP de novo variants—SUPT16H p.N685S and a mosaic fibroblast growth factor receptor 1 (FGFR1) p.Q288Rfs*3 detected by allelic imbalance—and was excluded from aggregate phenotypic analyses due to overlapping features of SUPT16H- and FGFR1-related disorders. Five additional variants (p.G93S in Individual 18, p.I172V in Individual 19, p.E311D in Individuals 20 and 21, p.A400S in Individual 22, and p.T736A in Individual 24) were classified as variants of uncertain significance (VUS) despite good phenotypic correlation and were similarly excluded from summary phenotype analyses.

Neurodevelopmental features

Unless otherwise specified, frequencies refer to individuals in the P/LP subset with available data (n/N, %). Individual-level data are provided in Supplementary Table 1, and cohort-level summaries in Table 1. Head size and shape anomalies—including dolichocephaly, scaphocephaly, macrocephaly, or microcephaly—were observed in 6 of 17 individuals (35.3%). Gross motor delay was noted in 10 of 17 (58.8%), although all participants aged ≥5 years ultimately achieved independent ambulation. Speech and language delay was frequent (15/17, 88.2%), with at least 5 of 17 (29.4%) individuals not producing sentences by ≥5 years of age; one individual demonstrated normal early language milestones but later developed selective mutism at 13 years. Intellectual disability (ID) was present in all but one assessed individual, and ASD—defined by clinical diagnosis or autistic traits—was reported in 11 of 14 (78.6%; 6 females, 5 males). Central hypotonia was also common (10/17, 58.8%). Among those with available data, epilepsy occurred in 2 of 11 (18.2%), with onset during the first two years of life and heterogeneous seizure semiology.

Table 1.

Core clinical features of 17 individuals with pathogenic or likely pathogenic SUPT16H variants.

Clinical finding	N assessed (this study)	% affected (this study)	N assessed (all studies)	% affected (all studies)
Developmental delay	17	88.2	25	92.0
Intellectual disability	10	9.0	16	93.8
Autism	14	78.6	17	76.5
Epilepsy	17	11.8	25	20.0
Abnormal MRI findings	11	45.5	18	66.7
Hypotonia	17	58.8	24	58.3
Sleep disturbance	16	31.3	21	42.9
Dysmorphic eyes	17	29.4	24	45.8
Vision abnormalities	17	17.6	24	16.7
Abnormal skull or head shape	17	41.2	24	58.3
Dysmorphic nose and/or mouth	17	41.2	24	50.0
Dysmorphic ears	17	11.8	24	25.0
Hearing loss	17	11.8	24	16.7
Congenital heart defects	17	5.9	25	20.0
Dysmorphic hands and/or feet	17	41.2	21	37.5
GI/GU abnormalities	17	47.1	21	33.3
Endocrine abnormalities	16	18.8	20	25.0
Feeding or swallowing difficulties	16	37.5	21	47.6
Respiratory distress at birth	17	17.6	23	21.7

Abbreviations: GI/GU, gastrointestinal/genitourinary.

Brain MRI findings

Postnatal brain magnetic resonance imaging (MRI) data were available for 11 individuals; abnormalities were identified in 5 of 11 (45.5%). Findings included hypoplastic corpus callosum (HCC) in 2 of 11 (18.2%), mild diffuse cerebral atrophy in 1 of 11 (9.1%), mild cerebellar vermis hypoplasia in 1 of 11 (9.1%), and periventricular T2-weighted white matter hyperintensity in 1 of 11 (9.1%) (Fig. 1B). Notably, Individual 12 (p.R847W)—a variant previously reported in an individual with HCC—showed a normal corpus callosum.

Other organ systems

Outside the nervous system, limb anomalies were observed in 7 of 17 individuals (41.2%), including tapered fingers, short metacarpals, and clinodactyly. Gastrointestinal and genitourinary involvement occurred in 7 of 17 (41.2%), with features such as duodenal atresia, fatty liver, hydronephrosis, multicystic kidneys, hypospadias, and constipation. Additional systemic findings were less frequent and included respiratory distress at birth in 3 of 17 (17.6%), growth-hormone deficiency with short stature in 2 of 16 (12.5%), precocious puberty in 1 of 16 (6.3%), vision deficits in 3 of 17 (17.6%), hearing loss in 2 of 17 (11.8%), and a congenital heart defect in 1 of 17 (5.9%) (Table 1).

Facial dysmorphism subcohort

Beyond the neurodevelopmental features emphasized in previous reports [4–8], approximately half of the cohort exhibited dysmorphic facial features. We highlight eight individuals with prominent craniofacial dysmorphism (Table 2), encompassing abnormal head shape, nasal and oral morphology, and external ear anomalies with or without hearing loss. While causality cannot be inferred from cross-sectional data, the co-occurrence of these features suggests perturbations in neural crest–derived structures.

Table 2.

Clinical findings in SUPT16H variant carriers with facial dysmorphism.

Clinical feature	Individual 1	Individual 3	Individual 4	Individual 5	Individual 7	Individual 9	Individual 11	Individual 14
SUPT16H variant (NM_007192.4)	c.123_125delTGA, p.Asp41del	c.1345G > A, p.Asp449Asn	c.1713C > A, p.Asn571Lys	c.1943A > G, p.Gln648Arg	c.2200C > T, p.Arg734Trp	c.2429C > T, p.Pro810Leu	c.2519 T > C, p.Val840Ala	c.662_663delTG, p.Val221Glyfs*5
Inheritance	de novo	Unknown	de novo	de novo	de novo	de novo	unknown	de novo
Sex	Male	Male	Male	Male	Female	Male	Male	Female
Gestational age (weeks)	35	34	37+5	38+1	40	40	38	32
Age at last evaluation	1.66 y	5.5 y	25 y	4.17 y	6.8 y	15 y	1 y	9.16 y
Weight	12.5 kg	16.6 kg	68 kg	N/A	21.3 kg	70.8 kg	10.9 kg	51.5 kg
Intellectual disability	N/A	Yes	Yes	N/A	Yes	No	N/A	Yes
Neurodevelopmental delay	Yes	Yes	Yes	Yes	Yes	No	Yes	Yes
Autistic behavior	Yes	No	Yes	N/A	No	Yes	N/A	Yes
Seizures	Yes	No	No	No	No	No	No	No
Speech/language delay	Yes	Yes	Yes	Yes	Yes	No	Yes	Yes
Hypotonia	Yes	Yes	Yes	Yes	Yes	No	Yes	Yes
Brain MRI findings	Cerebral atrophy, hyperintensity, restricted diffusion (19 m)	N/A	No (18 y)	No	Hypoplastic corpus callosum (6.8 y)	No (15 y)	Thinning of corpus callosum, enlarged ventricles (4 m)	Enlarged ventricles, superior vermis hypoplasia (3 y)
Gross motor delay (age at walking)	Yes (not yet)	No (12 m)	Yes (NA)	Yes (not yet)	Yes (5 y)	No (11 m)	Yes (not yet)	Yes (2.5 y)
Dysmorphic skull/head	Yes	Yes	No	Yes	Yes	Yes	Yes	Yes
Dysmorphic nose/mouth	Yes	No	Yes	Yes	Yes	Yes	Yes	Yes
Dysmorphic ear	No	No	Yes	Yes	No	No	No	No
Hearing loss	No	No	No	Yes	No	No	No	Yes
Congenital heart defect	No	No	No	Yes	No	No	No	No
Dysmorphic hands/feet	Yes	No	Yes	Yes	Yes	Yes	No	Yes

Abbreviations: MRI, magnetic resonance imaging; N/A, not available; m, months; y, years.

Zebrafish supt16h loss of function phenocopies patient-associated anomalies

To investigate the in vivo functions of SUPT16H underlying the diverse clinical manifestations observed in affected individuals, we generated a zebrafish supt16h mutant (supt16h^{+ kan1}) using CRISPR/Cas9. A guide RNA targeting exon 13 introduced a 7 bp deletion that produced a premature stop codon (Fig. 2A). Homozygous supt16h^kan1/kan1 (supt16h^−/−) embryos failed to survive beyond 4 days post-fertilization (dpf), whereas heterozygous supt16h^+/kan1 (supt16h^+/−) embryos were viable to adulthood. The mutant transcript containing the 7 bp deletion underwent marked RNA decay (Fig. 2B; Fig. S1).

Figure 2.

The zebrafish supt16h^kan1/kan1 mutant recapitulates patient phenotypes and induces apoptosis in the central nervous system and pharyngeal arches. (A) Schematic representation of the zebrafish supt16h^kan1 mutant generated by CRISPR/Cas9, highlighting the guide RNA target site (red arrow) with a 7 bp deletion within exon 13 of the supt16h gene. Black boxes indicate exons. (B) Genomic DNA (gDNA) analysis of supt16h^+/kan1 (supt16h^+/−) embryos showing two bands corresponding to WT and mutant alleles (left). Mutant RNA exhibits marked decay in both supt16h^+/kan1 (supt16h^+/−) and supt16h^kan1/kan1 (supt16h^−/−) embryos at 2 dpf (right; Fig. S1). (C) Bright-field lateral images of supt16h^+/− and supt16h^−/− embryos at 1 and 2 dpf. Arrow, head; arrowhead, tail; open arrowhead, yolk. Scale bars, 400 μm; n ≥ 30 embryos per group. (D) Target specificity of supt16h gRNA validated by comparing morphological phenotypes of supt16h^kan1/sd41 compound mutants with homozygous mutants at 2 dpf. Scale bars, 400 μm. (E) Touch-evoked response of 2 dpf supt16h embryos showing locomotor abnormalities in supt16h^−/− larvae. Data are presented as mean values; n ≥ 30 embryos per group. (F) Immunohistochemistry with anti-acetylated tubulin antibody (pan-neuronal marker) visualizing axonal projections in the developing brain of supt16h^+/− and supt16h^−/− embryos at 2.5 and 3 dpf. Arrowhead, midline-crossing axons in the midbrain; arrow, cerebellum. (G) Lateral views of live 2 dpf embryos stained with acridine orange showing apoptotic cells in the brain (arrow), pharyngeal arches (arrowhead), and spinal cord (arrow). Scale bars, 200 μm. Statistical significance was determined using one-way ANOVA followed by Dunnett’s multiple comparison test; ****P < 0.0001. Abbreviations: CNS, central nervous system; CRISPR, clustered regularly interspaced short palindromic repeats; dpf, days post-fertilization; gDNA, genomic DNA; supt16h, suppressor of ty 16 homolog; WT, wild type.

Morphologically, supt16h^−/− larvae exhibited developmental delay at 1 dpf and progressively severe phenotypes by 2 dpf, including microcephaly and shortened body length, whereas supt16h^+/+ and supt16h^+/− larvae appeared normal (Fig. 2C). Phenotypes in supt16h^kan1/kan1 mutants were indistinguishable from those in the previously generated N-ethyl-N-nitrosourea (ENU)–induced supt16h^sd45/sd45 mutant [10], and identical defects were observed in compound lines, confirming the specificity of supt16h disruption (Fig. 2D).

Neurological assessment revealed impaired sensorimotor responses in supt16h mutants. In touch-evoked response and locomotor activity assays, supt16h^+/+ and supt16h^+/− embryos displayed normal linear movement, whereas most supt16h^−/− embryos showed abnormal circular movement or were unresponsive (Fig. 2E). Immunostaining with anti–acetylated α-tubulin demonstrated delayed and defective brain development in supt16h^−/− embryos, including disrupted midline axonal projections (Fig. 2F). Live imaging using acridine-orange staining revealed extensive apoptosis in the brain, pharyngeal arches, and spinal cord (Fig. 2G). Together, these findings indicate that supt16h loss of function causes widespread cell death in the CNS and craniofacial structures during early embryogenesis, providing a mechanistic basis for the microcephaly and craniofacial abnormalities observed in affected individuals.

Comprehensive structural analysis of SUPT16H variants

To elucidate how SUPT16H variants give rise to the diverse clinical phenotypes observed, we performed a comprehensive structural assessment of all currently reported SUPT16H missense variants. The encoded protein, SPT16 (referred as SUPT16H in this study), contains four evolutionarily conserved domains: an N-terminal domain (NTD), a dimerization domain (DD), a histone chaperone Rtt106-like middle domain (MD), and a C-terminal domain (CTD). Together with structure-specific recognition protein 1 (SSRP1), SUPT16H forms the FAcilitates Chromatin Transcription (FACT) complex. Among 17 distinct missense variants identified, five were located in the NTD, one in the linker between the NTD and DD, two in the DD, and nine in the MD (Fig. 1A).

To assess their potential structural impact, we mapped 16 of these variants onto previously resolved SUPT16H molecular structures [17–19]. Except for Asp449 (D449), which resides in a flexible loop unresolved in available models, all variant residues were positioned on either the cryo-electron microscopy (cryo-EM) structure of the FACT–nucleosome complex (Protein Data Bank [PDB] ID: 6UPK), where the NTD is disordered, or on crystal structures of the isolated NTD (PDB ID: 5E5B) and the MD bound to the H3–H4 tetramer [(H3–H4)₂] (PDB ID: 4Z2M) (Fig. 3A; Fig. S2A–C).

Figure 3.

Structural and functional analysis of SUPT16H variants. (A) Positions of the two SUPT16H variants within the structure of the FACT–nucleosome complex (PDB ID: 6UPK). The side chains of N571 (labeled in red) and Q648 (labeled in blue) are shown as yellow sticks. SUPT16 and SSRP1 are depicted in green and cyan, respectively, along with DNA bound to the (H3–H4)₂ tetramer and the H2A–H2B dimer. Right: Close-up view of the polar interaction network surrounding N571, with polar contacts indicated by dotted lines (see also Figs. S2A–C). (B) Touch-evoked swimming patterns in embryos injected with human SUPT16H mRNA (WT, c.1712A > G [p.N571S], or c.1943A > G [p.Q648R]) assessed at 2 dpf. The control group (con) was injected with TagRFP mRNA. Data are presented as mean values; n ≥ 45 embryos per group. (C) Bright-field images of embryos injected with SUPT16H mRNA. Head size and body length were measured at 2 dpf (see also Fig. S2D). Data are presented as mean ± SD; n ≥ 10 embryos per group. (D) Brain morphology of embryos injected with SUPT16H mRNA was assessed by anti-acetylated tubulin immunohistochemistry to visualize neuronal structures (see also Fig. S2E). Brain size was quantified at 2.5 dpf. Data are presented as mean ± SD; n ≥ 25 embryos per group. Head size, body length, and brain size were measured using ImageJ software. Statistical significance was determined using one-way ANOVA followed by Dunnett’s multiple comparison test; *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. Abbreviations: ANOVA, analysis of variance; Con, control (TagRFP mRNA-injected); CTD, C-terminal domain; DD, dimerization domain; dpf, days post-fertilization; FACT, FAcilitates Chromatin Transcription; MD, middle domain; NTD, N-terminal domain; PDB, Protein Data Bank; SD, standard deviation; SUPT16H, suppressor of Ty 16 protein; WT, wild type.

Variants p.G93S, p.I172V, p.E311D, p.E355D, and p.A400S are located in the NTD (Fig. S2A) and are unlikely to disrupt nucleosome interaction or FACT complex formation. The isoleucine at position 172 (I172) forms hydrophobic interactions with W55 and V168; substitution with a less hydrophobic valine may modestly reduce structural stability. The aliphatic side chain of E311 participates in van der Waals contacts with L312 and V329; replacement with aspartate shortens the side chain and may locally destabilize the structure. Conversely, E355 contributes to NTD stability through ionic and van der Waals interactions (with K48, Y46, and A51), but its substitution to aspartate is predicted to have minimal effect due to the similar biochemical properties of the two residues. G93 and A400 do not form critical interactions, and substitution with serine is unlikely to significantly alter folding.

Within the MD, residues I747, P810, and R847 contact histones H3 and H4 in the crystal structure of the MD–(H3–H4)₂ complex (Fig. S2B). However, these contacts are absent in the FACT–nucleosome complex due to conformational rearrangements upon nucleosome binding [17]. Thus, the p.I747R, p.P810L, and p.R847W variants may destabilize (H3–H4)₂ tetramer formation during early nucleosome assembly in the absence of DNA.

In contrast, R734 and T736, also within the MD, do not form extensive intramolecular contacts; p.R734W and p.T736A are therefore unlikely to affect structural integrity. Another MD variant, p.T828M, may compromise stability because T828 forms polar interactions with C815, Q818, and N826, which would be lost upon substitution with methionine (Fig. S2C). Similarly, p.V840A may weaken hydrophobic interactions with I843 and I859, resulting in local destabilization (Fig. S2C).

Residue Q648, situated in the loop region between the DD and MD, forms polar contacts with N703 in the MD, as observed in the crystal structure of the SPT16–H3–H4 tetramer complex (PDB ID: 4Z2M). The p.Q648R substitution introduces a longer arginine side chain that may disrupt this polar contact but is unlikely to affect overall folding due to the flexibility of the loop (Fig. S2C). Conversely, N571, located within the DD, stabilizes the domain through polar interactions, including a key contact with K596 that links to the nucleosomal DNA backbone. The p.N571S variant likely disrupts this interaction network, potentially impairing domain stability and indirectly reducing DNA-binding affinity (Fig. 3A).

SUPT16H variants exhibit differential loss of function in rescuing supt16h^−/− phenotypes

Our structural analysis suggested that the p.N571S variant may have the most severe impact on SUPT16H function by destabilizing the structural framework critical for DNA binding (Fig. 3A), whereas p.Q648R was predicted to exert a milder effect on overall protein integrity. To evaluate these predictions experimentally, we assessed the ability of human SUPT16H wild-type (WT) and variant alleles—c.1712A > G (p.N571S) and c.1943A > G (p.Q648R)—to rescue the supt16h LOF phenotype in zebrafish. mRNA encoding WT or variant SUPT16H was injected into one-cell-stage supt16h^−/− embryos, and phenotypic rescue was subsequently evaluated. Embryos injected with WT SUPT16H mRNA exhibited substantial improvement in locomotor behavior, whereas those injected with c.1712A > G (p.N571S) mRNA failed to restore touch-evoked responses or normal locomotion. In contrast, embryos expressing c.1943A > G (p.Q648R) mRNA showed partial behavioral rescue, demonstrating improved touch response and linear movement, although the effect was less pronounced than that of WT mRNA (Fig. 3B).

Both SUPT16H variants displayed reduced capacity to rescue morphological abnormalities, including decreased head size and shortened body length (Fig. 3C; Fig. S2D). Similarly, neither variant fully restored brain development, as evidenced by persistent reductions in brain size and disrupted axonal projections; however, c.1943A > G (p.Q648R) exhibited greater rescue efficiency than c.1712A > G (p.N571S) at 2.5 dpf (Fig. 3D; Fig. S2E). Collectively, these findings indicate that both p.N571S and p.Q648R represent partial or complete LOF variants relative to WT SUPT16H, with p.N571S showing the more severe functional deficit. These results are consistent with the structural predictions and underscore the functional correlation between molecular perturbations in SUPT16H and the clinical spectrum of SUPT16H-related neurodevelopmental disorders.

p53 mediates apoptosis underlying neurodevelopmental defects and facial dysmorphism

Based on our findings that supt16h is essential for cell survival in the CNS and pharyngeal arches during embryogenesis (Fig. 2G), we compared the ability of WT and variant SUPT16H to rescue apoptosis in supt16h^−/− embryos. Both SUPT16H variants—c.1712A > G (p.N571S) and c.1943A > G (p.Q648R)—were markedly less effective than WT SUPT16H in preventing apoptosis within the brain, pharyngeal arches, and spinal cord (Fig. 4A and B).

Figure 4.

Reduced activity of SUPT16H variants in rescuing p53-mediated apoptosis in the CNS and pharyngeal arches of supt16h^−/− embryos. (A) Confocal images showing acridine orange-positive apoptotic cells (green) in supt16h^−/− embryos. Apoptosis was assessed at 36 hpf in the brain (white box) and pharyngeal arches (red box) of the head (left panel) and at 2 dpf in the spinal cord (white box) of the trunk (right panel). Scale bars, 200 μm. (B) Quantification of acridine orange-positive cells in the brain, pharyngeal arches, and spinal cord of supt16h^+/− and supt16h^−/− embryos injected with wild-type SUPT16H (WT), SUPT16H variants (c.1943A > G and c.1712A > G), or TagRFP mRNA (con). Data are presented as mean ± SD; n ≥ 7 embryos per group for head regions and n ≥ 14 for trunk regions. (C) p53 mediates apoptosis in supt16h^−/− embryos. Quantification of acridine orange–positive cells in the brain, pharyngeal arches, and spinal cord at 2 dpf shows that injection of p53 morpholino (p53MO) markedly reduces apoptosis (see also Fig. S3). Data are presented as mean ± SD; n ≥ 12 embryos per group. (D) p53 knockdown alleviates craniofacial dysmorphology in supt16h^−/− embryos. Data are presented as mean ± SD; n ≥ 13 embryos per group. Head size and body length were quantified from bright-field images using ImageJ in supt16h^+/− and supt16h^−/− embryos with or without p53MO injection. Data are presented as mean ± SD; n ≥ 13 embryos per group. Statistical significance was determined using one-way ANOVA followed by Dunnett’s multiple comparison test; *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. Abbreviations: ANOVA, analysis of variance; CNS, central nervous system; con, control (TagRFP mRNA-injected); dpf, days post-fertilization; hpf, hours post-fertilization; p53MO, p53 morpholino oligonucleotide; SD, standard deviation; SUPT16H, suppressor of ty 16 protein; supt16h, suppressor of ty 16 homolog (zebrafish gene); WT, wild type.

A recent study demonstrated that supt16h LOF upregulates p53 expression, thereby disrupting hematopoietic stem and progenitor cell (HSPC) specification through repression of Notch signaling [10]. To determine whether a similar mechanism contributes to the apoptosis observed in the CNS and pharyngeal arches of supt16h^−/− embryos, we inhibited p53 translation using a morpholino antisense oligonucleotide (p53MO) injected at the one-cell stage. p53 knockdown markedly reduced apoptosis in both anterior (brain and pharyngeal arches) and posterior (spinal cord) regions (Fig. 4C; Fig. S3) and significantly improved head size in supt16h^−/− embryos (Fig. 4D; Fig. S3). These findings indicate that p53-mediated apoptosis underlies the neurodevelopmental and craniofacial abnormalities associated with supt16h LOF during early embryonic development.

supt16h loss of function disrupts craniofacial development

Clinical characteristics in individuals with SUPT16H variants—including facial dysmorphism, external-ear malformations, and hearing loss—suggest that SUPT16H LOF may perturb neural crest development. To test this hypothesis, we examined the premigratory, migratory, and postmigratory stages of neural crest development in zebrafish supt16h mutants to define the pathogenic mechanisms underlying craniofacial defects. In situ hybridization using an mRNA probe for sox10 (SRY-box transcription factor 10), a key regulator of neural crest cell and otic placode development, showed normal premigratory cranial neural crest formation in the dorsal neural tube (arrow) and otic placode (arrowhead) of supt16h^−/− embryos at the 10-somite stage (ss) (Fig. 5A). Similarly, premigratory and early migrating vagal and trunk neural crest cells appeared intact in supt16h^−/− embryos at 26 ss (Fig. 5B). However, a marked reduction in migratory neural crest cells—labeled by crestin (ctn), which is expressed specifically in migrating neural crest cells—was observed at 26 ss (Fig. 5C). Furthermore, expression of dlx2a (distal-less homeobox 2a), a transcription factor essential for differentiation of cranial neural crest cells within the pharyngeal arches, was severely reduced in supt16h^−/− embryos at 32 hours post-fertilization (hpf) (Fig. 5D). Because dlx2a is critical for proper craniofacial cartilage formation, these findings demonstrate that supt16h LOF disrupts neural-crest-derived craniofacial development.

Figure 5.

Disruption of neural crest and glial development in supt16h^−/− embryos. Whole-mount in situ hybridization (WISH) was performed using antisense RNA probes targeting neural crest– And glial-marker genes. (A) sox10-positive premigratory cranial neural crest cells at the 10-somite stage (ss) in supt16h^+/− and supt16h^−/− embryos. A, anterior; p, posterior; arrow, cranial neural crest cells; arrowhead, otic placode. Scale bar, 200 μm. (B) sox10-positive premigratory and early migrating vagal/trunk neural crest cells at 26 ss in supt16h^+/− and supt16h^−/− embryos. A, anterior; p, posterior; arrow, cranial neural crest cells; black arrowhead, otic placode; open arrowhead, trunk neural crest cells. Scale bars, 200 μm. (C) Ctn-positive migrating cranial (arrow), vagal (arrowhead), and trunk (open arrowhead) neural crest cells at 26 ss in supt16h^+/− and supt16h^−/− embryos. A, anterior; p, posterior. Scale bars, 200 μm. The graph quantifies the migration distance of vagal and trunk neural crest cells, measured from the midline to the leading edge of laterally migrating cells (n ≥ 12 embryos per group). Data are presented as mean ± SD. (D) dlx2a-positive cranial neural crest cells in the ectomesenchyme of the pharyngeal arches (arrow) at 32 hpf in supt16h^+/− and supt16h^−/− embryos (n ≥ 16 per group). A, anterior; p, posterior. (E) olig2-positive oligodendrocytes in supt16h^+/− and supt16h^−/− embryos at 54 hpf (n ≥ 15 embryos per group). Arrow, midline in the CNS; arrowhead, forebrain; asterisk, cerebellum. Scale bars: 400 μm (lateral view) and 200 μm (dorsal view). Expression of olig2 mRNA was quantified in 2 dpf supt16h^+/− and supt16h^−/− embryos injected with wild-type SUPT16H (WT), SUPT16H variants [c.1943A >G (p.N571S) or c.1712A > G (p.Q648R)], or TagRFP mRNA (con, control) by qRT-PCR (n≥ 16 per group). Data are presented as mean ± SD. (F) Mbpa-positive myelinated oligodendrocytes and Schwann cells in supt16h^+/− and supt16h^−/− embryos at 3 dpf (n > 25 per group). A, anterior; p, posterior; arrow, oligodendrocytes at the midline of the CNS; arrowhead, Schwann cells in the PNS; open arrowhead, migrating neural crest cells undergoing premature differentiation. Scale bars: 400 μm (lateral view) and 200 μm (dorsal view). Statistical significance was determined using one-way ANOVA followed by Dunnett’s multiple comparison test, or Welch’s t-test; *P < 0.05, **P < 0.01, ****P < 0.0001. Abbreviations: ANOVA, analysis of variance; CNS, central nervous system; con, control (TagRFP mRNA-injected); ctn, crestin; dpf, days post-fertilization; dlx2a, distal-less homeobox 2a; hpf, hours post-fertilization; mbpa, myelin basic protein a; PNS, peripheral nervous system; qRT-PCR, quantitative reverse-transcription polymerase chain reaction; SD, standard deviation; ss, somite stage; SUPT16H, suppressor of ty 16 protein; supt16h, suppressor of ty 16 homolog (zebrafish gene); WT, wild type.

supt16h is essential for oligodendrocyte and Schwann cell development

In supt16h^−/− embryos, mRNA expression of olig2 (oligodendrocyte transcription factor 2), which is required for oligodendrocyte specification, was markedly reduced (Fig. 5E). To assess the functional relevance of human SUPT16H variants, we compared the ability of WT and mutant SUPT16H (c.1712A > G [p.N571S] and c.1943A > G [p.Q648R]) to rescue olig2 expression in supt16h^−/− embryos. Both variants were less effective than WT SUPT16H in restoring olig2 expression (Fig. 5E). Additionally, expression of mbpa (myelin basic protein a)—which marks both oligodendrocytes and Schwann cells at the onset of myelination—was markedly reduced in all supt16h^−/− embryos (Fig. 5F). These findings indicate that supt16h is essential for oligodendrocyte development in the CNS, suggesting that supt16h LOF may underlie the hypotonia observed in individuals with SUPT16H variants. Notably, ectopic mbpa expression was detected in the trunk region of supt16h^−/− embryos at 3 dpf (2/35 embryos), indicative of premature myelination by migrating Schwann cells derived from neural crest progenitors (Fig. 5F). This ectopic expression likely represents transient mis-regulation during Schwann-cell differentiation. Together, these results demonstrate that supt16h is required for proper migration and myelination of Schwann cells in the peripheral nervous system (PNS).

Discussion

Previous studies have described eight unrelated individuals harboring de novo SUPT16H missense variants associated with neurodevelopmental disorders [4–7]. However, due to limited case numbers and incomplete phenotypic data, the prevalence and clinical spectrum of heterozygous SUPT16H variants have remained uncertain. This prompted us to investigate the broader disease spectrum and underlying mechanisms through analysis of an expanded human cohort and functional studies in zebrafish supt16h mutants.

In this study, we report 24 individuals carrying SUPT16H variants, including five with PTVs, thereby extending both the genetic and phenotypic landscape of SUPT16H-associated disorders. Previously, only one individual with a likely pathogenic PTV had been identified [15]. We now describe five additional cases; however, current evidence does not suggest a distinct phenotypic separation between individuals with PTVs and those with pathogenic missense variants. A consistent pattern of neurodevelopmental impairment accompanied by multisystem involvement was observed across the cohort. Overall, intellectual disability was the most consistent clinical feature, whereas autism features, hypotonia, and facial dysmorphism were frequent. Less common findings included congenital heart defects, endocrine abnormalities, hearing loss, and vision deficits. Incomplete availability of brain MRI scans for centralized review represents a limitation, as subtle callosal thinning below the third percentile may have been underestimated. Although not universally observed, these findings underscore the need for comprehensive physical and neurological evaluations to better define the clinical phenotype associated with SUPT16H variants.

NCCs, which arise at the neural plate borders, undergo EMT and migrate extensively to form a diverse array of derivatives—including craniofacial bones and cartilage, peripheral and enteric neurons and glia, melanocytes, cardiac tissues, and adrenal medulla chromaffin cells [20]. Given their broad developmental potential, distinct neurocristopathies arise from defects in specific NCC-derived structures or cell types. However, the genetic causes and molecular mechanisms underlying most neurocristopathies remain poorly understood. Defects in cranial NCC migration or differentiation lead to craniofacial malformations, ear anomalies, hearing loss, and neurological impairment, as exemplified by CHARGE syndrome and Waardenburg syndrome [21, 22]. The constellation of craniofacial dysmorphism, external ear anomalies, and hearing loss in patients with SUPT16H variants suggests a potential role for SUPT16H dysfunction in neural crest development and neurocristopathy. Moreover, because NCCs represent a vertebrate-specific and transient embryonic population absent in invertebrate models such as Drosophila and Caenorhabditis elegans, traditional systems are inadequate for studying these mechanisms. In contrast, zebrafish, as a vertebrate model with conserved NCC ontogeny [11], provides an ideal system to delineate the developmental functions of SUPT16H.

In our study, zebrafish supt16h mutants phenocopied the major clinical features observed in individuals carrying SUPT16H variants, validating these as representative LOF models. Notably, supt16h LOF did not impair formation of premigratory sox10-positive cranial neural crest cells but resulted in a reduction of ctn-positive migratory neural crest cells and a loss of postmigratory dlx2a-positive ectomesenchymal cells within the pharyngeal arches. Because craniofacial structures originate from cranial neural crest cells that populate the pharyngeal arches, defects in this lineage likely account for the characteristic facial dysmorphism seen in patients. Likewise, neural crest cells give rise to the cartilage, bone, and connective tissues of the ear, as well as to glial progenitors that differentiate into Schwann cells. These Schwann cells myelinate and support spinal ganglion neurons within the inner ear, and their dysfunction could underlie the auditory deficits observed in some individuals with SUPT16H variants.

Beyond craniofacial malformations, supt16h mutants also exhibited impaired oligodendrocyte development and defective myelination, providing a plausible mechanistic link to the periventricular white matter hyperintensities observed on T2-weighted MRI in affected individuals. These findings highlight an essential role for supt16h in glial specification and differentiation within the CNS, potentially explaining the central hypotonia frequently observed in patients with SUPT16H variants. Schwann cell precursors (SCPs), which arise from the neural crest, migrate along developing peripheral nerves to populate the PNS. In supt16h mutants, we detected ectopic expression of mbpa in the trunk of a subset of embryos, suggesting that supt16h normally suppresses premature differentiation of SCPs before they reach their destinations. Accordingly, supt16h LOF disrupts proper migration and differentiation of Schwann cells in the PNS.

Previous studies have shown that the Drosophila ortholog of SUPT16H, dre4, is essential for cell survival; its loss causes early lethality, and tissue-specific knockdown results in neuronal and glial loss. Expression of patient-derived SUPT16H variants (p.T171I and p.G808R) in dre4-deficient flies only partially rescued these phenotypes, supporting a conserved requirement for SUPT16H function in neural maintenance [5]. Likewise, Supt16h^+/− mice exhibit cognitive and social behavior abnormalities accompanied by reduced neural progenitor populations in the cerebral cortex and hippocampus, attributed to impaired neural stem cell proliferation and increased apoptosis through inhibition of the mitogen-activated protein kinase (MAPK) signaling pathway [9]. Although neurological manifestations have been the primary focus of SUPT16H-related disorders, the contribution of vertebrate-specific neural crest defects has not been explored previously. Our findings demonstrate that supt16h LOF impairs neural crest development, establishing a novel role for SUPT16H in vertebrate neurocristogenesis and implicating SUPT16H deficiency as a cause of neurocristopathy.

The FACT complex is a heterodimeric histone chaperone involved in nucleosome disassembly and reassembly during transcription, replication, and repair. The SUPT16H subunit of FACT interacts with the H2A–H2B histone dimer to transiently unwind DNA and permit transcriptional elongation by RNA polymerase [23]. The subunits of FACT are highly conserved across eukaryotes, and SUPT16H is highly intolerant to both LOF and missense variants [21]. Through comprehensive structural analysis, we mapped patient-derived SUPT16H variants onto the SUPT16H protein structure. The N571S variant, which likely destabilizes the dimerization domain critical for DNA binding, showed reduced capacity to rescue mutant phenotypes compared with the Q648R variant, located in a f lexible loop that affects polar interactions with N703 in the middle domain. This functional disparity parallels the clinical observations, in which patients harboring Q648R exhibit milder manifestations than those carrying N571S, suggesting that structural context modulates variant-specific functional outcomes. These findings imply that altered chromatin regulation caused by distinct SUPT16H variants contributes to the broad phenotypic spectrum observed across multiple tissues.

In zebrafish supt16h mutants, supt16h LOF triggered p53-dependent apoptosis in the CNS and pharyngeal arches, establishing a mechanistic link between SUPT16H dysfunction and microcephaly or craniofacial malformations. These results suggest that SUPT16H normally restrains p53 activation during tissue development. Notably, this differs from prior findings that supt16h LOF upregulates p53 to disrupt hematopoietic stem and progenitor cell (HSPC) specification without inducing HSPC apoptosis [10]. Such divergence points to cell-type-specific apoptotic thresholds or cross-talk with other stress-response pathways, highlighting SUPT16H’s role in modulating p53 activity across distinct developmental contexts. Future studies should delineate the p53-regulated transcriptional programs and downstream effectors that mediate these context-dependent outcomes.

A recent study by Wang et al. [9] generated a Supt16h mouse model with deletion of exons 1–8, showing that Supt16h haploinsufficiency impairs cognitive and social behavior and reduces cortical and hippocampal neurons due to decreased neural stem-cell proliferation and increased apoptosis. In contrast, we observed no significant developmental differences between supt16h^+/+ and supt16h^+/− zebrafish embryos. These findings suggest possible species-specific differences in the developmental requirement for SUPT16H, potentially arising from genetic compensation mechanisms inherent to the non-mammalian zebrafish or differential phenotypic sensitivity thresholds across species.

In summary, our study defines the clinical, genetic, and mechanistic landscape of SUPT16H-associated neurodevelopmental disorder and establishes SUPT16H as a critical regulator of chromatin integrity and neural crest development. Through integrative analysis of24 affected individuals and complementary zebrafish models, we demonstrate that SUPT16H LOF leads to neurocristopathy-associated phenotypes encompassing craniofacial anomalies, hypotonia, and white matter defects. Structural and functional analyses further reveal variant-specific disruptions in SUPT16H architecture that correlate with clinical severity, highlighting a direct link between chromatin dysregulation and developmental outcomes. Collectively, these findings broaden our understanding of SUPT16H-related pathogenesis, identify vertebrate-specific mechanisms underlying neural crest vulnerability, and provide a framework for improved molecular diagnosis and future therapeutic exploration of SUPT16H spectrum disorders.

Materials and methods

Patient recruitment

The cohort of 24 individuals with SUPT16H variants was assembled through an international collaborative network of clinical laboratories, the GeneMatcher Exchange platform [13], and screening of public variant databases including ClinVar (https://www.ncbi.nlm.nih.gov/clinvar) and DECIPHER [14] (https://decipher.sanger.ac.uk/). Individuals 13 and 18 were enrolled in the Disorders of Cerebral Development: A Phenotypic and Genetic Analysis study within the Brain Development Research Program, UCSF, as part of a broader investigation of structural brain abnormalities. The study protocol was approved by the UCSF Institutional Review Board (IRB number 10–01008), and written informed consent was obtained from parents or legal guardians. The remaining individuals were identified through the Matchmaker Exchange platform or via independent institutional collaborations. All participants (or their legal guardians) provided written informed consent for genetic testing, data sharing, and publication of anonymized results in accordance with institutional ethical guidelines. Clinical and neuroimaging data were obtained from medical record review, and all sample collections were approved by local IRBs or equivalent ethics committees.

Genetic analysis

SUPT16H variants were identified by clinical or research-based exome sequencing (ES) using Illumina sequencing platforms, including the NextSeq 500, NovaSeq 6000, NovaSeq X, and HiSeq 4000 (Illumina Inc. San Diego, CA, USA), achieving a mean target coverage of > 80× and a minimum per-site read depth > 10× for variant calling. Individuals 1, 3, 6, 7, 12, 14, 17, and 24 underwent ES through GeneDx (Gaithersburg, MD, USA). ES protocols followed previously described methods [24], and all variants were confirmed by Sanger sequencing. For all other cases, bioinformatic analyses were performed according to the Genome Analysis Toolkit (GATK) best practices workf low (v3.4). Bidirectional reads were aligned to the human reference genome (GRCh37/hg19 or GRCh38/hg38), and sequence variants were annotated according to Human Genome Variation Society (HGVS) nomenclature. Pathogenicity assessments followed American College of Medical Genetics and Genomics (ACMG) guidelines, applying population frequency thresholds (gnomAD maximum non-founder subpopulation frequency < 0.001%) and in silico prediction tools including Combined Annotation Dependent Depletion (CADD) and Rare Exome Variant Ensemble Learner (REVEL) to support deleterious effects of missense variants.

Zebrafish maintenance

Wild-type AB zebrafish (Danio rerio) were obtained from the Zebrafish International Resource Center (Eugene, OR, USA). Embryos were produced by natural spawning of adult zebrafish and maintained in E3 medium (5 mM NaCl, 0.33 mM MgSO₄, 0.33 mM CaCl₂, and 0.17 mM KCl) at 28.5°C in an incubator. Embryos were staged as hpf and dpf, as previously described [25]. Adult zebrafish were raised in recirculating balanced saltwater at 27.5°C under a 14 h light/10 h dark photoperiod. All experimental procedures were approved by the Institutional Animal Care and Use Committee (IACUC) of Kangwon National University (Chuncheon, Republic of Korea; protocol no. KW-221129-1) and were conducted in accordance with national regulations and the Animal Protection Law for Laboratory Animal Research.

Generation and verification of zebrafish supt16h^kan1/kan1 zebrafish mutants

The zebrafish supt16h mutant line (supt16h^kan1/kan1) was generated using CRISPR/Cas9-mediated mutagenesis. A single guide RNA (sgRNA; 5′-TCAGAATGAGATGACTGCTGAGG-3′) targeting exon 13 was transcribed in vitro using the HiScribe T7 Quick High Yield RNA Synthesis Kit (E2050; New England Biolabs, Ipswich, MA, USA) and purified with the Monarch RNA Cleanup Kit (T2040; New England Biolabs). A 1 nl injection mixture containing 100 ng/μl sgRNA and 3.2 μg/μl Cas9 protein (M0646T; New England Biolabs) was microinjected into one-cell-stage embryos. Founder (F₀) fish were outcrossed to wild-type AB zebrafish to establish heterozygous supt16h^+/kan1 (F₁) carriers. Mutations at the target site in F₁ embryos were verified using a T7 endonuclease I cleavage assay (Sigma-Aldrich, St. Louis, MO, USA) followed by Sanger sequencing on an ABI 3730xl DNA Analyzer (Applied Biosystems, Waltham, MA, USA). The target specificity of the supt16h^kan1/kan1 allele was confirmed by comparing its phenotype with that of the previously generated supt16^sd45/sd45 mutant obtained by ENU mutagenesis [10].

Genomic DNA for genotyping was isolated from embryos or fin clips by lysis in 50 mM NaOH at 95°C for 20 min, followed by neutralization with 1 M Tris–HCl (pH 8.0). For CRISPR mutants, the supt16h genomic fragment encompassing the sgRNA target site was amplified by PCR using forward primer 5′-AATGACGATGAAGACGAGGAC-3′ and reverse primer 5′- CCAGCTCTTTCTGATGAGTCC-3′. PCR products were resolved on a 12% polyacrylamide gel to distinguish genotypes. For ENU mutants (supt16h^+/sd45 and supt16h^sd45/sd45), amplification was performed using forward primer 5′-AATGACGATGAAGACGAGGAC-3′ and reverse primer 5′-AGTGTTCCCGACCCACATAA-3′, and genotypes were confirmed by DNA sequencing. At least 30 embryos per genotype were analyzed to ensure reproducibility of the observed phenotypes.

RNA decay analysis in supt16h^kan1/kan1 zebrafish mutants

RNA decay was assessed in supt16h^kan1/kan1 zebrafish embryos to evaluate transcript stability. Total RNA was extracted from 2 dpf embryos using TRIzol Reagent (15 596 018; Thermo Fisher Scientific, Waltham, MA, USA), followed by purification with the RNeasy MinElute Cleanup Kit (74 204; Qiagen, Hilden, Germany). First-strand cDNA was synthesized from 1 μg of total RNA using the PrimeScript First Strand cDNA Synthesis Kit (6210A; Takara Bio, Seoul, Republic of Korea) according to the manufacturer’s protocol. The supt16h transcript was amplified by PCR using forward primer 5′-AATGACGATGAAGACGAGGAC-3′ and reverse primer 5′-CCAGCTCTTTCTGATGAGTCC-3′. PCR amplification was performed under the following cycling conditions: 95°C for 1 min; 30 cycles of 95°C for 30 s, 60°C for 30 s, and 72°C for 30 s; followed by a final extension at 72°C for 5 min.

Quantitative reverse transcription polymerase chain reaction (qRT–PCR) was performed using TOPreal qPCR 2× PreMix with low ROX (RT500M; Enzynomics, Daejeon, Republic of Korea) and SYBR Green detection on a QuantStudio 1 Real-Time PCR System (Applied Biosystems, Thermo Fisher Scientific). The cycling conditions were as follows: 95°C for 10 min, followed by 40 cycles of 95°C for 10 s, 60°C for 15 s, and 72°C for 15 s, followed by dissociation-curve analysis (95°C for 15 s, 60°C for 1 min, and 95°C for 0.1 s). All reactions were performed in technical triplicate. β-actin served as the endogenous reference gene, and relative transcript abundance was calculated using the 2^−ΔΔCt method after normalization to β-actin expression. Primer sequences were as follows: β-actin forward 5′-TGTCCCTGTATGCCTCTGGT-3′ and reverse 5′-AAGTCCAGACGGAGGATG-3′; supt16h forward 5′-AATGACGATGAAGACGAGGAC-3′ and reverse 5′- CCAGCTCTTTCTGATGAGTCC-3′.

Bright-field imaging

Embryos were anesthetized in 0.02% tricaine (Sigma-Aldrich) prepared in E3 medium and mounted in 3% methylcellulose (Sigma-Aldrich). Bright-field images were captured using an Olympus stereomicroscope (Olympus, Tokyo, Japan) equipped with an AxioCam GRC camera (Carl Zeiss, Oberkochen, Germany) and processed with Zeiss Zen 3.4 Blue Edition software (Carl Zeiss). Embryo body length and brain size were quantified using ImageJ software version 1.53 k (National Institutes of Health, Bethesda, MD, USA; accessed July 6, 2021). Each experiment was independently repeated at least three times to confirm reproducibility.

Touch-evoked response assay

Touch-evoked locomotor behavior was assessed in 2 dpf zebrafish embryos. Individual embryos were placed in the center of a Petri dish containing E3 medium, and a tactile stimulus was applied to the tail using a fine white tip. The resulting swimming movement was recorded under an Olympus stereomicroscope (Olympus). Behavioral responses to tactile stimulation were categorized as [1] linear movement, [2] circular movement, or [3] no response. At least 30 embryos per genotype were analyzed to ensure reproducibility of the observed phenotypes.

Immunohistochemistry

Immunostaining was performed using an anti-acetylated α-tubulin antibody (1:1000; T7451, Sigma-Aldrich), a pan-neuronal marker, to visualize structural defects in zebrafish brain development, following previously described protocols with minor modifications [26]. Zebrafish embryos were dechorionated and fixed in Dent’s fixative (80% methanol and 20% dimethyl sulfoxide) at 4°C for 12 h. Fixed embryos were rehydrated stepwise in 75%, 50%, and 25% methanol in phosphate-buffered saline (PBS), rinsed in 1× PBS containing 0.1% Tween-20 (PBST), and permeabilized with 10 μg/ml proteinase K in PBST at 23°C for 20 min. Samples were post-fixed in 4% paraformaldehyde in PBS for 30 min and washed twice in immunofluorescence (IF) buffer (0.1% Tween-20 and 1% bovine serum albumin [BSA] in PBS) for 10 min each at 23°C. Embryos were incubated in blocking buffer (10% fetal bovine serum and 1% BSA in PBS) at 23°C for 2 h, followed by overnight incubation at 4°C with anti-acetylated α-tubulin antibody (1:1000; T7451, Sigma-Aldrich) diluted in blocking buffer. The embryos were washed twice in IF buffer for 30 min at 23°C, incubated with Alexa Fluor 488-conjugated goat anti-mouse IgG (H + L) secondary antibody (1:1000; A11001, Invitrogen, Waltham, MA, USA) for 4 h and washed twice in IF buffer for 10 min each. Fluorescent images were acquired using either an SZX16 f luorescence stereomicroscope (Olympus, Tokyo, Japan) equipped with an AxioCam GRC camera (Carl Zeiss) or a Nikon Eclipse Ti2 confocal microscope (488 nm argon laser, 520–535 nm bandpass filter; Nikon, Tokyo, Japan) at the Kangwon Center for Systems Imaging (KCSI, Chuncheon, Republic of Korea).

Apoptotic cell death assay using acridine orange staining

Apoptotic cells were visualized in vivo in live zebrafish embryos using acridine orange (A6014; Sigma-Aldrich), as described previously [27]. Embryos at 36 hpf and 2 dpf were incubated in 9 μg/ml acridine orange prepared in E3 medium for 20 min at 23°C in the dark. After staining, embryos were washed three times with E3 medium for 5 min each, anesthetized with 0.02% tricaine, and mounted in 0.8% low-melting-point agarose in 35 mm confocal dishes. Fluorescence imaging was performed using a Nikon Eclipse Ti2 confocal microscope (488 nm argon laser, 520–535 nm bandpass filter; Nikon). Acridine orange–positive apoptotic cells were manually quantified within defined regions of interest corresponding to the embryonic brain (242.08 μm × 93.16 μm), pharyngeal arches (170.80 μm × 90.62 μm), and spinal cord (517.81 μm × 125.50 μm).

Site-directed mutagenesis and cloning of SUPT16H variants

The open reading frame of the human SUPT16H gene (NM_007192) was amplified by PCR using full-length SUPT16H cDNA (RC218797; OriGene, Rockville, MD, USA) as a template and the primers 5′-GTATGGGATCCACCATGGCTGTGACTCTGGA-3′ (forward) and 5′-GTATGTCTAGACTACTTCCTCTTTTTCTTGGGG-3′ (reverse). The PCR product was cloned into the pCS2+ vector using BamHI and XbaI restriction sites. To generate the SUPT16H missense variant c.1943A > G (p.Q648R), site-directed mutagenesis was performed via primer-directed PCR using primer pairs 5′-ACATATTCGGGAAATGAAGATCTAC-3′/5′-CCAGTGAGTCTCGTTTTACAATC-3′ and 5′-TGATCAATCTAAACCGGAGTAATCC-3′/5′-ATTCCGTAGCATTTACCAGCGCAC-3′. The amplified fragment carrying the mutation was cloned into the pCS2+ vector using XhoI and SpeI restriction sites. Another construct containing the SUPT16H variant c.1712A > G (p.N571S) was generated by inserting a custom-synthesized SUPT16H cDNA fragment encompassing exons 14–16 (Bioneer, Daejeon, Republic of Korea) into the pCS2+ vector using ClaI and XhoI restriction sites. All constructs were verified by DNA sequencing.

Microinjections of SUPT16H variants and p53 morpholino oligonucleotide

Microinjections were performed at the one-cell stage in zebrafish embryos using 1 nL of capped human SUPT16H mRNA together with a morpholino antisense oligonucleotide (MO) targeting p53 (Gene Tools, Philomath, OR, USA). The p53-MO sequence was 5′-GCGCCATTGCTTTGCAAGAATTG-3′, previously validated for specificity in zebrafish [28]. Capped mRNAs were synthesized from linearized pCS2+ constructs using the mMESSAGE mMA-CHINE SP6 Transcription Kit (AM1340; Invitrogen). Injection solutions contained 200 ng/μl SUPT16H mRNA (WT, c.1712A > G [p.N571S], or c.1943A > G [p.Q648R]), 0.5 ng p53-MO, and 25 ng/μl TagRFP mRNA (control). Injections were delivered into one-cell-stage embryos using a μPump microinjector (World Precision Instruments, Sarasota, FL, USA) fitted with a thin-wall glass capillary (TW100F-4; World Precision Instruments). TagRFP mRNA served as an injection control to normalize for injection volume and variability.

Whole-mount in situ hybridization (WISH)

Whole-mount in situ hybridization (WISH) was performed to visualize the expression of zebrafish marker genes sox10, ctn, dlx2a, olig2, mbpa, and myl7, as previously described [29] with minor modifications. Plasmid DNA templates for probe synthesis were linearized with appropriate restriction enzymes, and digoxigenin (DIG)-labeled antisense RNA probes were synthesized using the DIG RNA Labeling Mix (11 277 073 910; Sigma-Aldrich). WISH was conducted on embryos and larvae at developmental stages ranging from the 10-somite stage to 3 dpf. After colorimetric detection, mRNA expression patterns visualized by purple precipitate were examined under an SZX16 stereomicroscope (Olympus). Bright-field images were captured using an AxioCam GRC camera (Carl Zeiss) and processed with ZEN 3.4 software (Carl Zeiss).

Quantitative reverse transcription polymerase chain reaction (qRT–PCR)

qRT-PCR analysis was performed as described above, with the following modifications. Total RNA was extracted from pooled zebrafish embryos at 2 dpf (four embryos per group) using TRIzol Reagent (15 596 018; Thermo Fisher Scientific). First-strand cDNA was synthesized using the PrimeScript First Strand cDNA Synthesis Kit (6210A; Takara Bio) according to the manufacturer’s instructions. Quantitative PCR was conducted using TOPreal qPCR 2× PreMix with low ROX (RT500M; Enzynomics, Daejeon, Republic of Korea) and SYBR Green detection on a QuantStudio 1 Real-Time PCR System (Applied Biosystems, Thermo Fisher Scientific). Thermo Fisher Scientific). Cycling conditions and dissociation-curve analysis were identical to those described above. All reactions were performed in technical triplicate. β-actin was used as the endogenous reference gene, and relative transcript levels were calculated using the 2^−ΔΔCt method after normalization to β-actin expression. Primer sequences were as follows: β-actin forward 5′-TGTCCCTGTATGCCTCTGGT-3′ and reverse 5′-AAGTCCAGACGGAGGATG-3′; olig2 forward 5′-ATCCGTCCAGTTGTGGCACT-3′ and reverse 5′-TGGTGGAAGCAGAGGATGGT-3′.

Statistical analyses

All statistical analyses were conducted using GraphPad Prism v10 (GraphPad Software, San Diego, CA, USA). Data are presented as mean ± standard deviation (SD). Each experiment was performed with a minimum of three independent biological replicates. Statistical significance was evaluated using one-way analysis of variance (ANOVA) followed by Dunnett’s multiple-comparison test or Welch’s t-test, as appropriate. Significance thresholds were set at *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001.

Supplementary Material

Supplementary material is available at Human Molecular Genetics online.