Decreased H3K79 acetylation and dysregulation of neurodevelopmental genes in fetal down syndrome

Abstract

Background

Down syndrome (DS), the most prevalent chromosomal disorder caused by trisomy 21, manifests intellectual disability and cognitive dysfunction. Cumulative studies confirm epigenetic pathways including DNA methylation and non-coding RNAs drive DS pathological progression. Histone post-translational modifications (PTMs) are core epigenetic regulators of fetal brain development. However, genome-wide PTM alterations and their downstream functions in fetal DS brains remain poorly characterized, leaving a key gap in revealing epigenetic mechanisms underlying DS neurodevelopmental defects. To address this, we aimed to establish the first comprehensive landscape of histone PTMs in fetal DS cortex and investigate whether specific PTM changes contribute to aberrant neurodevelopmental gene expression.

Methods

Fetal cortexs from control and DS groups were subjected to global histone modification profiling via high-performance liquid chromatography-tandem mass spectrometry (HPLC-MS/MS). We detected mono-, di-, and tri-methylation, acetylation, homocysteinylation and malonylation on all four core histones (H2A, H2B, H3, H4). Chromatin immunoprecipitation sequencing (ChIP-seq) was used to map genomic binding profiles of H3 lysine 79 acetylation (H3K79ac). Quantitative reverse transcription polymerase chain reaction (qRT-PCR) was performed to quantify mRNA levels of candidate neurodevelopmental genes.

Results

HPLC-MS/MS analysis identified 172 distinct histone PTMs in control fetal cortices and 168 PTMs in DS fetal cortical samples. Quantitative comparison of 22 quantifiable histone PTMs revealed that H3K79ac showed the most prominent reduction in DS samples, with a 34% decrease (P<0.05). Chromatin immunoprecipitation (ChIP)-seq verified specific H3K79ac occupancy at the genomic loci of three vital neurodevelopmental genes: TNFSF13B, NXPH1 and CAMK4. Correspondingly, qRT-PCR revealed aberrant transcription levels of these three genes in DS fetal cortices.

Conclusions

This study establishes the first quantitative landscape of histone PTMs in in DS fetal cortical tissues. We demonstrate that depleted H3K79ac acts as a candidate epigenetic driver of DS neuropathology by disrupting the transcription of critical neurodevelopmental genes. This work reveals a novel epigenetic mechanism and a promising therapeutic target for DS-related neurodevelopmental disorders.

Highlight box.

Key findings

• The first comprehensive atlas covering methylation, acetylation, homocysteinylation and malonylation of histone modifications is constructed in human fetal cerebral cortex. Widespread locus-specific reduction of H3K79ac is detected in Down syndrome (DS) cerebral cortices. H3K79ac enrichment at CAMK4 correlates with significantly downregulated transcription, while depleted H3K79ac at TNFSF13B and NXPH1 accompanies markedly elevated mRNA levels. Parallel reaction monitoring (PRM), Western blot and chromatin immunoprecipitation (ChIP)-quantitative polymerase chain reaction (qPCR) independently verify these chromatin modification patterns.

What is known and what is new?

• Dysregulated epigenetics and impaired neurodevelopment are well-documented hallmarks of DS, yet systematic histone post-translational modification (PTM) profiling in DS fetal cortex remains absent.

• Multi-category histone modification landscape specific to DS fetal cerebral cortex is first reported. H3K79ac acts as a gene-specific repressive epigenetic mark at neurodevelopmental gene loci, serving as a candidate disease-associated epigenetic signature.

What is the implication, and what should change now?

• Locus-specific H3K79ac depletion provides a novel epigenetic molecular basis for disrupted neurodevelopment in DS.

• Functional validation in primary neural models is required to verify causal regulatory relationships. Small molecules targeting enzymes governing H3K79ac turnover can be further investigated to rescue DS-associated neurodevelopmental defects.

Introduction

Down syndrome (DS), also known as trisomy 21 and caused by an extra copy of chromosome 21, is the most common chromosomal condition, with an incidence of approximately 1 in 600–800 live births (1,2). While trisomy 21 is the definitive genetic driver of DS, the molecular mechanisms linking this chromosomal aberration to the characteristic neurodevelopmental abnormalities—such as impaired cognitive function, delayed neuronal maturation, and structural brain defects—remain incompletely understood (3). Epigenetic regulation, particularly histone post-translational modifications (PTMs), has emerged as a critical mediator of neurodevelopment and its perturbations in neurodevelopmental disorders (4). However, how histone PTM landscapes are remodeled in the developing DS brain and whether these changes contribute to the pathogenesis of DS-related neurodevelopmental deficits have been rarely explored, representing a key gap in current research.

Histones—around which DNA is wrapped—carry a versatile repertoire of PTMs that include methylation, acetylation, phosphorylation and homocysteinylation. Collectively, these marks constitute the “histone code”, a hypothesis proposing that combinatorial epigenetic signatures modulate gene-expression output by remodeling chromatin architecture and modulating DNA accessibility (5). Notably, the distribution and abundance of histone PTMs are cell-type- and tissue-specific, and mounting evidence implicates their dynamic regulation in brain development, synaptic plasticity and the pathogenesis of neurological disorders (6-9). Večeřa et al. suggested that acetylation marks at Lys9 of histone H3 (H3K9ac) affect embryonic brain development (7). Li et al. showed that an increase in histone H3 Lys14 acetylation (H3K14ac) activates the transcription of the bHLH factors Neurogenin1 and NeuroD1, which ultimately initiate and maintain the commitment to a neuronal fate (8). Additionally, histone H3 lysine 4 trimethylation (H3K4me3) and histone H3 lysine 27 trimethylation (H3K27me3) act as molecular “switches” to dynamically regulate the proliferation and development of neural stem cells (NSCs), modulate their differentiation toward GABAergic neurons, and influence spatial learning and memory in mice (10). Beyond normal neurodevelopment, dysregulation of histone PTMs is a key driver of multiple neurodevelopmental disorders. Kabuki syndrome arises from mutations in KMT2D (which result in reduced H3K4me2/me3) and also from mutations in lysine (K)-specific demethylase 6A (KDM6A) that impair H3K27 demethylation (11,12), whereas Sotos syndrome is associated with mutations in nuclear receptor SET domain-containing protein 1 (NSD1), which lead to dysregulated methylation of H3K36 and H4K20 (13). Beyond these, dysregulation of epigenetic modifications also plays a crucial role in other neurodevelopmental disorders, including Rett syndrome, Rubinstein-Taybi syndrome, and Fragile X syndrome (14-17). These findings highlight histone PTMs as critical mediators of neural development and its perturbations.

Despite this established role in other disorders, few studies have examined histone modifications in DS, a neurodevelopmental disorder.

Using quantitative mass spectrometry (MS) and chromatin immunoprecipitation sequencing (ChIP-seq) on DS cortex (n=4), we show that H3K79ac is the most depleted histone PTM. Intriguingly, this loss coincides with increased transcription of TNFSF13B, POGK and NXPH1, whereas H3K79ac retention at CAMK4 and TSHR is associated with reduced messenger ribonucleic acid (mRNA) levels. Together, these data underscore locus-specific regulatory modes of H3K79ac in the fetal DS brain. These data deliver the first epigenomic landscape of DS neocortex and establish H3K79ac modulation as a candidate therapeutic strategy for DS-associated neurodevelopmental deficits. We present this article in accordance with the MDAR reporting checklist (available at https://tp.amegroups.com/article/view/10.21037/tp-2026-0272/rc).

Methods

Human subjects

In this study, brain tissues were collected from four DS fetuses and four healthy gestational age-matched controls aged 22–33 weeks. All fetal cerebral cortex specimens were provided by the Biobank of Peking University People’s Hospital, Beijing, China. DS fetal specimens were obtained from medically indicated termination pregnancy, and trisomy 21 was confirmed by amniocentesis and karyotyping. Fetal sex and gestational age were recorded in detail (Table 1). Fetuses with DS accompanied by other congenital malformations were excluded. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Ethics Committee of Peking University People’s Hospital (approval No. 2022PHB283-01). Written informed consent was acquired from all pregnant women and their families after they were fully informed of the research purposes before sample collection.

Table 1. Clinical manifestations of the eight samples.

No.	Source of brain tissue	Gender	Gestational weeks
1	Normal control	Male	30
2	Normal control	Female	24
3	Normal control	Female	22
4	Normal control	Female	33
5	Down syndrome	Male	33
6	Down syndrome	Female	22
7	Down syndrome	Female	22
8	Down syndrome	Female	33

Clinical manifestations of the eight samples. No significant differences between the two groups were observed in sex distribution (Fisher’s exact test, P>0.05).

Histone extraction

Core histone proteins were extracted from the brain samples using a Histone Purification Mini Kit (ActiveMotif #400026) according to the manufacturer’s instructions. Briefly, no more than 200 mg of brain tissue was completely homogenized in an ice-cold Dounce homogenizer using 400 µL extraction buffer. Protease and phosphatase inhibitors were then added immediately before tissue lysis. The tube was left for 30 min on a rotating platform at 4 ℃. Crude histones were then obtained by centrifugation (14,000 ×g for 10 min), before 5× neutralization buffer was added until the pH reached 8.0. The crude histone sample was added to the spin column each time, centrifuged, and removed in the flow-through. Histone wash buffer was added to wash the columns three times, centrifuged, and removed in the wash buffer every time. Histone elution buffer was used to elute histone proteins and the flow-through was stored at −80 ℃ before analysis.

In-solution protein digestion

Histone protein mixture extracted from fetal cerebral cortex was transferred to a 3-kDa ultrafiltration filter by adding 300 µL UA solution (8 M CH₄N₂O in 0.1 M Tris/HCl) and centrifuging (14,000 ×g at 4 ℃) three times for 40 min each. Then, 10 mM (final concentration) DTT was used to reduce disulfide bonds for 1 h at 37 ℃. Next, alkylation was carried out by adding 50 mM (final concentration) iodoacetamide for 20 min at room temperature in the dark. Histones were further dissolved in 100 µL urea solution. At a ratio of 1:40 (enzyme: total protein), sequence-grade trypsin was added and histone proteins were then digested at 37 ℃ for 16 h. Finally, 1.0% trifluoroacetic acid was added to terminate the trypsin digestion and the insoluble material was removed by centrifugation. The supernatant was collected for MS.

High-performance liquid chromatography-tandem mass spectrometry (HPLC-MS/MS)

Trypsin-digested samples (3 µg) were reconstituted in 15 µL 0.1% formic acid and then loaded onto 15-cm C18 columns using the UltiMate 3000 RSLCnano System (Dionex, Germering, Germany). After gradient elution, the peptides were ionized and analyzed using a Q Exactive HF mass spectrometer (Thermo Fisher Scientific, Bremen, Germany). The parameters were as follows: spray voltage, 2.5 kV; first-stage full-scan maximum ion injection time, 50ms; and multi-stage scan maximum ion injection time, 150 ms. Full-scan MS spectra (m/z 350–2000) were acquired in an Orbitrap at a resolution of 60,000. The 20 most intense ions were isolated for MS/MS analysis. The fragmentation mode was collision-induced dissociation (CID) with a normalized collision energy of 35%. The raw MS data were processed using Proteome Discoverer (version 2.1.0.81, Thermo Fisher Scientific) against the database of human histones downloaded from Uniprot (http://www.uniprot.org, October 2015). Peptide spectral matches were validated using a percolator based on q-values at a 1% false discovery rate (FDR). The modifications were as follows: 14.01565 Da on Arg (R) and Lys (K) for monomethylation, 28.03130 Da on Arg and Lys for dimethylation, 42.04695 Da on Lys for trimethylation, 42.01057 Da on Lys for acetylation, 174.04600 Da on Lys for homocysteinylation, and 86.00039 Da on Lys for malonylation.

Parallel reaction monitoring (PRM)

PRM raw data were analyzed by extracting the targeted peptides from the corresponding histone database. The mass inclusion list involved mass, charge, polarity, and the time from start and end. The PRM method employed an Orbitrap resolution of 30,000 (at m/z 350) and a target AGC value of 2e⁵. Precursor ions were isolated with a window of ±0.8 m/z. Each sample was analyzed in triplicate. PRM data were processed using Xcalibur Qual Browser (version 4.0.27.19; Thermo Fisher Scientific) and validated with Skyline (version 3.5.0.9319). In the Xcalibur Qual Browser, the determination of the area under the curve (AUC) of selected fragment ions was based on the presence of product ion signals within ±2.5 min of the expected retention time, with a mass error within ±5 ppm. Skyline used raw files as input to generate and extract the normalized area of modified peptides at a 0.05 m/z ion match tolerance for each PRM spectrum.

Western blot

Histone mixtures (10 µg) were separated by SDS-PAGE on 15% gels and transferred to polyvinylidene difluoride membranes (Millipore, MA, USA). For immunoblotting of H3K79me1 (ab3594, Abcam) and H3K79me2 (ab2886, Abcam), membranes were blocked with 5% non-fat milk and then incubated with the respective primary antibodies (1:1,000). Separate membranes were blocked with 5% BSA for detection of H3K79ac (1:1,000; 39565, ActiveMotif). Peroxidase-labeled polyclonal goat anti-rabbit IgG (ZSGB-BIO, ZB-2301) was used as secondary antibody at a 1:4,000 dilution. The H3 non-modified antibody from Abcam was used as loading control antibody at a dilution of 1:2,000.

Band densitometry was performed following standard histone Western blot quantitative protocols using ImageJ software. Briefly, background-subtracted integrated density values of each modification band were normalized to the background-corrected signal of total H3 within the identical lane to correct for variations in protein loading and membrane transfer efficiency. Relative expression ratios of H3K79 modifications were calculated, and only normalized values were used for subsequent statistical comparisons.

ChIP-seq

ChIP analysis was performed using a SimpleChIP^® Enzymatic Chromatin IP Kit (CST #9005) according to the manufacturer’s protocol. Approximately 200 mg of tissue was used to obtain formaldehyde crosslinked chromatin. The chromatin was sonicated to generate DNA fragments ranging from 150 to 500 bp and then immunoprecipitated overnight at 4 ℃ with an antibody against H3K79ac. After de-crosslinking and elution, the purified immunoprecipitated DNA was sequenced. In-depth whole-genome DNA sequencing was performed by SHBIO (http://www.shbio.com, SHBIO, Shanghai, China). With alignment to the Homo sapiens reference genome (UCSC hg19), Bowtie (version 0.12.8) was used to analyze the raw sequence image data examined by the Illumina analysis pipeline, and peak calling was performed using MACS2 (version 2.2.7.1). After filtering using the control input, enriched binding peaks were generated. Differential peak analysis was conducted using the R package DiffBind, with P values adjusted for multiple testing

ChIP-quantitative polymerase chain reaction (qPCR)

ChIP was performed with the SimpleChIP^® Enzymatic Chromatin IP Kit (CST #9002) following the manufacturer’s instructions. Briefly, 100 mg of snap-frozen brain tissue was cross-linked with 1% formaldehyde, quenched, and fragmented by micrococcal nuclease to 150–500 bp. After overnight immunoprecipitation with 10 µg anti-H3K79ac (ActiveMotif #39565) or control IgG (CST #2729), DNA was recovered and quantified by SYBR Green qPCR with gene-specific primers (Table S1). Enrichment was calculated as percentage of input: %input = 2^(Ct_input – Ct_IP) * (dilution factor of input) *100. Relative enrichment (DS vs. control) = %input of DS/%input of controls. Ct=threshold cycle of the PCR reaction. Data are mean ± SD.

Quantitative reverse transcription polymerase chain reaction (qRT-PCR)

Total RNA was extracted with the RNeasy Mini Kit (QIAGEN, #74106) and reverse-transcribed using the RevertAid First-Strand cDNA Synthesis Kit (Thermo Fisher, #K1622). qRT-PCR was performed with SYBR Green Real-time PCR Master Mix (TOYOBO). Cycling condition: 95 ℃ 1 min → 40 cycles of 95 ℃ 15 s, 56 ℃ 15 s, 72 ℃ 45 s (fluorescence acquired at 72 ℃). Relative expression was calculated by the 2^−ΔΔCt method using GAPDH as endogenous control. The primers are shown in Table S1.

Statistical analysis

The statistical parameters for each experiment are reported in the corresponding figure and table legends. Sex distribution was compared using Fisher’s exact test. Quantitative data are presented as mean ± standard deviation (SD). P values were calculated with unpaired two-tailed Student’s t-tests and Benjamini-Hochberg correction was applied to all PTM comparisons to control for multiple testing; statistical significance was set at P<0.05 (raw) or FDR-adjusted P<0.05.

Results

Qualitative mapping of a high-confidence histone PTM atlas in human fetal cerebral cortices via HPLC-MS/MS

To explore the distribution of histone PTMs in the fetal cerebral cortex, histones were extracted from cerebral cortices of four DS fetal specimens and four healthy control fetal specimens. Four core histones (H2A, H2B, H3, and H4) were recognized by HPLC-MS/MS. Their sequence coverages were better than 51% and the lowest coefficient of variation of the histones was 5%, indicating that the proteome system was stable (Figure 1A). Histone PTMs were profiled by HPLC-MS/MS. A representative H3K79ac spectrum shows contiguous b- and y-ion series that unambiguously localize the acetylation to Lys79 (Figure S1).

Figure 1.

Identification of histone PTM sites in healthy and DS fetal cerebral cortices (n=4 per group). (A) Sequence coverage of four histones (H2A, H2B, H3, and H4) in healthy and DS samples. (B) Summary of the identified PTM sites in healthy and DS fetal cerebral cortices. (C) Distribution of histone modification types (mono-/di-/trimethylation, acetylation, homocysteinylation, malonylation) in healthy controls and DS. (D) Schematic of histone modifications (mono-/di-/trimethylation, acetylation, homocysteinylation, and malonylation) detected by HPLC-MS/MS in fetal cerebral cortices. In the diagram, K and R denote lysine and arginine, respectively, and the adjacent number indicates the residue position on the histone. Post-translational modification sites on core histones (H2A, H2B, H3, H4) are represented by distinct symbols: blue diamonds (methylation), yellow hexagons (acetylation), purple octagons (homocysteinylation), and green circles (malonylation). DS, Down syndrome; HPLC, high-performance liquid chromatography; MS, mass spectrometry; PTM, post-translational modification.

Differential distribution of histone PTM sites between DS and healthy control fetal cerebral cortices

Using HPLC-MS/MS, we identified 172 and 168 core-histone PTM sites in cerebral cortices from healthy control and DS fetal samples, respectively (Figure 1B,1C). The proportion of each modification among the six classes was comparable: mono-meth 29.7% vs. 31.0%, di-meth 22.1% vs. 22.0%, tri-meth 7.6% vs. 9.5%, acetyl 15.7% vs. 15.5%, Hcy 14.0% vs. 12.5%, malonyl 11.0% vs. 9.5%. Modified sites were distributed across four core histones: H2A (56 in control vs. 52 in DS), H2B (38 vs. 37), H3 (25 vs. 26) and H4 (53 vs. 53). Qualitative screening of histone PTMs in cerebral cortices from DS versus control fetal specimens revealed that DS samples lacked 18 modifications: monomethylation of H2AK15, H2AR81 and H3K64; dimethylation of H2AR17, H2AR88, H2BR33, H2BK34 and H2BK116; trimethylation of H2BK46 and H4K31; homocysteinylation of H2AK74, H2BK108, H2BK116, H3K56 and H3K64; malonylation of H2AK119, H2BK46 and H3K64; and acetylation of H2AK13. Notably, these absent modifications were detected in a subset of healthy controls (1/4 or 2/4 cases). Conversely, 17 histone-modification sites were present in ≤2/4 DS samples but absent in all controls: monomethylation of H2AK74, H2AK124, H2BK34 and H2BK116; dimethylation of H2AK74, H2AK95, H2AK99 and H4K77; trimethylation of H2AK74, H2AK95, H2AK118, H2BK34 and H4K79; and homocysteinylation of H2BK43, H3K18 and H3K122. These differentially modified residues are indicated in red in Figure 1D. Case-by-case analysis of the fetal cerebral cortex samples from four DS and four control fetuses identified no group-specific differentially modified sites (Table S2). Nevertheless, site frequencies differed markedly: H3K115me2 and H3R116me1 were present in 75% of DS cases versus 25% of controls; H2AK118ac and H3K119ac were found in 50% of DS samples but 100% of controls; H4K31ac showed the opposite trend (25% DS vs. 75% controls); and H2AK36ac was detected in 100% of DS but only 25% of controls (additional histone PTM data are provided in Table 2).

Table 2. Occurrences of histone modifications in healthy and DS fetal cerebral cortices.

Histone	PTM sites	Modification	Control†	DS†	Control					DS
					1	2	3	4		5	6	7	8
H2A	K13	Hcy	50%	0%	○	●	○	●		○	○	○	○
	K15	Hcy	75%	25%	○	●	●	●		○	●	○	○
H2A	R17	me	100%	50%	●	●	●	●		●	○	●	○
		me2	50%	0%	●	○	○	●		○	○	○	○
	K36	ac	50%	100%	○	●	●	○		●	●	●	●
	K74	me	0%	50%	○	○	○	○		●	●	○	○
	K95	me	75%	25%	●	●	○	●		○	●	○	○
		me2	0%	50%	○	○	○	○		●	●	○	○
		Hcy	25%	75%	○	○	●	○		○	●	●	●
	K99	me2	0%	50%	○	○	○	○		●	●	○	○
		me3	25%	75%	○	●	○	○		●	●	●	○
	K118	me2	75%	25%	○	●	●	●		○	●	○	○
		ac	100%	50%	●	●	●	●		○	●	●	○
		Hcy	100%	50%	●	●	●	●		○	●	●	○
		mal	100%	25%	●	●	●	●		○	○	●	○
	K119	ac	100%	50%	●	●	●	●		○	●	●	○
		Hcy	75%	25%	●	●	○	●		○	○	●	○
H2B	K34	me3	0%	50%	○	○	○	○		●	●	○	○
	K43	Hcy	0%	50%	○	○	○	○		●	●	○	○
H3	R63	me2	75%	25%	●	●	●	○		○	●	○	○
	K115	me2	25%	75%	○	○	●	○		●	●	●	○
	R116	me	25%	75%	○	○	●	○		●	●	●	○
H4	K31	ac	75%	25%	○	●	●	●		○	○	●	○
	K77	me2	0%	50%	○	○	○	○		○	●	●	○

^†, show the incidence of PTMs in healthy and DS groups (n=4 per group). ○, no modification detected at this site; ●, modification present at this site. ac, acetylation; DS, Down syndrome; Hcy, homocysteinylation; mal, malonylation; me, monomethylation; me2, dimethylation; me3, trimethylation; PTM, post-translational modification.

Quantitative PRM and Western blot validation reveal locus-specific downregulation of H3K79 acetylation in DS fetal cerebral cortices

While we did not identify any histone modification sites that showed an “all-or-none” pattern between DS and control samples, quantitative differences could still exist. We therefore applied PRM and detected significant inter-group variations in the abundance of multiple PTMs (Figure 2A). Relative to healthy controls, H3K79ac was significantly lower in DS samples, as were H3K79me, H3K79me2 and H2AK95ac. Conversely, H2AK99me, H2AK118ac and H4K59me were elevated in DS, whereas H4K31ac and H2AK36me showed no between-group differences (Figure S2, Table S3).

Figure 2.

Quantification of histone PTMs using PRM and western blot. (A) Targeted MS analysis of histone PTMs in fetal cerebral cortices. PRM was used to quantify relative PTM levels in healthy control and DS samples. Data are presented as mean ± SD (n=4 biological replicates). Unpaired two-tailed Student’s t-test was performed for statistical analysis. Raw P values were adjusted for multiple comparisons using the Benjamini-Hochberg FDR method. Asterisks indicate raw P<0.05, and statistical significance was defined as FDR-adjusted P<0.05. All FDR-adjusted P values are summarized in Table S3. (B) Western blot validation of H3K79 modifications. Upper panel: Representative blots of H3K79 acetylation (ac), monomethylation (me) and dimethylation (me2) in control and DS groups. Lower panel: quantitative analysis of band intensities. Individual data points are displayed in scatter plots, with results shown as mean ± SD. Quantitative values: H3K79ac (control: 0.44±0.08, DS: 0.31±0.06); H3K79me (control: 0.46±0.14, DS: 0.37±0.08); H3K79me2 (control: 0.46±0.15, DS: 0.39±0.05). Statistical differences were evaluated using unpaired two-tailed Student’s t-test. *, P<0.05. DS, Down syndrome; FDR, false discovery rate; MS, mass spectrometry; PRM, parallel reaction monitoring; PTM, post-translational modification; SD, standard deviation.

Given the marked reduction of H3K79ac and the established roles of both acetylation and methylation in neurodevelopment, we quantified these modifications by Western blotting. H3K79ac, H3K79me and H3K79me2 were clearly detected in all sample; however, each was consistently lower in DS cortex, with the reduction in H3K79ac reaching statistical significance (Figure 2B). These quantitative differences point to a DS-specific chromatin landscape that may alter gene-expression programs during fetal cerebral cortex development.

H3K79ac ChIP-seq identifies differentially targeted genes and their functional enrichment [Gene Ontology (GO)/Kyoto Encyclopedia of Genes and Genomes (KEGG)] in fetal DS cortex

H3K79ac ChIP-seq identified 386 high-confidence target genes (Figure 3A, available online: https://cdn.amegroups.cn/static/public/tp-2026-0272-1.xlsx). Applying a |log₂ fold change (FC)| >1 threshold (DS vs. control) revealed 70 genes with gained (red) and 107 genes with lost (blue) H3K79ac occupancy, as illustrated in Figure 3A. Applying a more stringent filtering (|log₂FC| >2, P<0.05) yielded 19 protein-coding loci that showed the largest differential occupancy (8 up, 11 down), which are displayed in Figure 3B. To specifically interrogate the link between H3K79ac and neurodevelopment, we performed enrichment analyses on this 177-gene subset (70 with increased H3K79ac enrichment, 107 with decreased H3K79ac enrichment). GO enrichment analysis revealed significant enrichment in the terms “nervous system development” (GO:0007399; 26 genes, FDR <0.0001) and “homophilic cell adhesion via plasma membrane adhesion molecules” (GO:0007156; 22 genes, FDR <0.0001) (Figure 3C,3D). Additional biological process (BP) terms were enriched in neurodevelopment-related processes, including neuron migration, cerebral cortex development, cerebral cortex cell migration, cerebral cortex radial glia-guided migration, and telencephalon glial cell migration; signal transduction processes such as calcium ion import and calcium import into the mitochondrion were also enriched (Figure 3D). KEGG pathway mapping assigned these genes to two major categories: Environmental Information Processing (e.g., hsa04020: calcium signaling) and Organismal Systems (e.g., hsa04360: axon guidance) (Figure 3E). Notably, the top three ranked KEGG pathways were Ubiquitin-mediated proteolysis (hsa04120), Calcium signaling pathway, and Axon guidance pathway (Figure 3F). Collectively, these findings implicate H3K79ac as a key regulator of the transcriptional program of axon-guidance and calcium-dependent genes—genes critical for neural circuit formation.

Figure 3.

Differential H3K79ac occupancy and functional annotation of target genes in DS fetal cerebral cortices. (A) Volcano plot of H3K79ac ChIP-seq signal at gene bodies (DS vs. control). Genes with a fold change greater than 2-fold (|log₂FC| >1) in DS versus control samples are highlighted in red (up) and blue (down) (n=2 per group). (B) Heatmap of H3K79ac enrichment for significantly altered protein-coding genes. Shown are the 19 genes that met the thresholds of |log₂FC| >2 and P<0.05. The color scale represents row Z-score normalized ChIP-seq signals. (C,D) GO and (E,F) KEGG enrichment analyses of genes with differential H3K79ac occupancy. Analyses were performed on 177 genes with |log₂FC| >1 in DS fetal cerebral cortex tissues compared to controls. (D) Top 20 enriched GO biological-process terms, ranked by adjusted P value. (F) Top 15 enriched KEGG pathways, ranked by adjusted P value. BP, biological process; CC, cellular component; ChIP-seq, chromatin immunoprecipitation sequencing; DS, Down syndrome; FC, fold change; GO, Gene Ontology; KEGG, Kyoto Encyclopedia of Genes and Genomes; MF, molecular function.

Differential H3K79ac occupancy is correlated with gene-specific expression changes in DS

Guided by prior reports and our GO/KEGG annotation, we selected CAMK4, TNFSF13B, POGK, NXPH1 and TSHR as neurodevelopment-related candidates and asked whether their expression is influenced by local H3K79ac. ChIP-seq detected prominent H3K79ac enrichment at CAMK4 and TSHR loci in DS cerebral cortices, whereas significant loss of this mark was observed at TNFSF13B, POGK and NXPH1, consistent with ChIP-qPCR validation (Figure 4A,4B). Genes gaining H3K79ac showed overall reduced expression tendencies: CAMK4 exhibited significantly downregulated mRNA levels in DS, while TSHR only presented a non-significant decreasing trend. Among genes with depleted H3K79ac, TNFSF13B and NXPH1 had markedly elevated transcript abundance in DS, whereas POGK merely trended upward without statistical significance (Figure 4C). Thus, within this gene set, elevated H3K79ac tends to correlate with transcriptional repression, while diminished H3K79ac coincides with transcriptional activation. These observations implicate H3K79ac as a potential negative regulator of expression at these loci in DS fetal cerebral cortices.

Figure 4.

H3K79ac regulates the transcription of neurodevelopment-associated genes. (A) Representative H3K79ac ChIP-seq profiles. Genome browser tracks display H3K79ac occupancy at TNFSF13B, POGK, NXPH1, CAMK4, and TSHR loci in control and DS fetal cortex samples. Data are presented as mean ± SD of normalized read coverage. (B,C) Validation of H3K79ac occupancy and its transcriptional effect. (B) ChIP-qPCR verification of H3K79ac enrichment at indicated gene regions. (C) Relative mRNA expression levels of target genes detected by qRT-PCR. Experiments were performed using identical samples (n=4 per group). Data are shown as mean ± SD relative to the control group. Statistical significance was assessed by unpaired two-tailed Student’s t-test (*P<0.05, **P<0.01, ***P<0.001, ****P<0.0001). ChIP-seq, chromatin immunoprecipitation sequencing; DS, Down syndrome; qPCR, quantitative polymerase chain reaction; qRT-PCR, quantitative reverse transcription polymerase chain reaction; SD, standard deviation.

Discussion

This study provides the first comprehensive atlas of histone PTMs—including methylation, acetylation, homocysteinylation, and malonylation—in the human fetal cerebral cortex and links their dysregulation to DS. Across 24 histone PTM sites exhibiting ≥50% difference in detection frequency between DS and control samples (Table 2), quantitative analyses identified H3K79ac as significantly reduced in DS cortices. Functional annotation (GO/KEGG) revealed that genes targeted by H3K79ac are enriched for neurodevelopmental pathways. Integrative ChIP-seq and transcriptomic data are consistent with H3K79ac acting as a transcriptional repressor at these loci; however, remains a working hypothesis pending functional validation. Collectively, our results provide a novel epigenetic framework for understanding DS pathogenesis and a valuable resource for future therapeutic targeting of histone modifications in DS.

Metabolic and enzymatic drivers of histone hypo-acetylation in DS

Accumulating studies have demonstrated mitochondrial dysfunction, disrupted glucose metabolism, and widespread derangement of intermediary metabolism in DS (18-21). In the brain, glycolytically derived pyruvate is converted to acetyl-CoA within mitochondria by the rate-limiting pyruvate dehydrogenase complex (PDHC) (22). Furthermore, Sorbi et al. reported significantly lower total and activated PDHC activities in fibroblasts from four DS patients (23). Consequently, acetyl-CoA is postulated to be reduced in DS, although absolute tissue levels have yet to be quantified. Additionally, overexpression of Hsa21 genes tilts the acetylation-deacetylation balance. The critical kinase DYRK1A, encoded on chromosome 21, phosphorylates and activates the NAD⁺-dependent deacetylase SIRT1, lowering H3K9ac and H4K16ac levels (24). Likewise, the Hsa21 gene product RIP140 recruits HDAC1/3/4/5 and potentiates their catalytic activity (25-27). These dual mechanisms—diminished acetyl-CoA supply and enhanced deacetylase activity—establish a testable metabolic–epigenetic axis linking perturbed acetyl-CoA metabolism to the histone acetylation deficit in DS. Ongoing work aims to quantify intracellular acetyl-CoA pools and systematically map the expression and activity of acetyl-transferases versus deacetylases to validate this hypothesis.

H3K79 and its modifications: chromatin plasticity at a non-tail hotspot governing cell cycle, genome stability and nervous system development

Unlike most histone marks confined to the N-terminal tail, Lys79 lies in the globular core of H3 adjacent to the DNA entry-exit point, where it remains accessible to modifying and effector proteins (e.g., writers and readers) (28,29). This residue can be methylated, acetylated or homocysteinylated, conferring diverse regulatory outputs: H3K79me controls gene expression, telomere silencing, checkpoint activation and DNA repair (30-34); H3K79me2 drives murine corticogenesis via Tbr1 and is depleted in human spina bifida (35,36); aberrant H3K79 homocysteinylation represses Cecr2, Smarca4 and Dnmt3b, thereby predisposing to neural-tube defects (37). Collectively, H3K79 modifications modulate chromatin architecture and transcription-factor binding, thereby influencing cell-cycle progression and nervous-system development (38,39).

H3K79ac loss in DS fetal cerebral cortices: a candidate epigenetic signature linked to neurodevelopmental gene dysregulation

Under the filtering criteria of |log₂FC| >1, we identified 177 H3K79ac target genes exhibiting robust between-group differences. These genes were significantly enriched in KEGG pathways critical for neurodevelopment, including the Calcium Signaling Pathway (hsa04020) and the Axon Guidance Pathway (hsa04360)—both of which are essential for neural circuit formation. Applying more stringent criteria (|log₂FC| >2 and P<0.05), we further obtained 19 high-confidence H3K79ac target genes that were specifically enriched in the axon guidance pathway (hsa04360) and the cAMP signaling pathway (hsa04024). This hierarchical enrichment pattern strengthens the association between H3K79ac dysregulation and early brain wiring during neurodevelopment.

Genes exhibiting aberrant H3K79ac occupancy, such as CAMK4 and TNFSF13B, show reciprocal transcriptional changes and are implicated in key neurobiological processes and neurological disorders. CAMK4 is a documented risk gene for Alzheimer’s disease, fragile X syndrome, autism and DS (40-43). Camk4-/- mice show blunted CREB phosphorylation and defective Ca²⁺/CREB-dependent transcription, resulting in impaired hippocampal LTP and late-phase long-term depression in cerebellar Purkinje neurons (44,45). Notably, arsenic-induced CAMK4 suppression also causes learning and memory deficits (46). Thus, activity-dependent CAMK4 signalling is essential for long-term memory retention, and its H3K79ac-associated down-regulation in DS may contribute to the cognitive phenotype. Similarly, elevated TNFSF13B/BAFF may also contribute to DS pathology. This TNF-family cytokine is produced by human astrocytes and is markedly up-regulated in activated astrocytes of multiple-sclerosis lesions (47). Although BAFF-BAFFR signalling can promote neuronal survival and microglia-neuron communication after ischemia, chronic overexpression is deleterious: BAFF-transgenic mice display brain inflammation, anxiety-like behaviour, impaired adult hippocampal neurogenesis, and defective LTP in the dentate gyrus (48-50). Moreover, BAFF acts as a potent inhibitor of neurite outgrowth, inducing axonal collapse via the Nogo-66 receptor in avian and rodent sensory neurons (51). Consequently, the H3K79ac-associated increase in TNFSF13B observed in fetal DS cortex could link early epigenetic dysregulation to inflammation-mediated neurodevelopmental deficits.

Reduced histone acetylation disrupts neural stem-cell differentiation, neuronal migration, dendritic spine morphogenesis and synapse formation, and is implicated in fragile X, Rubinstein-Taybi and Rett syndromes, underscoring the pathogenic importance of acetylation defects (7,16,52). Although we demonstrate reduced H3K79 acetylation in fetal DS cortex, its direct functional role in neurodevelopmental impairment remains to be established. Ongoing work will manipulate H3K79ac in neuronal cultures from a DS mouse model [e.g., Ts65Dn/Dp(16)1Yey] and euploid littermates to test its causal impact on differentiation, synaptogenesis, and cognition. Moreover, given that Sir2 deacetylates H3K79ac in yeast and romidepsin elevates this mark in human cell lines (53,54), the pharmacological modulation of H3K79ac represents a feasible therapeutic strategy.

H3K79ac: a candidate transcriptional repressor in human fetal cerebral cortex

Additionally, our results suggest that H3K79ac may act as a locus-specific negative regulator of transcription. Specifically, diminished H3K79ac occupancy at TNFSF13B, POGK and NXPH1 tended to accompany elevated transcription, whereas increased H3K79ac at CAMK4 and TSHR coincided with reduced mRNA abundance. These observations support a context-dependent, gene-specific repressive role of H3K79ac within DS fetal cerebral cortices, though several loci only displayed directional trends without statistical significance. This interpretation aligns with accumulating evidence implicates H3K79ac in transcriptional repression. Garcia et al. detected H3K79ac on silent cell-cycle promoters, and glucose-deprived hepatocytes exhibit H3K79ac accumulation at gluconeogenic loci that recruits HDAC3, erases neighboring H3K27ac, and silences transcription (39). Additionally, H3K79ac sterically hinders H3K79 methylation—a mark normally associated with transcriptional elongation—and has been observed on repressed genes (54,55). Collectively, these findings establish H3K79ac as a dual-function platform that can recruit repressive complexes and interfere with positive histone signals, providing a mechanistic rationale for its gene-specific suppressive activity.

The repressive effect of H3K79 acetylation may stem from its location within the H3 globular core, where it helps form the globular domain (29). Moreover, the functional output of histone PTM is dictated by its chemical nature. For instance, acetylation of H3K36, H3K79 and H2BK120 coexist with transcriptional repression, whereas methylation or monoubiquitination of the same residues correlates with activation. H3K79 is also subject to non-enzymatic homocysteinylation, a mark that silences neural-tube closure genes including Cecr2, Smarca4 and Dnmt3b (37). Conversely, H3K79me2 is an activation signature, and enzymatic succinylation of H3K79 by the succinyltransferase KAT2A up-regulates PIK3R1, JUN and PRKDC (56). Together, these data establish that the functional output of a histone PTM depends on both its specific site and the chemical properties of the modifying group. Key questions remain: How does H3K79ac enforce repression, and does it specifically shape the expression of DS-related genes? The underlying mechanisms demand direct experimental investigation.

Study limitations and future directions

This study has several limitations. First, the sample size is small (n=4 per group) because of the extreme rarity of human fetal post-mortem tissue and stringent ethical constraints, which limits statistical power and the generalizability. Second, we have not validated our observations in complementary models such as neural progenitor cell cultures or established DS mouse strains (e.g., Ts65Dn). Finally, our analysis focused primarily on H3K79ac; the potential contribution of the other histone PTMs detected here to DS-associated epigenetic perturbations remains largely unexplored.

To address these limitations and advance understanding of histone-PTM dysregulation in DS, we are actively expanding our cohort and establishing complementary in vitro and in vivo experimental models. Our key findings will be confirmed in an expanded human fetal cohort and cross-validated in cellular and animal models to ensure the robustness and cross-system consistency of H3K79ac-mediated gene regulation. Follow-up studies will systematically dissect how altered H3K79ac levels modulate neurodevelopmental phenotypes and will incorporate rescue experiments using small-molecule modulators that target H3K79-specific acetyltransferases or deacetylases. Ultimately, determining whether H3K79ac constitutes an actionable epigenetic target for ameliorating cognitive deficits in DS will require the integration of multi- omics datasets and with rigorous functional validation across models.

Conclusions

Collectively, this study constructs the first multi-type histone PTM atlas in human fetal cerebral cortex and identifies prominent locus-specific H3K79ac depletion in DS samples. H3K79ac correlates inversely with transcription of key neurodevelopmental genes at CAMK4, TNFSF13B and NXPH1, suggesting potential gene-specific repressive chromatin function. This work proposes H3K79ac as a candidate epigenetic signature linked to DS neurodevelopmental defects and offers a resource for follow-up mechanistic and therapeutic research.

Supplementary

The article’s supplementary files as

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Ethics Committee of Peking University People’s Hospital (approval No. 2022PHB283-01). Written informed consent was acquired from all pregnant women and their families after they were fully informed of the research purposes before sample collection.

Data Sharing Statement

Available at https://tp.amegroups.com/article/view/10.21037/tp-2026-0272/dss

tp-15-08-310-dss.pdf ^{(66.9KB, pdf)}

DOI: 10.21037/tp-2026-0272