This article is mentioned in:

Introduction

Abnormal alterations in Wnt genes are associated with neural tube defects (NTDs) (1,2). NTDs occur when the neural tube fails to close at the early stage of neural development, which can result in spina bifida, myelomeningocele (also known as spina bifida aperta), exencephaly or craniorachischisis (3). A previous study demonstrated that a null mutation of Wnt3a, a canonical Wnt, led to spina bifida aperta in mice (4). Wnt5a−/−, a non-canonical Wnt, mice did not suffer from NTDs; however, mice with double deletions of the Wnt5a−/− and Ltap/Vangl2 genes exhibited craniorachischisis (5). Furthermore, knockout mutations of the canonical Wnt3 or non-canonical Wnt genes, Wnt5a and Wnt7a, induce deformities of the anteroposterior (A/P) axis of the neural tube (6). These findings suggest that, in addition to roles of several known Wnt genes in the etiology of NTDs, other Wnt members may be involved in the pathogenesis of NTDs.

Wnt2b is a canonical Wnt, which is expressed in the primitive streak at embryonic day (E) 7–7.75, followed by the dorsal midline of the telencephalon and mesencephalon at E8.5–9.5, and in the cortical hem at E11.5–17.5 in murine embryos. This specific spatio-temporal expression of Wnt2b suggests that it functions in gastrulation, neurulation and cerebral cortex patterning (7,8). In addition, Wnt7b, which is associated with canonical and non-canonical Wnt signaling, is expressed at E5.5, and is expressed in the forebrain, and the ventral and intermediate spinal cord at the developmental stages during which patterning and neural specification occur (9). Notably, Wnt7b can assist in ensuring correct A/P guidance, and is involved in dorsoventral patterning of the amphibian embryonic neural tube (10). Although Wnt2b and Wnt7b mutations have not induced NTDs in mice (6), they are essential during neural tube patterning; therefore, the present study hypothesized that the expression levels of these two genes are likely to be disrupted in NTDs.

Epigenetic modifications typically regulate gene expression; histone 3 lysine 4 acetylation (H3K4ac) and trimethylation of H3 lysine 4 (H3K4me3), are commonly located in active promoters and are associated with gene activation, whereas H3K27me3 suppresses transcription (11). DNA methylation is predominantly catalyzed by DNA methyltransferase through the addition of a methyl group to cytosine residues on dinucleotide CpG islands, and is frequently distributed in the core promoter close to the transcription start site (TSS); once the promoter CpG islands are methylated, these genes are often transcriptionally inactivated (12). Our previous studies demonstrated that epigenetic modifications were significantly altered in patients with NTDs (13,14); however, the detailed mechanism and target genes remain to be elucidated.

The present study aimed to investigate dysregulation of epigenetic modification of Wnt genes in NTDs, using a mouse model of spina bifida. It was hypothesized that in NTDs, dysregulation of epigenetic modification is involved in the regulation of Wnt2b and Wnt7b gene expression.

Materials and methods

Animals

All animals experiments were performed in accordance with the standards of the Capital Institute of Pediatrics Ethics Committee (Beijing, China; permit no. SYXK 2008-0011). A total of 40 adult C57BL/6J mice (20 male and 20 female) were housed in light- and temperature-controlled rooms (12/12 h 1ight/dark cycle; 23±1°C) and maintained on pure mouse feed and tap water provided ad libitum. Adult virgin females (above 9 weeks of age and 21 g in weight) were mated for 2 h with males (1 male per 2 females, 8:00–10:00 am) of the same stock. A vaginal plug observed at 10:00 AM was considered as embryonic day 0 (E0). The mice were administered with a 20 mg/kg body weight dose of retinoic acid (RA; suspended in olive oil; Sigma-Aldrich, St. Louis, MO, USA) on E8.0–0 h, E8.0–6 h and E8.0–12 h. Mice treated with isovolumetric pure olive oil (Agric, Crete, Greece) at these same time points served as control groups, as described in our previous study (15). E18.0 fetuses were delivered by cesarean section, and 21 cases of spina bifida were selected (Fig. 1A and B). As the spinal cord of individual mice provides insufficient material for RNA extraction, the mice were randomly divided into three groups and the samples in each group were pooled. A total of 10 normal spinal cords were randomly pooled to create three control groups, two of which contained 3 mice and one contained 4 mice.

Figure 1 (A) RA-induced embryonic day 18.0 mouse embryos with typical exencephaly (top arrow) and spina bifida (bottom arrow), and (B) control mice. (C) Relative mRNA expression levels of Wnt2b and Wnt7b were significantly upregulated in the RA-induced mouse fetuses with NTDs, compared with the control group (*P<0.05). Data are presented as the mean ± standard error. RA, retinoic acid; NTD, neural tube defect.

Human subjects

Aborted NTD case subjects were obtained from Fenyang Hospital and Linxian County Hospital (Shanxi, China). The surgical details were, as previously described (14). The spinal cord tissues were used in the following experiments. A total of four NTD-affected fetuses and four age-matched controls were used in the present study. The clinical phenotypes of the four cases of NTD were as follows: One case of thoracic meningomyelocele (24 weeks gestation; female), one case of anencephalic and occipital encephalomeningocele (17 weeks gestation; female), and two cases of thoracic and lumbar spina bifida (20 and 26 weeks gestation, respectively; female). The present study was approved by the Committee of Medical Ethics at the Capital Institute of Pediatrics (SHERLLM2014002). Written informed consent was obtained from the parents on behalf of the fetuses.

RNA extraction and reverse transcription-quantitative polymerase chain reaction (RT-qPCR)

Total RNA was isolated from the samples using TRIzol® (cat. no. 15596-026; Invitrogen; Thermo Fisher Scientific, Inc., Waltham, MA, USA) and was reverse transcribed using a First Strand cDNA Synthesis kit (cat. no. K1612; Beijing TransGen Biotech Co., Ltd., Beijing, China). The cDNA samples were analyzed using an Applied Biosystems 7500 Real-Time PCR system (Thermo Fisher Scientific, Inc.) and a 2X PCR UltraSYBR Mixture kit (cat. no. CW0956; CWBIO, Beijing, China) containing carboxy-X-rhodamine, according to the manufacturer's protocol. The expression levels of the target genes were normalized to glyceraldehyde 3-phosphate dehydrogenase (Gapdh). The fold change in expression was determined using the 2−ΔΔCq method (16). The primer sequences designed using Primer 5 and provided by Thermo Fisher Scientific, Inc. are listed in Table I. The PCR thermocycling steps were as follows: 95°C for 10 min, then 40 cycles of 95°C for 15 sec, 60°C for 1 min and 72°C for 5 min.

Table I Primers for reverse transcription-quantitative polymerase chain reaction analyses of Wnt2b and Wnt7b mRNA.

Table I

Primers for reverse transcription-quantitative polymerase chain reaction analyses of Wnt2b and Wnt7b mRNA.

Gene	Primer (5′→3′)
mWnt2b	F: CGAGGTGGCAAACATCCTAT
	R: CTTTGAAGGCTCCACTCCTG
mWnt7b	F: GGCTCCTTCCTACTCGCTCT
	R: GAGTGTGACTGGTGGGTGTG
Gapdh	F: TGTGTCCGTCGTGGATCTGA
	R: CCTGCTTCACCACCTTCTTGA

[i] m, mouse; F, forward; R, reverse; Gapdh, glyceraldehyde 3-phosphate dehydrogenase.

Chromatin immunoprecipitation (ChIP) assays

A ChIP assay kit was purchased from Invitrogen (Thermo Fisher Scientific, Inc.; cat. no. 49–2024). Chromatin was prepared from targeted tissues and sonicated to DNA lengths between 300 and 500 bp. The following antibodies were used: Polyclonal anti-rabbit H3K4me (cat. no. ab106165; Abcam, Cambridge, UK), polyclonal anti-rabbit H3K4ac (cat. no. ab113672; Abcam) and anti-H3K27me3 (cat. no. 17–622; EMD Millipore, Billerica, MA, USA) antibodies. Alkaline phosphate-conjugated rabbit anti-immunoglobulin G (Beijing Zhongshan Golden Bridge Biotechnology Co., Ltd., Beijing, China)-treated chromatin immunoprecipitated samples served as a negative control. The antibodies were coupled with Dynabeads® Protein A/G (Thermo Fisher Scientific, Inc.; cat. no. 49–2024) at 4°C for 1 h, then chromatin was bound to the antibody-dynabeads complex at 4°C for 2 h. ChIP-qPCR analysis was performed using an ABI 7500 system (Applied Biosystems). The primers (Thermo Fisher Scientific, Inc.) used for ChIP-PCRs are listed in Table II. The relative enrichment of the histone modification was determined using the 2(input-Cq)NTDs/2(input-Cq)control method (14).

Table II Primer sequences for chromatin immunoprecipitation.

Table II

Primer sequences for chromatin immunoprecipitation.

Gene	Primer (5′→3′)	RefSeq accession	Product size (bp)	Target region
Wnt2b-1	F: CGAGAGTGGGTGGGAGAAG	NM_009520	125	chr3:104765115-104765239
	R: GGACTCCTTTGGCCACACTG
Wnt2b-2	F: GAAGTGTCCGCAACCCTCTC	NM_009520	103	chr3:104764911-104765013
	R: TCCTAGCGCCCTTCACTCA
Wnt7b	F: CGGAGCCAATACGCAGCA	NM_001163633	100	chr15:85411511-85411610
	R: CCAAACAGGTGAAAGTACGCG
WNT2B-1	F: GGGCGGTGATAGAAGTTGCT	NM_004185	122	chr1:113008796-113008917
	R: TCAGCTTTCCGAAGAAGGGC
WNT2B-2	F: CCCAGTGGAGTCAGGAAAGG	NM_004185	141	chr1:113009674-113009814
	R: TGGCAACTTGGTGCACTGTA
WNT7B	F: GAGACCACCCACTCCCATTG	NM_058238	145	chr22:46373432-46373576
	R: CCTGGGGTTGTGGGAATCTC

[i] F, forward; R, reverse.

DNA methylation analyses

Following ultrasonic sonication (P-141008; Diagenode SA, Seraing, Belgium) of mouse and human spinal cord tissues, genomic DNA was extracted from the homogenates using a Genomic DNA Miniprep kit (Axygen Scientific, Inc., Union City, CA, USA). A total of 500 ng genomic DNA from each sample was bisulfite-treated using an EZ DNA Methylation-Gold™ kit (cat. no. D5005; Zymo Research, Irvine, CA, USA). The Sequenom MassARRAY platform (CapitalBio Corporation, Beijing, China) was used to perform quantitative methylation analysis of Wnt2b and Wnt7b (13). The primers used to detect DNA methylation levels of the target regions are presented in Table III.

Table III Primer sequences for DNA methylation.

Table III

Primer sequences for DNA methylation.

Gene	Primer (5′→3′)	RefSeq accession	Product size (bp)	Target region
Wnt2b-1	F: aggaagagagGGGAAAGTAGTTTAGTTGTTTTTGA	NM_009520	349	chr3:104765115-104765463
	R: cagtaatacgactcactatagggagaaggctAAACTCCTTTAACCACACTACCCTC
Wnt2b-2	F: aggaagagagAGGGTAGTGTGGTTAAAGGAGTTTT	NM_009520	280	chr3:104764859-104765138
	R: cagtaatacgactcactatagggagaaggctCTTCCCTACCTACAACCCCCTAT
Wnt7b	F: aggaagagagTGAAGTGGTTAGGTTGGTTGGTAT	NM_001163633	205	chr15:85411502-85411706
	R: cagtaatacgactcactatagggagaaggctACCTAAAACCCAAACAAATAAAAAT
WNT2B-1	F: aggaagagagGAAGGATAGGGTTAGTGGTTAGGAA	NM_004185	340	chr1:113008996-113009335
	R: cagtaatacgactcactatagggagaaggctAACCCCTTTTTAACAATTATAAACATCT
WNT2B-2	F: aggaagagagTTTTTTTTAGGGTTTTGTGAGAATTT	NM_004185	338	chr1:113008683-113009020
	R: cagtaatacgactcactatagggagaaggctTTCCTAACCACTAACCCTATCCTTC
WNT2B-3	F: aggaagagagGATGGGGTTTTTAAGATTTAGGAGA	NM_004185	340	chr1:113008435-113008774
	R: cagtaatacgactcactatagggagaaggctACAAAAAAAACAACAAACCCATAAA
WNT2B-4	F: aggaagagagTTGTTATATTTGTTTGTGGAGGAGG	NM_004185	268	chr1:113007526-113007793
	R: cagtaatacgactcactatagggagaaggctAATCAAAATCTTCCTAAAAACACCC	NM_004185
WNT2B-5	F: aggaagagagTTTTGTTTTTATGTTTTTGAATAGG	NM_004185	328	chr1:113007314-113007641
	R: cagtaatacgactcactatagggagaaggctACCCCTTCCCAACAAAACTATTAT
WNT7B-1	F: aggaagagagGTTTTGGGGAGTAGGGGTTATTTA	NM_058238	326	chr22:46374152-46374477
	R: cagtaatacgactcactatagggagaaggctACTACCCTATCCCACCTAACCAAAC
WNT7B-2	F: aggaagagagGGGAGTTATTTAGGGTTGAGAAAGT	NM_058238	318	chr22:46373742-46374059
	R: cagtaatacgactcactatagggagaaggctACACTAAAAAAAACTCCTAACCCCC	NM_058238
WNT7B-3	F: aggaagagagGGTTAATGGGTTTTGTAGGAGGTTA	NM_058238	265	chr22:46373430-46373694
	R: cagtaatacgactcactatagggagaaggctAAAAAACCACCCACTCCCATTAC	NM_058238

[i] Upper case bases indicate the target region sequence, lower case bases indicate the probe sequence. F, forward; R, reverse.

Statistical analysis

Statistical analyses of the transcripts, histone modification enrichment and DNA methylation levels of Wnt genes were performed at least three times. Data were compared between the test and control groups using Student's t-test with SPSS software, version 20 (IBM SPSS, Armonk, NY, USA). P<0.05 was considered to indicate a statistically significant difference.

Results

Altered transcription levels of Wnt2b and Wnt7b in a mouse model of RA-induced spina bifida

To determine into the possible changes occurring in the expression levels of Wnt2b and Wnt7b during the development of RA-induced NTDs, the mRNA expression levels were determined. The results indicated that, in the mice with spina bifida, the mRNA expression levels of Wnt2b and Wnt7b were 3.1- and 2.8-fold higher, respectively, compared with the control mice (Fig. 1C). These results were consistent with those of a previous study, which detected the transcriptional upregulation of WNT2 (including WNT2B) and WNT7B in NT2/NTera2 cells following RA treatment (17). These results indicated that the transcription of Wnt2b and Wnt7b were abnormally activated in RA-induced spina bifida.

Abnormal histone modifications are present in the promoter regions of Wnt2b and Wnt7b

Due to the ectopic alterations in mRNA expression levels, epigenetic factors in the GC-rich promoter regions of Wnt2b and Wnt7b were examined, according to the current UCSC database (July 2007 NCBI37/mm9; http://genome.ucsc.edu/). As shown in Fig. 2A and B, H3K4me3 and H3K27me3 modifications target Wnt2b and Wnt7b in embryonic stem cells and the brain in the central nervous system, whereas H3K9me3 modifications are weak. Furthermore, in the cerebellum and mouse embryonic fibroblasts, these histone modification enrichments are markedly lower, indicating that the histone modifications, which occur in Wnt2b and Wnt7b are stable and spatio-temporal specific. Therefore, the present study aimed to determine whether H3K27me3 or H3K4me bind to the target regions of samples by performing ChIP assays. As H3K4ac is also a transcriptional activation marker (18), the enrichment of this modification was also assessed. In the target regions of Wnt2b and Wnt7b in the RA-treated NTD mice, enrichments of H3K4ac were upregulated, whereas H3K27me3 enrichments were significantly downregulated in the chr3: 104765115–104765239 segment of Wnt2b (Fig. 3A). The enrichment of H3K4me was undetectable in the target regions of the two genes (data not shown). The abnormally attenuated H3K27me3 and enhanced H3K4ac enrichment of Wnt2b and Wnt7b indicated their roles in the dysregulated increased transcription observed in the mice with NTDs.

Figure 2 Histone modification enrichment profiles of Wnt2b and Wnt7b in mice. Browser shots from the UCSC genome browser showing H3K4me3, H3K9me3 and H3K27me3 ChIP-seq data in various tissues and cells. The positions of UCSC CpG islands (green) are shown. ChIP-seq data are presented as the number of reads that overlap genomic windows. In mice, H3K4me3 and H3K27me3 modification levels of (A) Wnt2b and (B) Wnt7b in ES cells and brain tissue were higher, compared with in the MEF and cerebellum; however, H3K9me3 modifications in these two Wnt genes in cells and tissues were low. Orange dotted box indicates the target regions of genes in the present study. H3K, histone 3 lysine; me3, trimethylation; ChIP, chromatin immunoprecipitation; ES, embryonic stem cell; MEF, mouse embryonic fibroblast.

Figure 3 Alterations of epigenetic modification in NTDs. (A) In Wnt2b, H3K27me3 enrichment was decreased, whereas H3K4ac enrichment was increased in the promoter regions of Wnt2b and Wnt7b. DNA methylation levels of the specific CpG sites within (B and C) Wnt2b and (D) Wnt7b in a mouse model of retinoic acid-induced NTD, compared with in the control mice. *P<0.05; data are presented as the mean ± standard error. NTD, neural tube defect; H3K, histone 3 lysine; me3, trimethylation; ac, acetylation; TSS, transcription start site.

DNA methylation levels of Wnt2b and Wnt7b are abnormal in histone modification-enriched regions

Alterations in the histone modification profiles of the Wnt genes prompted the present study to examine DNA methylation in the histone modification-enriched regions. No difference was observed in the mean methylation levels of the target Wnt2b and Wnt7b regions between the NTD samples (Wnt2b, 5.36%; Wnt7b, 5.13%) and the control samples (Wnt2b, 5.79%; Wnt7b,4.93%; P=0.70 and P=0.92, respectively), as determined using the MassARRAY platform. However, the methylation levels of Wnt2b at the CpG-units chr3: 104765161-62 (−535 bp upstream from TSS) and chr3: 104765089-90 (−463 bp upstream from TSS) were significantly hypomethylated in the mice with spina bifida, compared with the controls (Fig. 3B and C). For Wnt7b, the chr15: 85411571-72 CpG site (407 bp upstream of the TSS) was also hypomethylated (Fig. 3D). These results suggested that, in NTDs, the increased mRNA expression levels of Wnt2b and Wnt7b may also be partially due to hypomethylation of specific CpG-units in the promoter regions.

The combination of the ChIP assay and DNA methylation assay data suggested that, in the mouse model of RA-induced spina bifida, increased H3K4ac and decreased H3K27me3 enrichment, and hypomethylation of CpG islands in the promoter regions potentially led to enhanced mRNA expression (Fig. 4).

Figure 4 Schematic of H3K4ac, H3K27me3 and DNA methylation on target regions of Wnt2b and Wnt7b in mice H3K, histone 3 lysine; ac, acetylation; me3, trimethylation.

Aberrant epigenetic modification of WNT2B and WNT7B in human NTDs

As abnormal epigenetic modifications of Wnt2b and Wnt7b were detected in the mice with NTDs, the present study subsequently aimed to validate whether similar changes were present in the human NTD samples. In H1 embryonic stem cells, at the early stage prior to neural development, the GC-rich promoters of the two WNT genes were enriched for H3K27me3 (Fig. 5A and B), and were bound to a series of enzymes for its catalysis, including enhancer of zeste homolog 2 (EZH2) and suppressor of zeste homolog 12 (SUZ12). However, low levels of HDAC2-6 were detected, which is known to regulate the deacetylation of H3. Furthermore, astrocytes (NH-A), which are a type of mature neuronal cell associated with the later stage of neural development, exhibited reduced H3K27me3 and EZH2 enrichment (19), indicating that H3K27me3 and H3K4ac may preferentially function in the early stage of neural development.

Figure 5 Histone modification enrichment profiles of WNT2B and WNT7B in humans. Browser shots from the UCSC genome browser showing H3K4me3, H3K9me3 and H3K27me3 ChIP-seq data in various tissues and cells. The positions of the UCSC CpG islands (green) are shown. ChIP-seq data is presented as the number of reads that overlap genomic windows. In humans, stable H3K27me3 modifications were detected in (A) WNT2B and (B) WNT7B in the ESCs, but not in NH-A. Orange dotted box indicates the target regions of genes in the present study. H3K, histone 3 lysine; me3, trimethylation; ChIP, chromatin immunoprecipitation; hESC, human embryonic stem cell; NH-A, astrocyte; EZH2, enhancer of zeste homolog 2; SUZ, suppressor of zeste homolog 12; HDAC, histone deacetylase.

In the human spinal cord samples, the GC-rich promoter regions of WNT2B and WNT7B were enriched for H3K4ac and H3K27me3. However, in the NTD samples, H3K4ac enrichment was downregulated, compared with the controls, and there were no differences in H3K27me3 enrichment between the groups (Fig. 6A). Furthermore, several CpG-units of WNT2B and WNT7B were significantly hypomethylated, compared with the control group, particularly those close to the TSS (Figs. 6B–D and 7). These results indicated that aberrant epigenetic modification of WNT2B and WNT7B occurred in the human NTDs.

Figure 6 Alterations of epigenetic modification in human NTDs. (A) In WNT2B and WNT7B, H3K4ac enrichment was reduced in NTDs, whereas no differences in H3K27me3 modification were observed between the NTD and control groups. DNA methylation levels of the specific CpG sites within (B and C) WNT2B and (D) WNT7B target regions in human patients with NTDs, compared with in the controls. *P<0.05; data are presented as the mean ± standard error. NTD, neural tube defect; H3K, histone 3 lysine; me3, trimethylation; ac, acetylation; TSS, transcription start site.

Figure 7 Schematic of H3K4ac, H3K27me3 and DNA methylation on target regions of WNT2B and WNT7B in human H3K, histone 3 lysine; ac, acetylation; me3, trimethylation.

Discussion

Previous studies have demonstrated that H3K27me3 binding assists in maintaining gene silencing in multi-cellular organisms, and is associated with the timed suppression of developmental regulatory genes (20,21). By contrast, H3K4ac is considered to be involved in the activation of gene transcription (18), whereas methylated DNA makes it difficult for histones to form nucleosomes, resulting in chromosomal instability and modified accessibility of regulatory proteins, which alter the pattern of gene expression (12,22). The present study hypothesized that the aberrant DNA methylation and histone modification profiles observed in NTDs may affect the binding ability of RNA polymerase and key transcription factors to promoter regions on genes, including EZH2, SUZ12, HDAC2, HDAC6, which in turn regulate histone modification and affect transcription initiation (19). In addition, H3K4ac and H3K27me3 modifications of the target genes in the present study were observed to function preferentially at the early stage of neural development, however not in the later stages. Furthermore, taking the data from our previous study into account, which reported DNA hypomethylation of long interspersed nuclear element-1 and low expression levels of H3K79me2 in human NTDs (13,14), the present study reiterates the hypothesis that abnormal epigenetic modifications have a destructive role in the etiology of NTDs.

It remains uncertain how alterations in Wnt2b and Wnt7b affect normal neural development. As ligands, Wnt2b and Wnt7b can function through the canonical β-catenin pathway, which maintains the proliferative state of neural progenitor cells (23,24). In addition, Wnt7b can signal through the non-canonical pathway (planar cell polarity), affect convergent extension movements prior to the closure of the neural tube, and control A/P axis formation and neural patterning (25). Therefore, the upregulated expression levels of Wnt2b and Wnt7b may lead to excessive signaling and result in disordered neural development, suggesting that aberrant Wnt gene expression may be involved in the etiology of NTDs. However, the Wnt5a genetic knockout mice demonstrated that individual deficiencies in Wnt genes do not induce an NTD phenotype (5). This suggests that, when multiple Wnt ligands are altered at one time, they may potentially contribute to the etiology of disease. However, further studies are required to confirm this hypothesis.

In conclusion, the present study provided novel insight into the synergistic function of transcriptional dysregulation and epigenetic modifications. Specifically, abnormal epigenetic modifications in the Wnt genes were detected following exposure to RA and may significantly contribute to the development and etiology of NTDs.

Acknowledgments

The present study was supported by the National Natural Science Foundation of China (Beijing, China; grant. nos. 81471163 and 81300489) and the National '973' project (grant. no. 2013CB945404). The authors would like to thank all participating hospitals and medical staff for their assistance in sample collection and recording of clinical information, and would like to thank all subjects and their family members for their cooperation in providing clinical information and samples for this study.

References