Regulation of the expression of tumor necrosis factor‑related genes by abnormal histone H3K27 acetylation: Implications for neural tube defects.

The association between apoptosis and neural tube defects (NTDs) is recognized as important, however, the precise link remains to be elucidated. Epigenetic modifications in human NTDs have been detected previously. In the present study, the occurrence of epigenetic modifications in apoptosis‑related genes was investigated in a retinoic acid (RA)‑induced mouse NTD model. Among 84 key genes involved in programmed cell death, 13 genes, including tumor necrosis factor (Tnf), annexin A5, apoptosis inhibitor 5, Bcl2‑associated athanogene 3, baculoviral IAP repeat‑containing 3, caspase (Casp)12, Casp4, Casp8, lymphotoxin β receptor, NLR family, apoptosis inhibitory protein 2, TNF receptor superfamily (Tnfrsf)1a, TNF superfamily (Tnfs)f10 and Tnfsf12, were downregulated, whereas nucleolar protein 3 was upregulated in the RA‑induced NTD mice. Chromatin immunoprecipitation assays revealed that the regulatory regions of these differentially expressed TNF‑related genes showed reduced histone H3K27 acetylation in NTDs, compared with control mice without NTDs. Reverse transcription‑quantitative polymerase chain reaction revealed that H3K27ac‑binding to the differentially regulated genes was markedly decreased in the NTD mice, whereas binding to the unchanged genes Casp3 and Nfkb1 was unaffected. In conclusion, certain TNF‑related genes appeared to be downregulated in NTDs, possibly as a result of abnormal histone H3K27 acetylation. These results shed new light on the epigenetic dysregulation of apoptosis‑related genes in NTDs.

Introduction

Neural tube defects (NTDs) are severe congenital birth malformations in the brain and spinal cord, which are caused by incomplete neural tube closure. They occur in 5.3 per 10,000 live births in the United States (1), and in >10/1,000 pregnancies in certain regions of China (2). Although the pathological mechanisms responsible for NTDs remain to be elucidated, at least 240 gene mutations in mice are known to increase the risk of NTDs, including several mutations in apoptosis-related genes (3).

Retinoids, a group of natural and synthetic derivatives of vitamin A, are important in the regulation of cellular proliferation, differentiation and apoptosis. Cell proliferation, differentiation, migration and apoptosis are key cellular processes in the regulation of neural tube closure (4). Vitamin A is required for central nervous system (CNS) development; a lack of vitamin A can result in the posterior hindbrain failing to develop as the cells destined to form this region of the CNS in the early embryo die through apoptosis (5). Cell apoptosis is increased in the neuroepithelium of the lumbosacral neural tube of rat embryos with all-trans-retinoic acid (ATRA)-induced spina bifida aperta (6). The increased cell death in the tail bud may be important in the formation of ATRA-induced spina bifida aperta (7).

Covalent histone modifications, including the methylation and acetylation of certain amino acid residues in chromatin histones, have been shown to be essential in gene regulation by modulating chromatin structure. Histone H3K27 acetylation (H3K27ac) is commonly located in gene-regulatory regions and is associated with gene activation, whereas trimethylated H3K27 (H3K27me3) represses transcription (8). Rada-Iglesias et al (9), showed that H3K27ac at specific differentiation gene enhancer regions is necessary for the conversion of human embryonic stem cells (hESCs) into neural stem cells. Balmer et al (10) used the differentiation of hESCs into neuroectodermal precursors as a model to investigate the mechanism underlying the effect of the histone deacetylase inhibitor trichostatin A, which was found to cause epigenetic changes and disturbed neural development. Mutations in genes encoding histone deacetylases, including EP300, CREBBP and CITED2, have also been shown to cause NTDs (11). In addition, our previous study demonstrated abnormal epigenetic modifications in human fetuses with NTDs (12,13), although the detailed mechanisms and target genes remain to be elucidated. The present study investigated the association between epigenetic modifications and the expression of apoptosis-related genes in NTDs.

Materials and methods

Animals

All animal experiments were performed in accordance with the standards of the Capital Institute of Pediatrics Ethics Committee (Beijing, China; permit no: DWLL2015015). Adult male and female C57BL/6J mice were placed together at 08:00 a.m. for timed mating, and when a vaginal plug was observed at 10:00 a.m., this was considered as embryonic day 0 (E0). Subsequently, three doses of ATRA (20 mg/kg body weight; Sigma; Merck KGaA, Darmstadt, Germany) suspended in olive oil were administered intraperitoneally into the mothers on gestational day (GD)8.0 at 0 h, GD8.0 at 6 h, and GD8.0 at 12 h when the neural tube and neuropore were fusing. The control consisted of olive oil administration, and mice were treated with the same volume of pure olive oil at the same time points as a control group, as described previously (14). E18.0 fetuses were delivered by cesarean section and 18 brain tissue samples from the fetuses with NTDs were obtained and pooled into three groups. At the same time, 18 brain tissue samples from 18 normal fetuses were pooled randomly to create three control groups.

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

Total RNA was isolated from tissue samples using TRIzol (cat. no. 15596026; Invitrogen; Thermo Fisher Scientific, Inc., Waltham, MA, USA) and analyzed on an Applied Biosystems 7500 Real-Time PCR system (Applied Biosystems; Thermo Fisher Scientific, Inc.) using a first strand cDNA synthesis kit (K1612; TransGen Biotech Co., Ltd., Beijing, China). Amplification was performed according to the following conditions: 1 cycle at 95°C for 10 min, 40 cycles at 95°C for 15 sec and 60°C for 1 min. The apoptosis-related genes were detected using the Mouse Apoptosis PCR Array from Qiagen GmbH (Hilden Germany; http://www.sabiosciences.com/rt_pcr_product/HTML/PAMM-012Z.html) according to the manufacturer's protocols. The Mouse Apoptosis PCR Array was used to profile the expression of 84 key genes involved in programmed cell death. This array includes tumor necrosis factor (TNF) ligands and their receptors, members of the B-cell lymphoma 2 (bcl-2), caspase, inhibitor of apoptosis (IAP), TNF receptor-associated factor, caspase activator of recruitment domain, death domain, death effector domain, and cell death-inducing DFFA-like effector families, in addition to genes involved in the p53 and DNA damage pathways. All the apoptosis-related genes are listed in Table I. The data were analyzed using Qiagen RT2 Profiler PCR Array Data Analysis version 3.5 (Qiagen GmbH; product. no. 330231; cat. no. PAMM-012Z). The expression levels of target genes were normalized to β-actin and β2 microglobulin. Fold changes were determined using the 2−ΔΔCq method (15).

Table I.

Functional gene grouping of the 84 key apoptosis-related genes.

Function	Genes
Induction of apoptosis
Induction of apoptosis	Cradd, Fadd, Tnf, Tnfrsf10b (Dr5)
DNA damage	Cidea, Cideb, Trp53, Trp63, Trp73
Extracellular signals	Cflar (Casper), Dapk1
Other	Bad, Bak1, Bax, Bcl10, Bcl2l11, Bid, Bnip3, Bok, Casp1 (Ice), Casp12, Casp14, Casp2, Casp3,Casp4, Casp6, Casp8, Cd70 (Tnfsf7), Dffa, Dffb, Diablo (Smac), Fas (Tnfrsf6), Fas1 (Tnfsf6), Mapk1 (Erk2), Nod1 (Card4), Pycard (Tms1/Asc), Tnfsf10 (Trail), Trp53bp2, Traf3
Regulation of apoptosis
Negative regulation	Bag1, Bag3, Bcl10, Bcl2, Bcl2a1a, Bcl2l2 (Bcl-X), Bcl2l10, Bcl2l2, Birc2 (C-Iap2), Bir3 (C-Iap1), Bnip2, Bnip3, Bnip3l, Casp3, Cd40lg (Tnfsf5), Nol3, Polb, Prdx2, Tnf, Trp63, Xiap (Birc4)
Positive regulation	Abl1, Akt1, Anxa5, Bad, Bak1, Bax, Bcl2l11, Bid, Bnip3, Bnip3l, Card10, Casp1 (Ice), Casp14, Casp2, Casp4, Casp6, Casp8, Cd40 (Tnfrsf5), Cd70 (Tnfsf7), Cideb, Cradd, Fadd, Fasl (Tnfsf6), Gadd45a, Ltbr, Nod1 (Card4), Pycard (Tms1/Acs), Tnf, Tnfrsf10b (Dr5), Tnfsf10 (Trail), Tnfsf12, Trp53, Trp53bp2, Traf1, Traf2, Traf3
Caspases and regulators
Caspases	Casp1 (Ice), Casp12, Casp14, Casp2, Casp3, Casp4, Casp6, Casp7, Casp8, Casp9, Cflar (Casper), Cradd, Pycard (Tms1/Asc)
Caspase activators	Aifm1 (Pdcd8), Apaf1, Bax, Bcl2l10, Casp1 (Ice), Casp9, Nod1 (Card4), Pycard (Tms1/Asc), Tnfrsf10B (Dr5), Trp53
Caspase inhibitors	Birc5, Xiap (Birc4)
Anti-apoptotic	Akt1, Api5, Atf5, Bag1, Bag3, Bax, Bcl2, Bcl2l1 (Bcl-x), Bcl2l10, Bcl2l2, Birc5, Bnip2, Bnip3, Bnip3l, Cd40lg (Tnfsf5), Cflar (Casper), Dad1, Dapk1, Fas (Tnfrsf6), Igf1R, Il10, Lhx4, Mcl1, Naip1 (Birc1), Naip2, Nfkb1, Nme5, Nol3, Polb, Prdx2, Tnf, Trp63, Xiap (Birc4)
DEATH domain proteins	Cradd, Dapk1, Fadd, Nfkb1, Ripk1, Tnfrsf10b (Dr5), Tnfrsf11b, Tnfrsf1a

Chromatin immunoprecipitation (ChIP) assays

The cross-linked, sonicated DNA products were incubated with H3K27ac (cat. no. 17-683; Abcam, Cambridge, UK) and H3K27me3 (cat. no. 17-622; EMD Millipore, Billerica, MA, USA) antibodies, with mouse IgG antibody as a negative control. ChIP (cat. no. 49-2024; Invitrogen; Thermo Fisher Scientific, Inc.) and RT-qPCR analysis was performed using the ABI 7500 system. All primers used for ChIP-PCR are listed in Table II. Relative enrichment of histone modification was determined as follows: 2(input-Cq)NTDs/2(input-Cq)control.

Table II.

Primer sequences for chromatin immunoprecipitation assay.

Gene	Primer	Primer sequence (5′-3′)
Tnf	Forward	CCATGCGTCCAGCTGACTAA
	Reverse	TCCCCTGGCCACATCTTTC
Tnfsf12	Forward	GAGTAGGTGTGGATGCTGCGT
	Reverse	GTCTCCCTGCACACACCTCA
Tnfsf10	Forward	AGTGGCCGTGTGCTCTCAG
	Reverse	TCTGCACTTGGGAAAGGCAT
Tnfsf1a	Forward	GAGGGTGGGTTGCTAGCGA
	Reverse	AGCAAACCTGGCGTCAAACTA
Ltbr	Forward	TTTGAAGAGGTAGCGGCATGT
	Reverse	GCCTCTCCTTGGTCTCTTTGTC
Casp8	Forward	CCATAGCAGGCCATTTAGCTG
	Reverse	ATGCAGGCTGCTGGAGAAA
Casp3	Forward	GCAAGACAGCTCACCCTGAA
	Reverse	TTCCCATTGCCTAAAGCCGA
Nfkb1	Forward	GTGGCGCTGAAGGACTGTAT
	Reverse	TCTACAGCACAGCACCGAAT

Tnf, tumor necrosis factor; Tnfsf, Tnf superfamily; Ltbr, lymphotoxin β receptor; Casp, caspase; Nfkb1, nuclear factor-κB subunit b1.

Pathway analysis

To investigate the possible biological interactions of the differentially regulated genes using GNCPro (http://gncpro.sabiosciences.com/gncpro.gncpro.php), an online software tool developed and maintained by SABiosciences (Qiagen GmbH, Hilden, Germany) for collating gene and pathway interactions.

Gene mapping in the genome

All TNF-related gene sets were mapped in the genome of mouse using the RefSeq track of the University of California-Santa Cruz (UCSC) Genome Browser (version NCBI37/mm9; July 2007), Histone H3K27ac and H3K27me3 ChIP-seq profiles of the mouse brain were analyzed together. ChIP-seq data is presented as the number of reads overlapping genomic windows.

Statistical analysis

The transcript and histone modification enrichment analyses were performed ≥3 times and the results were compared between the experimental and control groups using Student's t-tests. P<0.05 was considered to indicate a statistically significant difference. Statistical analysis was performed using SPSS for Windows, v13.0 (SPSS Inc., Chicago, IL, USA).

Results

Changes in the transcription of apoptosis-related genes in NTD mice

To detect changes in the expression of apoptosis-related genes during the development of RA-induced NTDs, the present study examined the mRNA expression levels of 84 key apoptosis-related genes in brain tissues obtained from E18 fetuses. According to the gene expression profiles, the three NTD samples were clustered into one group, whereas the three normal control samples were clustered into another group, indicating that the transcription levels of these genes differed significantly between the NTD and control mice (Fig. 1A). Among the 84 key genes involved in programmed cell death, 13 were downregulated and one was upregulated in the NTD group, compared with the control (Fig. 1B). The downregulated genes were Tnf, annexin A5 (Anxa5), apoptosis inhibitor 5 (Api5), Bcl2-associated athanogene 3 (Bag3), baculoviral IAP repeat-containing 3 (Birc3), caspase (Casp)12, Casp4, Casp8, lymphotoxin β receptor (Ltbr), NLR family, apoptosis inhibitory protein 2 (Naip2), TNF receptor superfamily (Tnfrsf)1a, Tnfsf10 and Tnfsf12, whereas the single upregulated gene was nucleolar protein 13 (Nol3), which belongs to the IAP family (Fig. 1C).

Figure 1.

Apoptosis-related gene expression profiles of retinoic acid-induced NTD and normal mouse brains. (A) Cluster analysis of data from the mouse apoptosis polymerase chain reaction array. Red indicates upregulation and green indicates downregulation. (B) Summary of expression profile analysis of the 84 key genes involved in programmed cell death. (C) Differentially expressed genes. NTD, neural tube defect.

Subsequently, analysis was performed for candidate genes using GNCPro online. The genes were classified as anti-apoptotic or pro-apoptotic regulators. There was a significant TNF-related network involved in NTD formation (Fig. 2). The Tnf, Ltbr, Tnfrsf1a, Tnfsf10 and Tnfsf12 genes all encode TNF superfamily or TNF-related apoptosis-inducing ligand (TRAIL) superfamily proteins. The activation of these proteins by their associated ligands rapidly activates caspase-8 and leads to cell destruction (16,17). Casp12, Casp4 and Casp8 genes encode caspase family proteins, which are key proteins regulating apoptosis. Casp8-knockout mice show abnormal yolk sac vasculature and various defects in the developing heart and neural tube (18,19). The proteins encoded by Birc3, Naip2 and Nol3 belong to the IAP family, the members of which bind and inhibit caspases, and are important regulators of cell differentiation and cell death (20). The functions of Anxa5 and Bag3 remain to be elucidated. These data indicated that the transcription levels of certain apoptosis-related genes were abnormally activated in RA-induced NTDs.

Figure 2.

Interaction analysis of differentially expressed genes. The gene expression profile of the NTD mice differed from that of the normal controls. One gene was upregulated and 13 were downregulated in the NTD mice. The TNF pathway was the pathway with the highest representation among the differentially expressed genes. NTD, neural tube defect.

Abnormal histone modifications in TNF-related gene promoters

To determine the mechanisms responsible for the ectopic changes in TNF-related gene expression, the present study examined the histone-modification status in the regulatory regions of the altered genes. H3K27ac and H3K27me3 profiles were analyzed according to the current UCSC database (July 2007; NCBI37/mm9). The H3K27ac and H3K27me3 modifications were all targeted to the regulatory region of TNF-related genes (Tnf, Tnfrsf1a, Tnfsf10, Tnfsf12 and Casp8) in the brain (Fig. 3A-E). All of these TNF-related genes were regulated by p300, which functions as a histone acetyltransferase and regulates transcription via chromatin remodeling. A search was performed of the regulatory regions of TNF-related genes in the brain using the Genome Browser, and it was found that the TNF-related genes were also bound to p300, which indicated that these genes were regulated by histone acetylransferase P300 (Fig. 3). A ChIP-qPCR primer was designed from the 1 kb relative to the transcription start site in the highly targeted region, to determine whether the altered expression of apoptosis-related genes was caused by abnormal regulation of H3K27ac and H3K27me3.

Figure 3.

Histone H3K27ac and H3K27me3 enrichment profiles of TNF-related genes in the mouse brain. Screenshots from the UCSC genome browser showing H3K27ac and H3K27me3 ChIP-seq data of the mouse brain. ChIP-seq data is presented as the number of reads overlapping genomic windows. There was a high association between H3K27ac and H3K27me3 modification in the 1,000 bp downstream of the expression of TNF-related genes in the mice. The red boxes indicate the region 1 kb downstream from transcription start site. (A) Tnf, (B) Tnfrsf1a, (C) Tnfsf10, (D) Tnfsf12, (E) Casp8. TNF, tumor necrosis factor; ChIP, chromatin immunoprecipitation; H3K27ac, H3K27 acetylation; H3K27me3, trimethylated H3K27.

According to the results of the ChIP analysis, H3K27ac bound to the regulatory regions of differentially regulated (Tnf, Casp8, Ltbr, Tnfrsf1a, Tnfsf10 and Tnfsf12) and non-differentially regulated genes, Casp3 and nuclear factor (NF)-κB subunit 1 (Nfkb1), in the normal control mice (Fig. 4). However, H3K27ac-binding to the differentially regulated genes was markedly decreased in the NTD mice (Fig. 4), whereas binding to the unchanged genes Casp3 and Nfkb1 was unaffected. There was no significant difference in the binding of H3K27me3 to apoptosis-related genes between the NTD and normal control mice. These results suggested that the downregulated expression of TNF-related genes in the RA-induced NTD mice may have been caused by abnormal H3K27ac binding in their regulatory regions; this confirmed the aberrant epigenetic control of these genes in NTDs.

Figure 4.

Chromatin immunoprecipitation assay of H3K27ac and H3K27me3 binding to TNF-related genes in the normal and NTD fetal brains. Results for all TNF-related differentially expressed and two non-differentially expressed (casp3 and Nfkb1) genes. In the TNF-related gene promoter, H3K27ac enrichment was markedly downregulated in the NTD group. *P<0.05. NTD, neural tube defect; TNF, tumor necrosis factor; H3K27ac, H3K27 acetylation; H3K27me3, trimethylated H3K27.

Discussion

NTDs result from the failure of complete neurulation during the fourth week of embryogenesis. Neurulation is a complex and multistep process, involving the precise temporal and spatial regulation of gene expression (21–23). Cell apoptosis involves a process of autonomic ordered programmed cell death, which is essential for maintaining homeostasis. Apoptosis is crucial in vital biological processes, including normal embryonic development, cell signaling, the immune response, and cell death (24). Apoptosis is also important in the morphogenesis and homeostasis of the developing CNS through its involvement in the folding, pinching off and fusion of neural walls, in neural precursor selection, and in the postmitotic competition of neurons for their cellular targets (25–27). The inhibition of caspase activity in chickens prevents neural tube closure, suggesting that apoptosis is required for neurulation, whereas reduced or excessive apoptosis is associated with the failure of neural tube closure in mutant mice (28). Caspases controlling apoptosis have also been shown to be involved in metazoan embryogenesis and homeostasis (29). Previous studies have revealed that apoptosis occurs during normal development of the CNS (30). Two fundamentally different mechanisms involving apoptosis-related genes can lead to NTDs. In the 129/SvJ genetic background, apoptotic peptidase activating factor 1 (apaf-1)-, casp-9-, or casp-3-knockout can cause neurodevelopmental abnormalities. A deficiency of proapoptotic regulators in these mutants leads to reduced apoptosis, resulting in an excess of neuroepithelial cells in the hindbrain, midbrain and forebrain, and subsequent gross neuronal disorders (31–34). This suggests that apoptosis is the primary mechanism whereby mammals prevent the survival of excess cells in the developing nervous system. By contrast, NTDs can also occur as a result of excessive cell death, leaving insufficient cells for neural tube closure. A previous study showed that genetic defects in Bcl-10 and inhibitor of NF-κB kinase, which activate NF-κB in the anti-apoptotic pathway, cause exencephaly during neurulation (31–34). The activation of NF-κB in vivo is important in protecting neurons against excessive apoptosis during development. These apoptosis-related genes may be involved in NTD formation via the positive regulation of neuronal survival. Therefore, inhibited and increased apoptosis can cause NTDs and increase the risk of exencephaly.

Retinoids, a group of natural and synthetic derivatives of vitamin A, have important regulatory roles in cellular proliferation, differentiation and apoptosis. The results of the present study indicated that pro- and anti-apoptotic genes were abnormally expressed, suggesting that the two mechanisms are involved in RA-induced NTD formation in mice. Rodriguez-Leon et al revealed a mechanism by which RA, in conjunction with bone morphogenetic proteins, may regulate programmed cell death (35). Following ATRA treatment, Boutzen et al found that the reduction of cell viability and the induction of apoptosis led to a specific and marked reduction in the size and number of L-CFUs (36). Apoptosis can be affected by RA in certain developmental systems, including the neural crest cells of the rhombencephalic segments (37). Sulik and Dehart (38) investigated limb malformations, and it was found that RA resulted in excessive cell death in regions of programmed cell death and led to subsequent malformations of the affected regions. Liu et al (39) exposed the human K562 myeloid leukemia cell line to ATRA, which led to apoptosis increasing significantly. ATRA also can exert anti-apoptotic effects through the myeloid cell leukemia 1 protein at the transcriptional level (40).

The results of the present showed that TNF family genes, including Tnf, Tnfrsf1a, Tnfsf10, Tnfsf12 and Ltbr and Casp8, were involved in NTD formation. Activation by their associated ligands rapidly activates caspase-8 and leads to cell death (17,18). Casp8 has been shown to have dual roles in the activation of apoptosis and suppression of necrosis, which are two important processes in mouse embryonic development (41). Previous studies have focused on the critical role of intrinsic apoptosis in mammalian neural tube development, whereas the present study demonstrated that extrinsic-apoptosis-related genes, including the Casp8 gene, were also aberrantly expressed in RA-induced NTD.

The results of the present study demonstrated that the expression levels of TNF-related genes were decreased as a result of the abnormal regulation of H3K27ac. H3K27ac is associated with transcription activation (41,42); therefore, decreased H3K27ac binding to the gene regulatory region results in the downregulation of gene expression.

A change between H3K27me3 and H3K27ac at specific differentiation gene enhancer regions has been shown to be essential for the conversion of hESCs into neural stem cells (9). Hayashi-Takanaka et al (43) monitored the levels of H3K27ac in mouse preimplantation embryos produced by in vitro fertilization and somatic cell nuclear transfer, which revealed that a high level of H3K27ac was required for normal embryonic development. Overall, these findings indicate that abnormalities in H3K27ac are involved in the etiology of NTDs, including via the downregulation of apoptosis-related genes.

In conclusion, the present study demonstrated that apoptosis-related gene expression levels were altered in NTDs, possibly as a result of epigenetic modifications involving H3K27ac. Therefore, aberrant histone modifications during early pregnancy may contribute to the formation of NTDs by affecting neural development. Further investigations are required to confirm these results.