Reduced H3K27me3 suppresses Wnt/β-catenin signaling by S-adenosylmethionine deciency in neural tube development

S-adenosylmethionine (SAM) as a major methyl donor play a key role in methylation modication in vivo, and its disorder was closely related to neural tube defects (NTDs). However, the underlying mechanism between SAM deciency and NTDs remained unclear. Here, we investigated the association between histone methylation modication and Wnt/β-catenin signaling pathway in NTDs induced by SAM deciency. The levels of SAM and SAH were determined by enzyme linked immunosorbent assay (ELISA). The expressions of H3K27me3 and Wnt/β-catenin signaling pathway specic markers were demonstrated by western blotting, reverse transcription, and quantitative PCR (RT-qPCR) and immunouorescence in ethionine induced E11.5 mouse NTDs and NSCs models. The results showed that the incidence rate of NTDs induced by ethionine were 46.2%, post treatment of ethionine combined with SAM, the incidence rate of NTDs was reduced to 26.2%. The level of SAM was signicantly decreased (P<0.05) and a reduction in the SAM/SAH ratio was observed. The SAM depletion caused the reduction of both H3K27me3 modications and UTX activity, and inhibited the marker proteins (β-catenin, TCF-4, Axin-2, p-GSK-3β, CyclinD1, and C-myc) in Wnt/β-catenin signaling pathway (P<0.05). The differentiations of neural stem cells (NSCs) into neurons and oligodendrocytes were inhibited under SAM deciency (P<0.05). These results indicated that the depletion of SAM led to reduced H3K27me3 modications, prevented the activation of Wnt/β-catenin signaling pathway and NSCs differentiation, which provided an understanding of the novel function of epigenetic regulation in NTDs. group, and reversed by SAM supplementation. These results suggested that SAM deciency inhibited Wnt/β-catenin signaling pathway. As above, SAM deciency caused an abnormal histone methylation modication. Therefore, we speculated that SAM deciency induced abnormal expression of histone methylation modication, inhibited the activation of Wnt/β-catenin signaling pathway, and thus participated in the occurrence of NTDs. Here, our results showed that UTX, a H3K27me3 demethylase, activity and the expression of β-catenin, CyclinD1, C-myc, and GSK-3β were signicantly decreased, the expression of Axin-2 were signicantly increased after the treatment of ethionine. On contrary, the UTX demethylase inhibitor-GSK-4J activated Wnt/β-catenin signaling pathway, by up-regulating the expression level of β-catenin, CyclinD1, C-myc, and GSK-3β and down-regulating the expression of Axin-2. The results indicated that SAM deciency induced abnormal expression of H3K27me3, and inhibited the activation of Wnt/β-catenin signaling pathway. However, these ndings were not in agreement with the reports that SAM could inhibit β-catenin signaling in dedifferentiated cells [36, 37]. Lastly, we evaluated the expression and activity of H3K27me3 in ethionine-treated NSCs. UTX, also known as Kdm6b, belongs to a small family of JmjC-domain containing enzymes that mediate demethylation of H3K27me3 repressive chromatin marks Our study showed that ethionine caused an increased UTX activity, which lead to increased H3K27me3 expression level. RT-qPCR results suggested that reduced UTX activate Wnt/β-catenin signaling pathway by increasing the level of H3K27me3. In short, we inferred that this inconsistency was due to SAM-SAH metabolic disorder and these abnormal metabolites were caused by ethionine. It was implied another possibility that SAM the depletion to H3K27me3 modications and UTX activity downregulation, prevented the activation of Wnt/β-catenin signaling pathway and accelerated NSCs differentiation. provided a new understanding of the novel function of epigenetic regulation in NTDs which will have far-reaching implications in and neurobiology. Hence, the mechanisms and inuences of these altered epigenetic histone needs to concrete studies


Introduction
Abnormal development of central nervous system (CNS) caused by neural tube defects (NTDs), including anencephaly, spina bifda, and encephalocele et al, not only remained a major contributor in the prevalence of stillbirths and neonatal deaths, but also a signi cant cause of lifelong physical disability in the surviving infants. Human NTDs was considered to be associated with genetic and environmental factors [1]. The genetic predisposition is not yet well understood, but environmental factors such as maternal folic acid have been identi ed as closely related to NTDs. It had been established that maternal folate supplementation decreased/prevented the occurrence of NTDs [2], and folate forti cation was mandated in some other countries. However, the underline molecular mechanisms between folate de ciency and NTDs remained unclear.
Folate metabolism is essential nutrients required for the two major biological processes, including purine and thymidine monophosphate biosynthesis and methionine regeneration. S-adenosylmethionine (SAM), a key metabolite in the methionine regeneration, acts as a major methyl donor in numerous biochemical reactions [3,4]. It has reported that that abnormalities of SAM were associated with NTDs [5]. Ethionine is biologically antagonistic to methionine, which is an Sethyl homologue of methionine [6,7]. It competed with methionine and thus resulted in reduction of SAM during DNA, RNA and proteins synthesis [8]. Furthermore, previous studies have con rmed that ethionine could cause NTDs in the whole embryo culture [5,8]. However, the pathogenesis of methionine circulation in NTDs was not clearly clari ed. Thus, it is crucial to explore the molecular mechanisms of SAM involved both in prevention of NTDs and its potential role in human brain development.
Experimental studies indicate that β-catenin was the primary effector molecule in the Wnt/β-catenin signaling pathway [9], and playing a major role in cell growth and survival [10]. In addition, Wnt/β-catenin signaling had a major role in maintaining self-renewal as well as in regulating NSCs differentiation [11]. Increasing evidence implicated that the SAM was closely related to the Wnt/β-catenin signaling. It was reported that SAM treatment reduced the expression level of total β-catenin, in dedifferentiated cells [12]. Moreover, β-catenin signaling was regulated by SAM level in hepatocytes [13]. βcatenin signaling was required for caudal neural tube closure, and closely related to the occurrence of NTDs. However, it has not reported that whether the depleted of SAM affected the histone methylation, activated Wnt/β-catenin signaling pathway, and inhibited the differentiation of NSCs into neurons and oligodendrocytes, nally resulted in the occurrence of NTDs. Here, our studies showed that SAM activated Wnt/β-catenin signaling pathway by increasing histone methylation and strengthening cell differentiation. This study not only contribute to a better understanding of the molecular mechanisms responsible for the regulation of SAM but also provide new therapeutic targets for the treatment of NTDs.

Materials And Methods
Animals C57BL/6 mice (9-10 week, weighing 19-23 g), were provided by Shanxi Medical University which were housed in speci c pathogen-free cage with an approve facilities following a 12 hours light/dark cycle. Mature male and female C57BL/6 mice were allowed to mate overnight and vaginal plug was observed in the following morning. Noon of the day on which vaginal plug was detected, was considered as embryonic day 0.5 (E0.5). On E7.5, mice were randomly divided into two groups: control group and NTD mice model (established by intra-peritoneal injection of ethionine). The pregnant mice were sacri ced by cervical dislocation on E11.5. All embryos were dissected under a stereoscopic microscope, and the phenotype was recorded. Some embryos were xed in cold 10% neutral buffered formalin for subsequent hematoxylin and eosin staining. All procedures involving animal handling were approved by the Animal Research Ethics Board of Shanxi Medical University, China.
Cell culture and treatment Primary neural stem cells (NSCs) were isolated from mice embryos on E13.5. Brie y, the brain vesicle of mouse embryo was collected under sterile conditions, suspended as single cell in 5 mL DMEM/F12 medium (Hyclone, Logan, UT, USA) supplemented with 2% B27 (Gibco, USA), EGF (20 ng/mL, Peprotech, USA), and b-FGF (20 ng/mL, Peprotech, USA). The cells were incubated at 37°C with 5% CO 2 and passaged after every 3 days. NSCs were passaged at least twice before they were used for the subsequent experiments. The cells were treated with 20 mmol/L ethionine, 2mmol/l SAM for 48 hours, and 30 nmol/l of UTX inhibitor GSK J4 sc-391114 (Santa Cruz, USA) for 6 hours.
Hematoxylin-eosin staining The E11.5 embryos para n sections were depara nized with xylene, soaked in 100, 95, 80 and 75% ethanol for 3 minutes each. After washing with distilled water for 2 minutes, hematoxylin stain was performed for 5 minutes, and washed with tap water. Hydrochloric acid and ethanol were applied for 30 seconds, washed again 5 times under running tap water and nally soaked in tap water for 15 minutes. Post keeping in eosin solution for 2 minutes, normal dehydration, clearing and neutral resin sealing were carried out.

Western blot analysis
Proteins were extracted from E11.5 embryo brain tissues and NSCs. Total protein concentration was determined using a BCA protein assay kit (Pierce/Thermo Fisher, USA). The proteins were denatured in SDS gel-loading buffer at 95℃ for 8 minutes, 30 µg of each sample was separated on 12% SDS-PAGE gel and then transferred onto polyvinylidene uoride (PVDF) membranes (Millipore, Billerica, MA, USA). After blocking with 5% skim milk at room temperature for 1 h, the membrane was incubated with primary antibodies overnight at 4℃ followed by incubation with the secondary antibody at room temperature for 2 hours. The protein bands were visualized using an enhanced chemiluminescent (ECL) blot detection system (ChemiDoc TM Imaging Systems, BIO-RAD, USA) following manufacture's instruction. The protein bands were quanti ed using ImageJ software, and β-Tubulin, β-actin, and GAPDH were used as a control.
Enzyme linked immunosorbent assay (ELISA) On E11.5, mice were anesthetized with chloral hydrate and blood was collected from the eyeball. Serum was separated and the concentration of SAM and SAH were evaluated using the SAM and SAH ELISA kit (R&D Systems; Minnesota, USA) according to the manufacturer's protocol.
UTX activity detection UTX activity was detected according to the manufacturer's protocol using apparent enzyme JMJD3/UTX demethylase Activity/Inhibition Assay Kit (Epigentek, Farmingdale, NY). 10μg nuclear extract was used for detection.

Functional Interact Network Analysis of DEGs
The key DEG in the development of the mouse neural tube is Screening is performed by using an intersection analysis between control and ethionine. The expression patterns of the selected DEGs were analyzed by clustering 3.0 [14] and Java Treeview [15].

Statistical Analysis
Statistical analysis was performed using GraphPad Prism version 6.0 (GraphPad Software, CA) and the data was presented as mean ± SEM. The t-test was used for the the comparisons of two groups, while for multiple comparison test, ANOVA followed by Tukey post hoc test were performed. P values<0.05 were considered as statistically signi cant.

Ethionine induces neural tube defects by disrupting methionine cycle
Methionine is an important morphogen for the normal development of brain. In previous study, we had demonstrated that 500 mg/kg of ethionine caused the highest incidence rate (54.8%) of NTDs with a lower embryonic resorption rate (8.2%) [16]. Therefore, this dose of ethionine (500mg/kg) was used to establish the NTDs mouse model. The results showed that a full appearance of the structural characteristics was observed in the control embryos ( Figure 1A). Compared with E11.5 control mouse embryos, 500 mg/kg ethionine treated embryos showed an obvious overall growth retardation along with a small and hypoplastic brain vesicle ( gures 1B-D). Compared with the control, NTDs revealed that neural tube closure was failed in the hindbrain region (Figures 1E-F).
In order to investigate whether ethionine induced NTDs by disrupting the methionine cycle. SAM were intraperitoneally injected into the NTDs model on E7.5 to observe the incidence of NTDs. The results showed that the incidence rate of NTDs induced by ethionine were 46.2%. Post treatment of ethionine combined with SAM, the incidence rate of NTDs was reduced to 26.2% as shown in Table 1. However, it was found that ethionine at 500 mg/kg and SAM at 30 mg/kg dose had no effect on the weight gains during pregnancy (Table 3, Figures 1G-I). The level of SAM was signi cantly decreased in 500mg/kg ethionine treated group compared with the control group in mouse embryonic tissue ( Figure 1J). Compared with the control group, SAH level was signi cantly decreased in NTDs mouse embryonic tissue ( Figure 1K), and a consequent reduction in the SAM/SAH ratio was observed ( Figure 1L). These ndings indicated that ethionine induced NTDs via blocking the generation of SAM. SAM enhances the histone methylation level of E11.5 embryonic brain tissue and NSCs SAM, as a methyl donor, affects histone methylation modi cation. To investigate whether the depleted of SAM could affected the methylation of histones in NTDs. The levels of histone methylation were determined in E11.5 embryonic brain tissue and NSCs. The results showed that the expression of H3K4me3, H3K27me3, H3K36me3 and H3K79me3 in E11.5 ethionine-induced embryonic brain tissues were signi cantly decreased compared with normal embryonic brain tissue, especially H3K27me3 expression ( Figure 2A). Meanwhile, high expression level of H3K27me3 was observed in normal embryos, which was lowered in NTDs embryos. Consistent with the abovementioned results, H3K27me3 was signi cantly down-regulated in NTDs embryos ( Figure 2B). In contrary, the expression of H3K27me3 was markedly increased in ethionine and SAM-treated group compared with the ethionine-treated group ( Figure 2B). H3K27me3 possessed speci c methylation modi cation enzymes, including methyltransferase Ezh2, demethylase UTX and JMJD3, which are encoded by Ezh2, Kdm6a and Kdm6b genes, respectively. Here, the expression of Ezh2, Kdm6a and Kdm6b were detected to investigate the association between those enzymes and H3K27me3. The results demonstrated that the expression of Ezh2 mRNA was signi cantly decreased in ethionine group compared with control. Simultaneously, ethionine also increased the expression of Kdm6a and Kdm6b which negatively regulated the expression of H3K27me3 ( Figure 2C). Conversely, the situation was reversed after SAM supplementation. Thus, it was concluded that SAM improved histone methylation levels by increasing methyl donors.
To further prove the role of SAM in histone methylation, we have determined the histone methylation levels in neural stem cells. The neural stem cells (NSCs) were obtained from E13.5 mouse embryos, which were major cell types in CNS development [17]. The morphology of NSCs isolated from the mouse embryos was normal and globular in appearance ( Figures 3A-D). It was strongly expression of Nestin, neural stem cell marker, in the isolated NSCs ( Figures 3I-K Similarly, SAM supplementation could partially reverse this phenomenon ( Figure 3N). showed that the proteins related to the Wnt/β-catenin pathway (β-catenin, TCF-4, Axin-2, p-GSK-3β, CyclinD1, and C-myc) were inhibited in ethionine treated group compared with the control ( Figure 4A). It was found that the expression levels of β-catenin, TCF-4, Axin-2, p-GSK-3β, CyclinD1, and C-myc were upregulated in ethionine+SAM group compared with the ethionine group ( Figures 4A-B). Compared to the control group, an obvious decrease in the proportion of β-catenin + , TCF-4 + , CyclinD1 + cells relative to that of DAPI + cells were observed in ethionine-treated group ( Figure 4C). Meanwhile, SAM supplementation reversed this situation in embryos.
In order to further investigate whether the decreased level of SAM caused the inactivation of Wnt/β-catenin pathway, the protein expression level of β-catenin, TCF-4, Axin-2, p-GSK-3β, CyclinD1, and C-myc in NSCs treated with 20mM ethionine and 2mM SAM for 48h were measured. The results showed that there were signi cant decreased levels of CyclinD1, Axin-2, C-myc, β-catenin and TCF-4 in ethionine treated group ( Figure 5A-B). GSK-3β plays a key role in the canonical Wnt pathway which operates through regulating the phosphorylation and degradation of the transcription activator β-catenin. The effect of ethionine on GSK-3β showed that p-GSK-3β and GSK-3β were down-regulated in ethionine treated group compared with control, and p-GSK-3β and GSK-3β were up-regulated in ethionine+SAM group compared with ethionine group ( Figure 5C).

Reduced Utx Activates Wnt/β-catenin Signaling Pathway
Recent studies reported that UTX, a key factor for neural tube development, was H3K27me3 demethylase. In order to ensure the decreased level of H3K27me3 caused by UTX inactivated Wnt/β-catenin signaling pathway in ethionine-induced NSCs, the UTX function was determined. Results showed that UTX activity was signi cantly down-regulated after the treatment of ethionine ( Figure 6A), and ethionine treatment suppressed the expression of β-catenin, CyclinD1, C-myc, and GSK-3β ( Figures 7C-F) and increased the expression level of Axin-2 ( Figure 6B) which negatively regulates the process of Wnt/β-catenin signaling pathway. On contrary, the UTX demethylase inhibitor-GSK-4J activated Wnt/β-catenin signaling pathway, by up-regulating the expression level of β-catenin, CyclinD1, C-myc, and GSK-3β and down-regulating the expression of Axin-2 ( Figures 6B-F). Furthermore, the expression levels of β-catenin, CyclinD1, C-myc, and GSK-3β genes were signi cantly up-regulated with ethionine combined with GSK-4J treatment, while slightly down-regulated Axin-2 ( Figures 6B-F

Discussion
In this study, we reported that the depleted of SAM was involved in the pathogenesis of NTDs. Furthermore, experiments using a mouse model of maternal SAM depletion revealed that intracellular SAM depletion led to an impairment of histone methylation that was associated with activation of the Wnt/β-catenin pathway. Additionally, further analyses suggested that SAM de ciency could inhibit terminal mitosis of NSCs and hinder neuronal differentiation. Taken  Here, we had simultaneously given ethionine along with SAM to the pregnant mice. The results showed that the incidence of NTDs was signi cantly decreased, and the levels of SAM and SAM/SAH ratio were signi cantly higher in ethionine+SAM group compared with ethionine group. It was further con rmed that the NTDs mouse model were induced by SAM metabolic disorder. These results suggested that SAM played crucial role in neural tube development.
The epigenetic modi cation in the neural tube development had received increased attention in recent research, and speci c histone methylation spatiotemporal expression patterns might be essential for neurogenesis. Histone methylation played important roles in implicating chromatin modi cation and gene activation or repression, which depended on the methylation of site-speci c residue. Studies had reported that there were 24 histone methylation sites, including 17 on lysine residues and 7 on the arginine residues, in which H3K4me3, H3K36me3 and H3K79me3 were related to gene activation, and H3K9me3 and H3K27me3 were related to gene silencing [24][25][26]. The regulation of histone methylation modi cation were widely involved in neural tube development [27,28]. The modi cation level of H3K9me3 and H4K20me3 were signi cantly reduced after feeding with methyl de cient food in rats [29]. Previous studies also showed that histone methylation modi cation could directly or indirectly activate and inhibit gene expression, thereby affecting embryonic development [30]. It was revealed that valproic acid caused the increased of H3K4me3, and decreased of H3K9me3 in the NTDs mouse model, respectively [31]. In our study, the results showed that histone methylation modi cation was signi cantly decreased in ethionine-induced NTDs embryos, which were consistent with the alteration of H3K79me3 and H3K27me3 in human NTDs samples [27,32]. Moreover, we also found that the SAM metabolism disorders caused the alteration of histone methylation modi cations in NSCs, and the alteration could be improved by SAM supplementation. It was suggested that SAM metabolic disorder could affect the histone methylation modi cation, nally caused the occurrence of NTDs.
It was well known that Wnt/β-catenin signaling pathway was essential for neural development [33]. Some studies showed that β-catenin, a key transcriptional regulator, was closely relate with NTDs occurrence [34]. It was reported that SAM treatment reduced the expression level of total β-catenin, in dedifferentiated cells [23]. Moreover, β-catenin signaling was regulated by SAM level in hepatocytes [35]. Based on this, we speculated that the depleted of SAM affected the Wnt/βcatenin signaling pathway, and resulted in NTDs. In our study, the results showed that the expression of β-catenin, TCF-4 and CyclinD1 was signi cantly reduced in the ethionine group, and reversed by SAM supplementation. These results suggested that SAM de ciency inhibited Wnt/β-catenin signaling pathway. As above, SAM de ciency caused an abnormal histone methylation modi cation. Therefore, we speculated that SAM de ciency induced abnormal expression of histone methylation modi cation, inhibited the activation of Wnt/β-catenin signaling pathway, and thus participated in the occurrence of NTDs. Here, our results showed that UTX, a H3K27me3 demethylase, activity and the expression of βcatenin, CyclinD1, C-myc, and GSK-3β were signi cantly decreased, the expression of Axin-2 were signi cantly increased after the treatment of ethionine. On contrary, the UTX demethylase inhibitor-GSK-4J activated Wnt/β-catenin signaling pathway, by up-regulating the expression level of β-catenin, CyclinD1, C-myc, and GSK-3β and down-regulating the expression of Axin-2. The results indicated that SAM de ciency induced abnormal expression of H3K27me3, and inhibited the activation of Wnt/β-catenin signaling pathway. However, these ndings were not in agreement with the reports that SAM could inhibit β-catenin signaling in dedifferentiated cells [36,37]. Lastly, we evaluated the expression and activity of H3K27me3 in ethionine-treated NSCs. UTX, also known as Kdm6b, belongs to a small family of JmjC-domain containing enzymes that mediate demethylation of H3K27me3 repressive chromatin marks [38]. Our study showed that ethionine caused an increased UTX activity, which lead to increased H3K27me3 expression level. RT-qPCR results suggested that reduced UTX activate Wnt/β-catenin signaling pathway by increasing the level of H3K27me3. In short, we inferred that this inconsistency was due to SAM-SAH metabolic disorder and these abnormal metabolites were caused by ethionine. It was implied another possibility that SAM might stimulate Wnt/β-catenin signaling pathway through activation of histone methylation.
Wnt/β-catenin signaling had a major role in maintaining self-renewal as well as in regulating NSCs differentiation [39]. In order to investigate the association between the Wnt/β-catenin signaling pathway and the NSCs differentiation, we choose the primary neural stem cells extracted from E13.5 embryos as a cell model treated with ethionine. The results demonstrated that SAM depletion prevented NSCs differentiate into neurons and oligodendrocytes and inhibited Wnt/βcatenin signaling pathway activation. Furthermore, ethionine impaired NSCs differentiation which was partially rescued by SAM supplementation. These results reinforced the idea that SAM positively regulated NSCs differentiation. Similar notion was raised in a previous research revealing that SAM metabolism level was related to cell differentiation [23].

Conclusions
Collectively, our results uncovered the essential role of SAM in modulating Wnt/β-catenin signaling pathway during neurogenesis and normal neurodevelopment. SAM reduction disturbed NSCs differentiation and resulted in developmental disorders. Our research also revealed that the depletion of SAM led to reduced H3K27me3 modi cations and UTX activity downregulation, prevented the activation of Wnt/β-catenin signaling pathway and accelerated NSCs differentiation. These results provided a new understanding of the novel function of epigenetic regulation in NTDs which will have far-reaching implications in diseases and neurobiology. Hence, the mechanisms and in uences of these altered epigenetic histone modi cations needs to concrete studies in future.

Declarations
Author contributions LZ and XW wrote the manuscript. AK and FW helped in revising the manuscript. LZ, DL and ZL analyzed the data. KW contributed to the manuscript for literature research. JX, JX and BN revised and approved the manuscript. All authors read and approved the manuscript for submission.
Ethics approval and consent to participate Not applicable.

Consent for publication
Not applicable.

Competing interests
The authors declare that they have no competing interests.    with eithionine and SAM. Bar graphs for protein abundance were quantitative data from three independent experiments.

Figure 7
Ethionine inhibited NSCs differentiation. (A) At the level of neural stem cells, after the intervention with 20mM eithionine and 2mM SAM for 24h, the ability of GLAC, GFAP, MAP2, NF-H to differentiate was detected by Fluorescence microscopy analysis (P<0.001). Bar graph was the quanti cation data of density. *indicates signi cant difference compared with other groups in one-way ANOVA followed by Tukey tests. *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001. (B) GO functional classi cation of DEGs (corrected P value< 0.05). Top enriched GO terms of 10 cellular components, 10 molecular functions and 10 biological processes were shown.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.